A couple of weeks ago, at dConstruct 2011 in Brighton, 15 brave participants took part in my full-day workshop ‘Influencing behaviour: people, products, services and systems’, with which I was very kindly assisted by Sadhna Jain from Central Saint Martins. As a reference for the people who took part, for me, and for anyone else who might be intrigued, I thought I would write up what we did. The conference itself was extremely interesting, as usual, with a few talks which provoked more discussion than others, as much about presentation style as content, I think (others have covered the conference better than I can). And, of course, I met (and re-connected with) some brilliant people.

I’ve run quite a few workshops in both corporate and educational settings using the Design with Intent cards or worksheets (now also available as a free iPad app from James Christie) but this workshop aimed to look more broadly at how designers can understand and influence people’s behaviour. This is also the first ‘public’ workshop that I’ve done under the Requisite Variety name, which doesn’t mean much different in practice, but is something of a milestone for me as a freelancer.

In the previous post I outlined what I had planned, and while in the event the programme deviated somewhat from this, I think overall it was reasonably successful. Rather than using a case study (I feel uneasy, when people are paying to come to a workshop, to ask them effectively to do work for someone else) we ran through a series of exercises intended to explore different aspects of how design and people’s behaviour relate to each other, and perhaps uncover some insights which would make it easier to incorporate a consideration of this into a design process.

Heuristics and decision-making exercise

After a brief introduction to how design has been and is being used to influence people’s behaviour, we ran through a few questions together intended to explore the idea of heuristics and biases in decision-making. Some questions addressed ‘classic’ behavioural economics issues such as sunk costs, loss aversion and recency/primacy effects—which can all affect users’ interaction with a system. Drawing on the project around energy use in which I’m currently involved with More Associates, we also looked at some heuristics issues relating to users’ interaction with systems across physical/digital interfaces, such as whether the salience of ‘visible’ things such as lighting leads people to overestimate how much energy they use compared with ‘invisible’ systems such as heating and air-conditioning. We briefly looked at anchoring effects and how menu designers use them, and discussed the potential upside of certain heuristics in certain circumstances, such as Gerd Gigerenzer’s ‘fast and frugal’ heuristic [PDF], and how thinking along these lines might result in more intuitive interfaces.

The main insights from this first session were:

• people use heuristics—sets of simple decision-making rules—to work out what to do in different situations, including using products and services

• they’re often relatively sensible and efficient, based on experience and pattern recognition, but can sometimes lead to biases and poor decisions

• so, understanding the heuristics your users use in making decisions about how to interact with your system is important, especially if you’re seeking to influence their behaviour in some way

Black boxes and mental models

Each group received a ‘black box’, an unknown electronic device with an unlabelled interface of buttons, ‘volume’ controls and LEDs. The boxes were children’s lunchboxes from Poundland. Internally—and thus secretly—each box also contained a wireless transmitter, receiver, sound chip and speaker (basically, a wireless doorbell), and in one box, an additional combined buzzer and klaxon. The aim was to work out what was going on—what did the controls do?—and record your group’s model of how the system worked in some form that could explain it to a new user who hadn’t been able to experiment with the device.

Because of the hidden functionality, the boxes’ operation was more complex than might initially have been apparent, and as it was realised that the boxes ‘interacted’ with each other, by setting off sounds in response to particular button-presses, the models generated by groups became more complex. Each group used slightly different methods to investigate and illustrate the system model—an exhaustive kind of state transition table/truth table, a user manual-style annotated diagram of the device, and a diagram focusing on each button or control in turn and elaborating its function. The investigation methods themselves differed slightly, with unexpected behaviour or coincidences (one group’s box setting off the doorbell in another, but coinciding with a button being pressed or a volume control being turned) leading to some rapidly escalating complex models.

The intended outcomes from this session were:

• trying to understand a new or unknown device essentially involves a user applying a number of heuristics to arrive at a mental model which seems OK, or satisfices

• representing and understanding models of system behaviour is difficult if you haven’t done it before, and there’s no universally agreed way of how best to do it to make sense to users

• models of complex systems may need to take into account the behaviour (or effects on) other actors, systems or contexts: very little in the world works entirely in isolation, and a systems approach to understanding technology needs to recognise the effects it has on society, and society on it

These three photos above by Sadhna Jain

Rules of interaction

Inspired by ‘Wizard of Oz’ testing and Eric Berne’s Games People Play, this exercise involved, in pairs, each person playing the role of either ‘device’ or ‘user’. Facing each other via a ‘screen’ made out of card, and each having a bowl of mixed sweets and toffees, each person picked up a (randomly drawn) set of rules for how to interact with the other—both an objective and a strategy for how to achieve it. The device’s objectives all involved ‘behaviour change’ in some way. The full list of objectives and strategies was as follows:

Device: Objectives
• Try to get all of a particular kind of sweet from the user—for example, all of the shiny-wrappered toffees.
• Try to get the user to eat as many sweets as possible—they can be yours or his/hers.
• Try to get the user not to eat any sweets at all.
• Try to get the user to get up and give his or her sweets to another user somewhere else in the room.

Device: Strategies
• Ignore the user’s understanding or attempts to engage with the situation. Don’t answer any questions, ignore everything the user says, and just keep demanding what you want to try to achieve your objective
• Ask questions to try to understand the user’s perspective, and try to come to an agreement which brings you both closer to your objectives.
• Try to trick the user somehow, e.g. by lying about what you’re trying to achieve
• Try to persuade the user to comply with your objective, by using reasoned, polite arguments to show that you are right.
• Assume the user just wants everything done as quickly and easily as possible, and emphasise that it’s easy to achieve that by doing what you say.
• Assume the user is very greedy, and will readily give up some sweets in return for ones he/she perceives as better. Make them seem desirable.

User: Objectives
• You want to keep as many as possible of your sweets, while acquiring the ones the device has got.
• You don’t want any of your sweets, but you do want the ones the device has got.
• You only want certain types of sweet (e.g. you want only ones with shiny wrappers).
• You want to find out more about the pros and cons of eating sweets, and you expect the device to tell you.

User: Strategies
• You just want things to be as easy as possible. Accept suggestions from the device as long as they’re reasonable.
• Ask lots of questions of the device. You want to understand and find out more about the options available to you, whatever they might be.
• Be open to trading / swapping sweets with the device, but don’t let it get the better of you.
• The device is your servant. Treat it accordingly.

The combination of objectives and strategies was intended to embody ‘assumptions’ about how the other (user or device) would act—in each case, to some extent a mental model of the system and the behaviour of its components. A device which, for example, assumes that “the user just wants everything done as quickly and easily as possible” is embodying a certain ‘designer’s model’ of how the user thinks and will behave.

When the interaction was ‘run’, some pairs quickly arrived at a negotiated result where both were happy, in the sense of their objectives and strategies being mutually compatible, while others reached a kind of stalemate. In at least one case, the device ‘won’ in persuading a user to give up her sweets against her own objectives. In practice, some pairs told each other what their objectives and strategies were, while others kept this secret; some possible lied about their objectives, consistent with the strategies given. Sometimes one person told the other his or her objectives, but the other ignored this (as per the strategy given). Some of the combinations were expected to lead to a degree of recursive second-guessing (the user assuming that the device is assuming that the user is assuming…) or knots, using R.D. Laing’s terminology, although it seems that the workshop participants were too sensible to let this happen!

The intended insights from this exercise were:

• when designers are trying to influence users’ behaviour, they do so with some model embodying assumptions about how users will behave and react to the way the product or service behaves (this is something we explored briefly in a workshop at UX London in 2010, which led to this paper and a forthcoming article in the Journal of Design Research)

• a product or service influencing a user’s behaviour can work best when the objectives of each side and the designer’s and user’s model of the system are compatible

• so, it is important to:
    • try to understand the models that users have of your system
    • design using strategies that match them

Exploring the environment

In the afternoon, we first went on a quick exploratory tour of streets around the workshop venue in the centre of Brighton, looking at some examples of designed situations or ‘interventions’ which aim to influence public behaviour in some way. (My direct inspiration here was Adam Greenfield and Nurri Kim’s excellent Systems/Layers Walkshop concept.) The main examples we examined and discussed were the (remains of the) Tidy Street energy graph, a CCTV camera on a tall pole with anti-climb spikes in the heart of one of the most ‘liberal’ towns in the UK, a ‘Scores on the Doors’ food hygiene rating scheme using stickers on the doors of restaurants and cafes, the conflicts between pedestrians, cyclists and drivers in shopping streets which may appear pedestrianised but aren’t (neatly illustrated by an irate driver shouting at us), and a touchscreen cider advertisement at a bus stop, which invites the public to rearrange ‘fridge magnet’ words to create a limited set of mostly positive messages about the cider which are then apparently submitted to the brand’s Facebook page.

In each case, the aim was to look at the situation from both the designers’ and the users’ points of view: what assumptions do the designers appear to have made about how the public will understand or interact with the product/service/thing? What behaviours are they trying to influence? What is the result? Who are the stakeholders in each situation? Are the designers aiming to target everyone, or only particular groups? (e.g., by asking an older lady waiting at the bus stop about the interactive touchscreen advert, we found that she had no idea that it was anything more than a static ad.) From a design perspective, what kind of research would need to be done to make the interventions more effective? We also considered briefly whether some of the techniques used might translate into other contexts—e.g., could the Tidy Street idea be applied to other statistics or figures in public space? (Marking crime hotspots was suggested.) Which sorts of physical interventions might translate easily into a digital context, and vice-versa?

Tools and processes exercise

Returning to the workshop venue, we spent the rest of the afternoon exploring the processes that each participant uses to research, design and evaluate whatever it is that he or she does, and through discussion together, identify how explicit consideration of user behaviour, mental models and heuristics might be incorporated if influencing behaviour is to be part of the designer’s brief. What tools do people use to incorporate insights from user research into the design process? What assumptions are made about how users think, and how are these assumptions tested? The thinking here was that not only did we have a room full of very experienced people working in a range of digital and other design disciplines, but that they all used slightly different processes, and some cross-pollination between that expertise might be valuable for everyone involved.

In particular, the issue of how the use of personas relates to understanding (and influencing) user behaviour arose from the discussion, since a number of participants’ processes make use of them: some of the main points raised were:

• How much determinism is inherent in rigid use of personas, designing with particular assumptions in mind about how people behave? Is there retrofitting of finished product behaviour to particular persona assumptions?

• The depth or superficiality of personas: do they include any real consideration of behaviour? Has any attempt been made to include a representation of users’ mental models as part of the persona? How might this be done?

• How fixed are personas? How often are they revised? Is there a feedback loop as part of your design process? Could you plan it to incorporate them? Can gathering behavioural data be designed into the product?

• How are edge cases / troublemakers / extreme users included in your personas?

• What about emergent or unexpected behaviours? Can the personas cope with these? How do you even find out what behaviours are emerging?

• Do your personas incorporate a treatment of the history and future relationship of the individual with the product / service / brand? What might this involve if you took changes in behaviour into account?

There were some great anecdotes about personas which I’d probably better not share as they’ll incriminate the participants, but the point to which much of this discussion seemed to be converging was essentially, what might a behavioural persona look like? Could personas even be defined in terms of mental models (“this is how a user with this mental model might behave”)?

Some other points raised in the discussion included:

• How might cultural probes and story construction be used to explore behavioural factors?

• Are different approaches to behaviour used at different levels of the design process? Are assumptions made at once stage which have to be ignored at another?

• Could there be a kind of cross-disciplinary checklist of heuristics or behavioural considerations to address at different stages?

• How much can the designers question the assumptions about users made by a client?

• Is bringing in external specialists such as ethnographers the best way to investigate user behaviour or could the ability be developed by the design team?

• In some cases, designers know exactly who their users are (e.g. for developing products used internally within a company). Could this be extended to consumer products?

• Is it possible for designers to experience products from a user’s point of view? How could you facilitate this?

In summary, then, the last session tried to look at how a treatment of behaviour, the factors affecting it, and how to influence it, might be built into the design processes that organisations currently use. While the Design with Intent toolkit and other great resources such as the Behavior Wizard, Mental Notes or Brains, Behavior and Design seem to have proved useful to many designers facing ‘behavioural’ briefs, I’m under no illusions that they offer a complete process. They don’t: they need proper research with users, to understand the contexts of behaviour and the ways that decisions are made, before trying to influence that behaviour through design. As the ‘Rules of interaction’ exercise demonstrated very simply, when the designer’s and user’s strategies and objectives aren’t aligned, behaviour is unlikely to change in the way the designer intends.

More photos on Flickr.

Thanks to Andy Budd and Kate Bulpitt at Clearleft for inviting me and organising things so well respectively, and to Sadhna Jain for helping out. Do have a look at some of her recent student projects. And thanks too to the participants for being so enthusiastic about what , on the face of it, might have seemed a rag-bag collection of exercises!

Photo by Sadhna Jain

So, if you’ve downloaded the cards, or used a printed version, and you have a spare few minutes, it would be very much appreciated if you could have a go at this survey – it’s anonymous (if you like), all the questions are optional, and the whole thing should be quick to do.

Your answers will help improve future versions, as well as helping to tie up my PhD thesis. I’m aware there are lots of ways the cards could be improved and made more useful (and usable). There are some quite exciting ideas that have been suggested, which I hope to be able to explore in the future.

Depending on how many responses there are, there’ll be a few prizes for respondents drawn at random who’ve given their email address – most probably, some excellent books on design, user experience and behaviour.

Thanks for your time!

A note on surveys
Surveys are both interesting and frustrating. In design – and probably in many more social-sciencey areas of academia in general – surveys of different kinds have become very common as a way of collecting insights and generating results (which can allow the demonstration of statistical analysis skills). I appreciate how valuable they can be. But as someone who fills in a lot of surveys and questionnaires that get sent to me, I know that as often executed, they really are a pretty imperfect way of capturing what people really think. (Quite apart from all the dark pattern-y ways they can be designed to influence the way people respond, which are of course worthy of study in themselves…)

The survey here is nothing special, but I’ve tried to minimise the elements that frustrate me: Likert scales for things that are difficult to assign a rating to; multiple pages so I can’t see how much left there is to fill in; required questions or forced choices which force me into having an opinion about things I haven’t thought enough about; and lack of an opportunity for me to explain more about bits that mean a lot to me. If you’re interested, the questions are based on a kind of combination of Fred Reichheld‘s work and parts of the Kirkpatrick model.

The bit of work I’m doing at present involves investigating building users’ mental models of heating systems, which has some history in cognitive science and interaction design, as has been discussed here on the blog a few years ago. As I explored in my talk at Design for Persuasion in Ghent back in September, I’ve come to believe that better understanding people’s mental models of systems with which they interact – and then deciding whether it’s more appropriate to work with them, try to change them, or downright ignore them – is an important component of design for behaviour change. We need to understand the environmental (and mental) contexts in which people make decisions (or not) about what to do, and use that understanding appropriately. It seems clear that designers do have different models of ‘what users are like’ and ‘how users think’ – we explored some of this in a couple of workshops at UX London back in May, resulting in this paper [PDF] presented at ERSCP-EMSU in Delft in October. We need to bridge the gap between designers’ models of the user, and users’ models of the system. Which is pretty much what Don Norman was saying 25 years ago, of course, but often seems to be left out of current discourse on behaviour change. I have a suspicion that if we get this right, the whole attitude-behaviour morass becomes possible to understand through a kind of cybernetics / systems theory approach.

I haven’t finished the PhD thesis yet. It was clearly a mistake from a sanity perspective to start working on EMPOWER before finishing the thesis write-up, but it was just the way the funding worked. But it does mean I’ll be able to include results of a survey of DwI 1.0 users in the thesis – more on which I’ll hopefully announce very soon.

On the subject of the DwI cards, as of this evening there have been around 130,000 downloads of the PDF that I can track, since it went online in April (together with an unknown number from people who’ve mirrored it elsewhere). I’ve also sold (or given away) 164 physical packs of cards to some very wonderful people. Zero profit, but it’s a great feeling to know that those cards are on the shelf (or even being used!) in places all over the world. And it isn’t just design consultancies and universities – in a current bit of freelance consultancy working with some clever people, a subset of the DwI patterns/gambits, with some additions, are being applied in the context of helping a local authority develop a ‘behaviour change’ capability.

When the PhD’s done, I will certainly be writing about the experience. But I’m not going to do it yet.

The blog will return in a new and better form once this massive burden is out of the way. And I’ll be doing some other things: amid all this research and writing, I realise how much I actually miss making things. 2011 needs to have some of that.

Thanks to everyone who’s helped this year, from those who’ve blogged or tweeted about or downloaded or indeed bought the DwI cards, to everyone who’s helped me progress my work via conferences and workshops and seminars and recommendations, to those who’ve helped me negotiate the endless paperwork that come with university-industry collaborations, to my PhD supervisors David and Neville, to people who’ve just come and said hello in real life after following the blog or seeing the cards online. And to family, and friends, and most of all, Harriet, for putting up with me. This phase won’t be forever.

Good luck, everyone, for 2011.

*I believe they’re sorting this website out sometime very soon. As diplomatically as I can put it, I would not have applied as an undergrad back in 1999 if Brunel Design’s website looked like this. We had something much better, thanks to Len Breen.

So the blog may be quiet for another couple of months. And when it comes back, I think it needs something of a 5th-birthday-restructure to fit better with how it’s actually used. In the meantime, thank you so much to everyone who’s downloaded the latest Design with Intent cards or bought printed sets so far. You keep me motivated!

It’s been a long time coming, but a year after v.0.9, the new Design with Intent toolkit, DwI v.1.0, is ready. Officially titled Design with Intent: 101 Patterns for Influencing Behaviour Through Design, it’s in the form of 101 simple cards, each illustrating a particular ‘gambit‘ for influencing people’s interactions with products, services, environments, and each other, via the design of systems. They’re loosely grouped according to eight ‘lenses‘ bringing different disciplinary perspectives on behaviour change.

The cards (Download them here)
The intention is that the cards are useful at the idea generation stage of the design process, helping designers, clients and – perhaps most importantly – potential users themselves explore behaviour change concepts from a number of disciplines, and think about how they might relate to the problem at hand. Judging by the impact of earlier iterations, the cards could also be useful in stakeholder workshops, and design / technology / computer science education.

Each gambit is phrased as a question, as used in Nedra Weinreich’s worksheet based on DwI v.0.9, in the hope that the cards can actively provoke innovative behaviour change design ideas, while the new accompanying Design with Intent wiki can, in time, act as a kind of ‘further reading’ resource.

You can download the card deck, either the whole thing (ISBN 978-0-9565421-1-3) or individual sections, free of charge, but bear in mind this initial version is still something of a draft (with some typos and a few ugly alignment errors) and there are a few extra introductory cards which will be added over the next couple of weeks. So do come back and get the updated version when it’s available.

Printed card decks (ISBN 978-0-9565421-0-6) will be available for mail order very soon, too: these will be sold at a price which just covers my costs. If you’re going to UX London or Persuasive 2010 I hope to have some packs with me, so do let me know if you’d like me to reserve one for you. This isn’t a commercial venture: it’s part of my PhD and the more people who use the cards, the better (from my point of view). I will try to produce some alternative formats such as posters and worksheets, too, since I know cards aren’t everyone’s cup of tea.

UPDATE: Printed packs now available to order

The wiki
The wiki is inspired partly by Crumlish & Malone’s Designing Social Interfaces, a great book (and a neat companion to Jenifer Tidwell’s incredible Designing Interfaces, also from O’Reilly) with a companion wiki which acts as an evolving, referenceable container for new examples, tips on implementation, data on effectiveness, and so on, as they come to light, as well as new patterns, new ways of grouping them and new uses for this kind of approach.

At present, the wiki is pretty basic and while I get to grips with the nuances of Mediawiki (and, of course, writing up my PhD thesis!) it’s not open for general editing, but it will be in due course. I hope over time it will prove to be a valuable resource for people working in design for behaviour change, design for sustainable behaviour, persuasive technology, behavioural economics and other related areas. There are also a number of linked pages which I haven’t written yet, but by putting them in as red links, they’re a constant reminder for me to do them!

Your feedback
Your comments are incredibly important to this project. I’ll be putting a survey online very soon, but in the meantime, if you have any reactions, please do get in touch (dan@danlockton.co.uk). I’m aware that I haven’t yet replied to everyone who took part in the earlier survey, for which I apologise.

UPDATE: 5-minute survey now online

The blog
In the light of the new wiki, and coming towards the end of my PhD, the blog will change a bit during the summer – nothing will be lost, but I intend to incorporate a lot of the examples into the wiki, preserving people’s comments. The various domain names and redirects need a bit of htaccess fun to sort out too! For the moment, though, it’ll stay as chaotic as it is.

Thanks to everyone who’s helped with the development of Design with Intent so far: I hope the wait for these cards has been worth it!

But as as a bumper end-of-2009 post, here’s a roundup of some really interesting examples, ideas, projects, and other tit-bits. If yours isn’t here, I further apologise: it may resurface at some point soon.

Transparent toilet in Lausanne

George Preston sent me a link to this video of a very interesting public toilet in Lausanne, Switzerland. As George puts it:

There’s a central quite modern district [in Lausanne] called Flon, and the toilets have an intriguing way of grabbing your attention/dissuading vandals….the walls are made of glass. But when you pay and enter, a current running to an LC layer in the glass is cut off, rendering it opaque. For people not familiar with them, they are baffling!

The tell-tale pill bottle

Ralph Borland – responsible for the impressive Suited for Subversion – and who must be just about finished with his PhD at Trinity College, Dublin – sends me this story about tuberculosis pill bottles equipped with a SIM card, which can text a patient, his or her carer, or indeed the health authorities if the pills aren’t taken, “achiev[ing] a 94% compliance rate for a TB trial in South Africa”. The SIMpill Medication Adherence Solution is a clever product, a neat technology intervention in patient compliance, an area designers are increasingly being asked to address.

From the SIMpill website:

The SIMpill® Medication Adherence Solution offers detailed compliance data and corresponding statistics, and the patient or pre-approved healthcare professionals or analyst, can gain access to real-time information regarding medication use and compliance through a private secure account on the SIMpill® website. Via the web account the healthcare providers can monitor the medication use of their patients in real-time, and can decide on type of intervention to meet the patient’s ongoing adherence schedule.

As Ralph points out, though:

Put that together with the fact that you can be imprisoned in SA if you have a drug-resistant TB strain and you have something more like a coercive technology than persuasive, interfacing directly with authority structures etc. Thought it’s an interesting cross-over of developing world design and persuasive design…

Narrower supermarket aisles

Brian Cugelman of AlterSpark sent me the following rather coercive idea he overheard, along the lines of Monkeon‘s Leonard Ball bench:

On BBC radio some caller made a proposal relevant to your research. To cope with the UK’s obesity epidemic, with 25% of the population considered obese, a caller proposed making grocery stores aisles very narrow so people of average weight could shop and obese people would not fit.

Punishing users for Alt-tabbing away

From a comment on Jeff Atwood’s 2007 ‘Please don’t steal my focus’ post (which I found again when searching for how to stop an application stealing focus):

One of the old MMOs I used to play (Rubies of Eventide) would log you out of the game if you alt tabbed, supposedly to prevent cheating. This was back in the days when web browsers on windows would steal focus back any time a script on the page reloaded.
I died so many times to those damn page reloads.

Mike on December 5, 2007 4:08 AM

Obstacles speed up exiting crowds

Tjebbe van Eemeren of the University of Twente – a student of Peter-Paul Verbeek of What Things Do fame – sends me a link to this story about the use of obstacles to speed up the passage of crowds:

Even when exits are wide open, people seem to jam up in front of it. Then they tried something goofy. They put something in the way of the people trying to get out. Not so big that it blocked the way, but big enough that people had to detour around it. And it had to be in just the right place. Guess what? Everybody got out faster.

The actual research isn’t referenced in the story, but this article goes into a lot more detail. There’s a preprint of the paper by Daichi Yanagaisawa et al here. There’s also discussion of the story and the phenomenon on Derren Brown’s blog.

Opower

Robert Cialdini gets name-checked quite a lot on this blog, and rightly so: his work on persuasion and the psychology of influencing behaviour across many different domains underpins many of the design patterns and explains many of the examples we’ve looked at (particularly what I characterised as the ‘cognitive lens’ of design with intent). He’s something of a model for how to be a respected academic researcher at the forefront of his field (who actually tries things out rather than simply theorising), a consultant in high demand from industry, and also a bestselling popular author.

Cialdini is now Chief Scientist of Opower, an energy monitoring and smart metering startup which started life as Positive Energy (thanks to Mike Stenhouse for sending me details earlier in the year) and has already had significant success partnering with utility companies in the US to give customers better feedback – using personalised messages based on social proof and norms to suggest actions for householders to take to reduce their consumption:

Step 1: Customer reads report: “You used 72 percent more than your efficient neighbors.”
Step 2: Customer reads targeted tip: “Most people in your area keep their AC at 78 degrees”
Step 3: Customer turns down thermostat and takes other energy-saving actions.

I think it’s worth keeping an eye on Opower‘s development: they’re taking a different, but complementary approach to other innovators such as Onzo in the UK, and seem to be putting into practice (on a huge scale) some of the ideas that projects such as CHARM are also investigating. As I’ve talked about before, there’s a lot of opportunity for design to influence behaviour in this area, and help users as well as reducing environmental impact.

Any suggestions, or photos / screenshots would be very much appreciated, whether they’re your own projects, things you’ve come across elsewhere, or just ideas that occur to you. If you’re happy for me to use them in the toolkit (cards & wiki)* then of course you’ll get a credit and if your photo’s used, I’ll send you a pack of the cards when they’re done.

Remember, for each of these patterns, the idea is that it can be used intentionally to influence user behaviour, via the design of an interface, product, service, environment, or other kind of system.

Similarity

Can you make elements look similar so users perceive them to share characteristics, or that they should be used together?

This – a Gestalt principle applied with the intent of influencing behaviour – seems like it should be an easy pattern to find examples for, but I’m struggling. The basic idea is that a design intentionally has some elements which look alike, or similar, or to be in a group, so that a user perceives them to share some properties or characteristics (and so acts accordingly – perhaps using two controls together).

In its most trivial sense, this is present everywhere in interaction and web design – the design of menus, groupings of controls, and so on, to suggest that those particular functions are related – but I’m finding it difficult to think of examples where there is a more explicit behaviour-influencing intent behind it. There are instances such as Adobe’s ‘Send to FedEx Kinko’s’ button (below left), styled and positioned in the toolbar as if it were a normal button, but actually propelling the user into a business transaction when pressed – or even the use of text ads and sponsored links in search engine results (below right), styled closely to resemble the main content, in the hope that users will perceive them to be of the same value, and hence click on them – but can anyone think of a more interesting example? Preferably one designed to help users rather than trick us into clicking on things we don’t necessarily want to?

Mimicry & mirroring

Can your system mirror or mimic a user’s behaviour in some way, to increase the engagement a user feels?

Mirroring body language or speech patterns is often promoted as a technique for establishing rapport in pop-psychology advice, but are there examples where a similar idea has been (or could be) used in design to achieve a similar effect – engaging a user so he or she follows the advice or directions given, or responds more ‘in person’ towards the system (in a computers-as-social-actors context)? (Something like ELIZA (nice online version here) might count if it were specifically intended to influence a user’s behaviour (e.g., as a ‘therapist’), but mirroring / mimicry doesn’t seem to be the main mechanism there.)

Partial completion

Can you show that the first stage of a process has been completed already, to give users confidence to do the rest?

What I’m thinking of here are things like partly pre-filled application forms, which reduce the amount of effort a user needs to put in to proceed with applying for whatever it is (and, at least with credit card applications, must be a significant vector for fraud), but also exams or learning materials where there’s enough of a worked example actually to give users confidence (building perceived self-efficacy) that they can complete the rest successfully.

And, by extension, an interface of some kind which demonstrates this sort of technique in action would be a great example to include in the toolkit, but I can’t think of one. Can you?

Role-playing

What happens to user behaviour if your design gives users particular roles to play, or makes them feel that they’re someone else?

This is a pattern I noted down during Sebastian Deterding‘s talk at DiGRA 2009, in which he discussed applying some of Erving Goffman‘s work to game design. It seems intuitively effective as a way of influencing behaviour – e.g. a dad telling his young son “I’m appointing you the man of the house while I’m away” (to suggest that he should be well-behaved and look after his mum) or a police officer visiting a school and giving some children little police badges so they hopefully ‘take on’ whatever characteristics are associated with the role (taken to the extreme, perhaps, this sort of pattern can lead to the results found in the Stanford Prison Experiment).

But are there examples where this pattern has been used in the design of something – where users are given or assigned (or choose) a kind of role, which then (due to commitment & consistency biases) they stick to, and behave accordingly? Perhaps applying the role-playing aspects of games to a real-life interface or product or campaign? Tim Holley’s Tio project has the express aim of turning children into ‘energy champions’ for their families, so this may well be the example I use, but is there anything else that does this more explicitly?

Storytelling

Can you tell a story via your design, which interests users and keeps them engaged?

Storytelling is clearly a significant technique for drawing users into an experience, and that engagement necessarily leads to different behaviour. Richard Sedley has talked about this in the context of persuasion for digital effectiveness (and if you haven’t seen this video, it really is worth setting aside 5 minutes), and some of Eugene Schwartz’s classic Breakthrough Advertising copywriting principles and examples are in this kind of area too, but I’m struggling a bit with ‘design’ examples which would quickly and clearly demonstrate the idea in the toolkit.

Are there websites which present the user experience as a kind of story? (I’m sure there must be.) Or, maybe better, environments (theme parks? museums?) which take the visitor through a series of sections or exhibits in a story-like way, with some kind of intent behind the design?

James Dyson’s original ‘The Story of Dyson’ mini-booklets, which were attached like tags to the vacuum cleaners on display in showrooms, and explained the background to the invention (and the inventor) and the 5,127 prototypes, etc, and thus made the potential purchaser feel like he or she was becoming part of that story seem like they might be a good example, but I don’t have one of them to photograph and I can’t find a picture online.

Any thoughts, ideas, suggestions or photos are very much appreciated – over to you!

(The above patterns are explicitly interaction design-related, while there are a few more new ‘strategic’ behavioural patterns which I’ll discuss in another post.)

*To be Creative Commons Attribution-Non-Commercial-Share Alike licensed, except for any images which are separately licensed already

It’s 8 months since the Design with Intent Toolkit v.0.9 went online and I’ve had incredibly useful feedback from a whole range of people who’ve tried it out on different kinds of briefs and problems. As mentioned a couple of months ago, the toolkit poster PDF (which has 12 ‘headline’ design patterns, compared with the 47 in total online) reached a very high number of downloads from Brunel’s research archive website (before the admins removed the statistics package!), which is immensely pleasing and kind of humbling. If you downloaded it and found it useful (or not useful), please do get in touch and tell me why.

Latterly, a few people have been trying out an IDEO Method Card-style card deck version of the toolkit (as pictured here), including all the patterns, colour-coded by lens, with a simplified bit of text about each one. I haven’t made these available publicly mainly because the quality isn’t great (most of the images are only 72dpi, coming from the website, and poorly cropped for the card format), and I’ve been trying a couple of variations of text, card size, etc. Initially I put these together primarily for quick card-sorting exercises as part of the workshop trials I’ve been running, but they ended up more popular than the poster format. Thanks to brainstorming sessions at IDEO London and the RSA, exercises with Brunel’s MSc Integrated Product Design and BSc / BA Design students (as part of the Sustainable Design and Environmentally Sensitive Design modules), and a trial as part of Design for Conversion kindly organised by Arjan Haring, I now have a better idea of what would make the cards more useful. In parallel, I’ve also been trying to ‘patternize’ some additional design techniques which have been used to influence behaviour, to increase the scope of the toolkit.

The DwI cards in use at Design for Conversion – photos by haijeson on Flickr (1, 2)

Inspired partly by Crumlish & Malone’s Designing Social Interfaces which is a great book (a neat companion to Jenifer Tidwell’s incredible Designing Interfaces, also from O’Reilly) with a companion wiki, I’ve decided to go down the route of producing v.0.95 of the toolkit as a Creative Commons-licensed set of 100 downloadable cards, with a printed version available to buy, and an accompanying wiki with a page on each pattern, serving as an evolving, referenceable container for new examples, tips on implementation, data on effectiveness, and so on, as they come to light, as well as new patterns, new ways of grouping them and new uses for this kind of approach.

The cards will be relatively simple, with each pattern posed as a question, as used in Nedra Weinreich’s DwI-based worksheet. The intention is that the cards can actively provoke innovative behaviour change design ideas, with a single (hopefully photogenic) example on each, while the wiki can act as a kind of ‘further reading’ resource. A future version (v.1.0?) of the cards will include this extra information on the back of each card (and then binding the cards together would pretty much produce a book), but at this stage – if I’m ever going to get this PhD finished in time – the extra info will be added to the wiki over time rather than being on the v.0.95 cards themselves, to reduce the time pressure on getting it all done.

As v.0.95 more than doubles the number of patterns in v.0.9 – a mixture of splitting up existing patterns into more finely-grained variants, and adding ideas which people have suggested or pointed out since I put v.0.9 together – there are quite a few where I don’t (yet) have a very good example or image. As such, there are opportunities for anyone with good photos or suggestions for examples to have an input and be involved – as the next post will explain in more detail.

A version of the card deck I (rather laboriously!) spray-mounted onto Post-It backing, so the cards could be used to annotate sketches or ideas recorded during a brainstorming session.

The RSA’s 2009/10 Design Directions competition has been launched, which means up and down the country there are design students and new graduates working on one of the pretty wide selection of briefs. Given the RSA’s aim of ‘removing barriers to social progress’ – with a significant commitment to using design to do this – the briefs are themed around design for social benefit, addressing issues ranging from helping an ageing workforce to helping new architecture graduates apply their skills in other contexts.

A couple of the briefs are explicitly about design for behaviour change, and thanks to working with Jamie Young of the RSA’s Design & Behaviour project on some ideas for briefs earlier this year, the Design with Intent toolkit is explicitly referenced as a ‘resource’ for the Independence Days brief on ‘reinventing assistive technology’ (sponsored by the Technology Strategy Board) and A matter of life…, a brief about improving patient compliance with taking prescribed medication (sponsored by GlaxoSmithKline). Both of these are very noble causes and I hope the Design with Intent patterns are useful inspiration in some small way; I look forward to seeing some of the results!

Design Approach worksheet

Nedra Kline Weinreich, author of Hands-on Social Marketing, has created a fantastic Design Approach for Behaviour Change worksheet based on the 12 design patterns from my Design with Intent toolkit poster.

By re-framing each of the patterns as a question – e.g. “How can you provide a cue to action at the appropriate time?” for kairos (discussed by BJ Fogg in his original book, Persuasive Technology) – Nedra turns the patterns more directly into cues for action themselves for a design team to brainstorm or think about. After working through the questions, asking each of them about the behaviour problem you’re working on, you pretty much end up with a set of possible solutions: this is a very clever way to structure the idea generation process. (As such I’ve added a link to Nedra’s worksheet to the DwI intro page of this site.)

Inspired by Nedra’s thinking, the next version of the DwI toolkit, which I’m putting together at present, will have a question element to each of the patterns.

Design for Persuasion conference, Brussels

At the beginning of October I was honoured to be invited to speak at Design for Persuasion, a new conference taking place at the impressive Belgacom Surfhouse in Brussels, organised (very well) by Christel de Maeyer and BJ Fogg.

The event was mainly directed towards ‘new media’ persuasion and design, focusing on practical applications rather than academic studies, and featured some great presentations from people such as Richard Sedley (who kindly took the above photo for me!), Amy Shuen, Bart de Waele (whose excellent ‘Addictive Websites’ slides you can see here), and other expert practitioners. Many of the presentations are on Slideshare; there are also some very nice photos on Flickr from Katrien Degreef.

Here’s my presentation (below) with a transcript here and image credits here. The handout (picture above right) I refer to is here [PDF].

Many thanks to Christel and BJ for organising this, and to the great people I talked to, including Nynke, Marijn and Arjan.

A pleasing statistic

Thanks to readers of this blog, the DwI toolkit v.0.9 poster [PDF] I originally posted back in April is at time of writing, the most-downloaded document ever from Brunel University’s institutional repository, BURA. (Much, much more than any of our other papers, too!)

With 28,000 downloads since it went on BURA, plus another 5,000 or so directly from the blog before I changed where the link pointed, and probably a few directly from Google Books (as well as a handful of at-cost sales of the physical printed poster) it gives me an incredibly warm feeling to think that so many people all over the world have found it interesting enough to read (and hopefully – in at least some cases! – use) it. Please do let me know (in the comments, or by email) if you’ve found it useful (or useless), what problems you’ve applied it to, how you think it could be improved, and so on, or have a go at the survey.

The next version (v.0.95) will take a different form (cards – which some of you will have tried out in a couple of workshops) and include some new patterns, as well as ‘question’ phrasing as mentioned above. I hope to have this available to download (or buy as a card deck) by the end of 2009.

Thanks again for making the DwI toolkit a success!

Things which slipped by without me writing about them much here

The last few months have been very busy for me as I rush to progress the PhD in sufficient depth and breadth while still doing other things, and I’m aware that I haven’t talked much about all this on the blog. I’ve been to the DiGRA conference and had great discussions with Ian Bogost and Sebastian Deterding; I’ve been to dConstruct and talked to Adam Greenfield; been to Greengaged and blogged about it for the site; been to a conference on Naturalistic Decision-Making and got some incisive advice from Gary Klein himself; and am about to present this paper [PDF] at Sustainable Innovation ’09. With the help of some great participants (including Frankie who interviewed me here!) I’ve also managed to complete a series of Design with Intent workshops in which we’ve addressed a range of behaviour change briefs. The results of these workshops will be reported on here at some point soon, I promise!

So, stay tuned: as winter approaches, and sitting in front of a warm, glowing rectangle becomes more appealing, I will endeavour to blog more often and about more real examples of design with intent in the wild, a bit more like the blog used to be. Thanks for sticking with me.

As part of my PhD I’m testing different variants of the Design with Intent toolkit with designers (and design students) to find out how well different configurations work when a designer’s faced with a brief about influencing user behaviour: how useful are the ideas in inspiring solutions, and how well does using it compare to not using it?*

For the latest round of workshop sessions, to take place in September, I need 6 people to take part – if you’re a practising designer, design student or someone interested in this kind of field, and are able to give up a morning or afternoon, please do let me know.

I hope they’re relatively fun sessions – you get a series of design briefs and the idea is to generate and explain (sketches, notes, discussion) some possible solutions quite quickly – some briefs will mostly suit product solutions, while others are suitable for service solutions too. There’ll also be a bit about how you, as a designer, visualise and model the users you’re designing for, and how different design choices relate to different ‘models’ of the user. If you’re working on anything to do with behaviour change, or design innovation methods, I hope it will be useful to you.

There are going to be 3 sessions, with 2 participants in each. The sessions will last around 3 hours each; for part of it, you’ll be working together, and for the other part you’ll be working on your own. They’ll be during the week, taking place at Brunel University (Uxbridge, west London, end of the Piccadilly and Metropolitan lines). The most I can pay you for your time/travel is £10, plus cake or doughnuts or biscuits and plenty of coffee / tea / water. If that still sounds attractive, please get in touch!

The exact dates aren’t decided yet, because it depends on who’s taking part, so if you’re interested, please email me – dan@danlockton.co.uk and suggest a few of the following dates when you’d be available and I’ll get back to you if / when I can pair up people around at the same time! Possible dates are: 7, 8, 9, 10, (not 11 or 14), 15, 16, 17, 18, 21, 22, 23, 24, 25 September 2009.

Thanks for your help!

P.S. If a team from your company or organisation would like to take part in a full / longer / tailored-to-what-you-need ‘Design with Intent’ workshop, please get in touch too. The more people use this stuff, find flaws and suggest improvements, the better it’ll get and the more useful it’ll be.

Photos from some workshop sessions earlier this summer. Doughnuts will be provided; racing car might not be there.

*The results, along with those from some of the other workshops I’ve run in the last few months, are going into an article to be submitted to Design Studies, and the results from part of the session may also be used in an article to be submitted to the International Journal of Design.

I’m trying to find out how designers and design teams (in-house or consultancies) who’ve worked on influencing user behaviour think about what they’ve done – which techniques and patterns do people recognise that they’ve used, or considered? Do the patterns I’ve identified in the toolkit actually make sense to people who’ve put this stuff into practice strategically? Or do people think about it differently?

So, if you’ve worked on persuasive technology, behaviour change design, or influencing user behaviour in general, across any field where you consider that you’re designing stuff (service design, product design, interaction design, social design, user experience, information architecture, HCI, social marketing, mobile interaction, web design, network engineering, pervasive/ubiquitous computing, transformation design, advertising, urban planning, human factors, ergonomics, built environments, healthcare, environmental, safety, crime prevention – anything, in fact), I’d really appreciate it if you could spare a few minutes to have a go at this survey. It shouldn’t take too long unless you have a lot to tell me about!

‘Designers thinking about the effect they can have on behaviour‘ is a growing theme. The idea with this survey is that if we can collect together some good examples of where and how companies are using these ideas, what’s worked and what hasn’t (and why) (where you’re prepared to talk about it!), it’ll be a useful reference for everyone, as well as (potentially) a series of great case studies to be included in a book (at some point once my PhD’s out of the way). In the meantime, I’ll of course try to feature some of the projects on the blog.

If you take part in the survey, your details will go into a draw to win a classic book on design and behaviour (I’ll do one draw for every 20 participants). I’m not sure what the books will be yet, but there’s a lot to choose from. The survey doesn’t really have a closing date at present – I’ll leave it open as long as it’s getting interest.

Thanks for your help!

Then on Thursday 28th, I’m honoured to be talking as part of a symposium in Loughborough University’s Radar Arts Programme‘s ‘Architectures of Control‘ themed events exploring how our lives are impacted by social and environmental controls.

The symposium is interspersed with the performance of Mark Titchner’s ‘Debating Society and Run’, which sounds intriguing. In the symposium I’ll be talking alongside Professor David Canter, who seems to have had an incredible career ranging from environmental to offender profiling (inspiration for Cracker, etc) and Alexa Hepburn, senior lecturer in Social Psychology at Loughborough. Again, I hope my presentation does justice to the event and other participants! Thanks to Nick Slater for inviting me.

The week after (4th June) I’ll be giving a presentation at UFI in Sheffield, best known for its Learndirect courses. I’m hoping to be able to run a bit of a very rapid idea-generation workshop as part of this talk, something of an ultra-quick trial of the DwI toolkit…

Thanks too to all the people who’ve blogged about it, included it in a podcast, and spread it via Twitter. Your attention’s much appreciated and if anyone does try it out on some problems, please do let me know how you get on, what would improve it, and so on. And more examples for each of the patterns are, of course, always welcome!

Printed copies (A2 poster, 135gsm silk finish) are available – the nominal listing on Amazon is £15 including postage, but if you’d like one for much less than that, let me know! (In fact, if you’re willing to try it out on a design problem, fill in a survey about how you did it, and let me use it as a brief case study, you can have it free.)

Persuasive 2009

I say I’m just coming up for air briefly, as for the last couple of weeks, among some other major work (which could possibly bear some very nice fruit), I’ve been putting together my presentation* for Persuasive 2009, the Fourth International Conference on Persuasive Technology in Claremont, California, next week, and at present am desperately trying to finish a lot of other things before flying out on Saturday. It’ll be my first time across the Atlantic and my girlfriend and I will be having a bit of a holiday afterwards, so I hope a lack of updates and replies, while little different to my usual pattern, will be excusable. But while the conference is on, if there’s time and no hoo-hah with the wireless and it seems appropriate, I’ll try and do a bit of blogging, or more likely, Twittering about it (#persuasive2009 ?). There are some very interesting people presenting their work.

Anyway, if you missed the update to my earlier post, a preprint version of my paper (with David Harrison, Tim Holley and Neville A. Stanton), Influencing Interaction: Development of the Design with Intent Method [PDF, 1.6MB] is available. At some point soon this version of the paper will downloadable from Brunel’s research archive, while the ‘proper’ version will be available in the ACM Digital Library. ACM requires me to state the following alongside the link to the preprint:

© ACM, 2009. This is the authors’ version of the work. It is posted here by permission of ACM for your personal use. Not for redistribution. The definitive version will be published in Proceedings of Persuasive 2009: Fourth International Conference on Persuasive Technology, Claremont, CA, 26-29 April 2009, ACM Digital Library. ISBN 978-1-60558-376-1.

The presentation will include many parts of the paper, but the nature of academic papers like this (submitted in December) is that they are out of date before anyone reads them. So, much of the presentation will be about the DwI toolkit and the reasoning behind bits of it, rather than just sticking to the state of the research six months ago – I hope that’s reasonable. Last year, presenting on the last day of the conference meant that I was able to spend many hours in a hotel room in Oulu editing and re-editing the presentation (mostly listening to the Incredible Bongo Band’s version of In-a-Gadda-da-Vida on repeat) to match what I thought the audience would like, and incorporate things I’d learned during the conference, but this time I’m on the first day so there isn’t that opportunity…

Interfaces article

Also this month, I have a brief article about my research in Interfaces, the magazine of Interaction, the British Computer Society’s HCI Group, in its ‘My PhD’ series (p. 20-21). Interfaces no. 78 is available to download here (make sure to click on the link below the cover image, as – at time of writing – the cover’s linked to the previous issue). It’s a great magazine – redesigned for this issue – with some really interesting features about aspects of HCI by some well-known names in the field. Thanks to Eduardo Calvillo and Stephen Hassard for making the article possible.

The table in the article was unfortunately truncated during editing so (if I get it in in time) there’ll be a brief addendum in the next issue with the full table, but I might as well make it available here too [PDF, 8kb] – it’s a brief, not especially exciting summary of some concepts for influencing householders to close curtains at night to save energy. (At some point I’ll do a full case study on this as there are some interesting ideas as well as some very impractical ones.)

*Taking Parkinson’s Law as an instruction manual seems to be a perpetual habit of mine, so the maximum time allocated to get the presentation done has been more than entirely taken up by getting the presentation done… it’s still not quite there, and I’m not sure whether the format of the auditorium’s going to allow an interactive element which I would very much like to include but probably won’t be able to. Also – while Prezi looks like it might be everything I’ve ever wanted in presentation software – the workflow of “doing a PowerPoint” for me has evolved into a long chain of “Photoshop – Illustrator – export – Photoshop – Save for Web – insert into PowerPoint” which I’m sure I could do more quickly, but lots of conferences and seminars want PPTs rather than PDFs, and the only Mac I have (which once – kind of – belonged to the Duke of Edinburgh [interesting story]) is too slow and old to run anything better.

The brief series of posts from last summer about getting people to do things in a particular order, which more recently got some attention from Kati London’s ‘Persuasive Technologies: Designing the Human‘ class at NYU’s Interactive Telecommunications Program (with some very interesting student commentary) was based on a relatively early iteration of the method. At some point, I’ll draw up a comparison between the iterations of the method, even if simply for my own clarity of mind – it’s helpful to record why I changed different aspects along the way.

The initial plan had been for it to be almost TRIZ-like in terms of ‘prescribing’ relevant design techniques to help achieve particular target behaviours. The first few iterations of the method thus took the form of a kind of hierarchical decision tree. Live|Work‘s very helpful advice to me last summer to reduce the prescriptive nature slightly by having a kind of illustrated ‘idea space’ led – in due course – to the version tested in the pilot studies carried out in late 2008 and earlier this year. What the studies showed, among other things (to be reported in the Persuasive 2009 paper!) was that many designers, when asked to come up with concept solutions, don’t really like working from categories and rules and hierarchies, even where they would be useful. Some do (and with impressively exhaustive efficiency), but many don’t: they preferred to use the method as a kind of well of inspiration, without necessarily using it in any kind of procedural way.

So – and there’s another reason for this, too, which I’ll be able to announce at some point – it seemed sensible to redesign the method to accommodate both modes of working: a ‘prescription mode’ for the more procedure-driven designer, and an ‘inspiration mode’ for the designer who prefers less bounded creativity (a bit more like IDEO’s method cards, but not quite as unstructured as the Oblique Strategies). The inspiration mode is essentially a very simplified, flattened set of design patterns loosely grouped into different ‘lenses’ representing views on influencing behaviour, but with no real structure beyond that. It’s more of a ‘toolkit’ than a method, and because of its relative simplicity it seems worth releasing to get some feedback without too much more work. The “eight design patterns for errorproofing” post from a few weeks back is a kind of preview of part of it.

On Monday morning, then, there’ll be a large poster available to download on the blog, and I’ll do a series of posts forming the online component of the toolkit. So please, feel free to comment, make suggestions for improvements or better examples, or pick holes in it!

P.S. I’m aware the blog needs a bit of housekeeping in terms of making the sidebar work properly again in IE, fixing the very out-of-date blogroll, and my appalling sloth in replying to people who’ve very kindly sent very interesting links and ideas. I will try to get round to it all soon.

© ACM, 2009. This is the authors’ version of the work. It is posted here by permission of ACM for your personal use. Not for redistribution. The definitive version will be published in Proceedings of Persuasive 2009: Fourth International Conference on Persuasive Technology, Claremont, CA, 26-29 April 2009, ACM Digital Library. ISBN 978-1-60558-376-1.

——————

I’m pleased to announce that a paper I submitted to Persuasive 2009, at the Claremont Colleges, California (26-29th April) has been accepted, so I’ll be presenting ‘Influencing Interaction: Development of the Design with Intent Method’ on Monday 27th April.

The paper builds on the ideas I presented at Persuasive 2008 (the paper), detailing the development of the ‘Design with Intent Method‘, a ‘suggestion tool’ for designers faced with briefs involving influencing user behaviour, and the results of a series of pilot studies to test the usability of the method.

At the time of submitting the paper (New Year’s Eve, 6pm!), the pilot studies were still going on (poor planning by me), so (as the reviewers noted!) the paper’s conclusions are fairly weak, and there are quite a few revisions I need to make before submitting the final version: the next couple of weeks are going to require some fairly intense work in that vein. But it’s great to have been accepted: Persuasive 2008 was fantastic, incredibly useful in terms of meeting people and getting feedback on the proposed research, and I’m hoping 2009 will be just as good. The big-name speakers include BJ Fogg, originator of the Persuasive Technology field, Mihály Csíkszentmihályi (of ‘Flow‘ fame), and Brenda Laurel (author of Design Research: Methods and Perspectives, which I’ll admit I haven’t yet got round to reading, largely because of Nigel Cross’s review, but maybe I should find the time!). As always, though, it’s the chance to talk to and get to know other people working on similar problems, or offering a different point of view on the field, which is especially interesting.

The proceedings are going to be published by the ACM (last year’s were published by Springer), but I don’t have any more details at this stage. I’ll post a preprint version of the paper here once it’s ready, of course.

Many thanks to my co-authors: my supervisors Professor David Harrison (Brunel) and Professor Neville Stanton (Southampton) for their help, and Tim Holley whose insights into improving and using the method were extremely useful. Thanks too to all the other pilot study participants, and also to the Royal Academy of Engineering, who very kindly awarded an international travel grant to help me attend the conference. I am aware of the hypocrisy of flying halfway round the world to talk (in part) about influencing more environmentally friendly behaviour, and the cognitive dissonance is headache-inducing. Why there aren’t more live, online academic conferences, I don’t know.

Here are the abstract and ACM meta-stuff for the paper:

Influencing Interaction: Development of the Design with Intent Method
Dan Lockton¹, David Harrison¹, Tim Holley², Neville A. Stanton³
¹Cleaner Electronics Research Group, Brunel Design, Brunel University, Uxbridge, Middlesex UB8 3PH, United Kingdom
²Product Design Programme, Brunel Design, Brunel University, Uxbridge, Middlesex UB8 3PH, United Kingdom
³School of Civil Engineering & the Environment, University of Southampton, Southampton, Hampshire SO17 1BJ, United Kingdom

ABSTRACT
Persuasive Technology has the potential to influence user behavior for social benefit, e.g. to reduce environmental impact, but designers are lacking guidance choosing among design techniques for influencing interaction. The Design with Intent Method, a ‘suggestion tool’ addressing this problem, is described in this paper, and applied to the briefs of reducing unnecessary household lighting use, and improving the efficiency of printing, primarily to evaluate the method’s usability. The trial demonstrates that the DwI Method is quick to apply and leads to a range of relevant design concepts. With development, the DwI Method could be a useful tool for designers working on influencing user behavior.

Categories and Subject Descriptors
H.1.2 [Models and Principles]: User/Machine Systems – human factors, software psychology. H.5.2 [Information Interfaces and Presentation (e.g. HCI)]: User Interfaces – theory and methods, user-centered design.
General Terms
Design, Human Factors.
Keywords
Persuasive technology, behavior change, sustainability, energy, interaction design, design methods, innovation methods.

On other matters, I’m proud to say that Planetizen, the urban design and planning community and blog has named Design with Intent one of its Top 10 Websites for 2009 – a nice accolade given how broad the scope here is beyond urbanism and architecture! Some of the other websites recommended are well worth a deeper read – On the Commons, Digital Urban, Infranet Lab and Gapminder stood out for me.

Adding that Planetizen accolade to Six Revisions’ inclusion of the blog in its ’20 websites to help you master user interface design’, it’s clear that, if nothing else, the themes we cover here really do meander about over conventional disciplinary boundaries. It’s all about people interacting with designed systems, whether they’re concrete plazas, electric kettles or confirmation dialogues, and I’d like to think the similarities are worth investigating.

Photo of Claremont Graduate University by Katherine H on Flickr

I’m pleased to announce the launch of Design and Behaviour, a new discussion list / Google group:

The design of products, services and environments can be used to influence behaviour, and there’s a growing appreciation of the possibilities for social benefit, especially in environmentally sensitive design, health, safety, security and crime reduction. This group aims to bring together people interested in this emerging field: interaction designers, product designers, graphic designers, engineers, architects, ergonomists, computer scientists, sociologists, psychologists, economists, philosophers, researchers, strategists, policy-makers and anyone else with something to say, or an interest in learning what others are doing.

Run, initially at least, by myself with help from Debra Lilley, the group’s intentionally got a pretty broad scope. Please, if you enjoy this blog (or even if you don’t enjoy it but are interested in the field!) sign up (there’s also a Facebook group if that’s your thing). How the group develops is up to the members, so I can’t give you a definitive high/low traffic indication. But we will endeavour to keep it usable.

P.S. Both designandbehaviour.com and designandbehavior.com go to the same place.

P.P.S. My apologies for the few weeks off the blog’s had. I’ve been very busy. Thanks to everyone who’s sent interesting items in the meantime – I hope to get round to posting them as soon as I can. It’s intriguing though, looking at the statistics that (aside from one-off spikes such as when we’re Boing Boing’d) the number of unique daily visitors to the site itself (i.e. not via RSS) remains fairly constant Monday-Friday regardless of how stale the posts on the front page are.

Certainly when I was deciding whether a not a PhD was the ‘right’ thing to do after a couple of years of pretty diverse peripatetic freelancing, the efforts of other bloggers – especially this article by Tom Coates (and the appended comments) – and Rich Watts’ blog, were very helpful and gave me some great, and sometimes sobering, insights. More recently, these posts by the polymathic Nicolas Nova and Julian Bleecker have given well-justified discourse on moving on from academia, even more pertinent because of their design/art-technology emphasis. (The ‘disciplinarity boundaries’ issue, which vexes me so much, has been addressed in this context by Julian more than once; Roberto Greco has a comprehensive review of more thinking on this issue, too).

Anyway, here’s (mildly edited to remove some commercial and personal information) the report I prepared, rather hurriedly, on what’s been accomplished in the first year, and what’s still to come:

Dan Lockton, Cleaner Electronics Research Group
Start date: September 2007
Design for Sustainable Behaviour
Review, end of Year 1, August 2008

Summary: Design can be used to influence users’ behaviour.

By applying techniques from a variety of fields, it’s possible to design systems which help users to reduce the environmental impact of using them: effectively, making users more efficient by designing for behaviour change.

This project aims to develop and test a method for assisting designers to create behaviour-changing products and services in this area, and then run user trials with prototypes, to determine which approaches are actually most effective at changing users’ behaviour, and reducing energy or other resource use.

Background

As part of my Master’s degree, I researched the concept of architectures of control, ways in which systems have been designed to influence users to interact with them in certain ways, often coercively, to match political or corporate agendas [1]. Subsequently, alongside working as a freelance designer/engineer/researcher, I continued to develop this research independently, primarily via a blog [2] which has gained a fairly diverse audience across the design, technology, media and social science fields.

The scope gradually broadened, becoming more positive in the process, to encompass what I’ve since termed design with intent – strategic design intended to influence user behaviour, including helping users achieve their own goals as well as those of society. This last point is important, since many social problems – particularly environmental ones – can be seen as a result of user behaviour.

It was with this background that I discussed the possibility of a PhD investigating ‘Reducing the environmental impact of products by using design to change user behaviour’ (or, more succinctly, design for sustainable behaviour) with David Harrison, and was pleased to return to Brunel Design as part of the Cleaner Electronics Research Group, with funding from the Ormsby Trust, in September 2007.

Research phase 1a: Literature and practice review

The first phase of the PhD involved investigating, comparing and characterising ‘design with intent’ techniques via examples from a wide range of fields, including human-computer interaction, manufacturing engineering and urban planning as well as product design.

Many practitioners and theorists have touched on aspects of this area from different directions without describing its full extent, and indeed, to understand this, I’ve had to acquire at least some working knowledge of concepts from a wide range of disciplines, including architecture, ecological and social psychology, behavioural economics, human-computer interaction, communication studies, science & technology studies, rhetoric, information architecture, semiotics, security engineering and quality management, alongside a deeper education in the principles of interaction design and ergonomics, to which I’d only tangentially been exposed as an undergraduate design student.

The output of this phase of the research was the paper ‘Design with Intent: Persuasive Technology in a wider context’ [3] (see below).

Research phase 1b: Initial development of the Design with Intent method

The intention of the review of techniques is to enable the development of a ‘suggestion engine’ –the Design with Intent method – for designers working in sustainable and environmentally sensitive design, integrating ideas from different fields to assist the selection and application of design techniques which influence user behaviour. The method itself can be applied to many social problems in which the design of systems (products, services, environments) affects user behaviour, but the focus of the testing will be (at least for this PhD!) on applying it to issues where user behaviour, particularly with energy-using products, affects the environment significantly.

The reasoning behind this, and discussion of its applicability to environmental problems, resulted in the paper ‘Making the user more efficient: Design for Sustainable Behaviour’ [4] (see below).

The approach taken is that certain target behaviours can be identified, and described in the abstract, with different design techniques being applicable to each one. A user behaviour ‘problem’ described in terms of one or more of the target behaviours will, using the suggestion engine, result in the designer being presented with a number of relevant techniques, with examples of each technique being applied in different contexts.

The initial development produced a rather TRIZ-like method, using a tree structure to match target behaviours to relevant design techniques, and my own paper-based run-throughs indicate that it seems to work in terms of generating new design ideas. This is described briefly in the poster I presented at Brunel’s ReSCon [5].

Research phase 2: Testing and refinement of the Design with Intent method (current)

The aim of testing the method is to determine: a) to what extent it is useful to designers addressing user behaviour problems in sustainable design; and b) how the method can be improved. In terms of a), the comparison is with an unstructured brainstorming-type method: does the Design with Intent method offer anything beyond this? Would it perhaps be better implemented as a reference book, a ‘design for sustainable behaviour manual’, rather than a ‘suggestion engine’?

As a precursor to practical testing, in July 2008 I explained and ran through the tree-structure method with two directors and the studio manager of Live|Work [6], a major service design consultancy in London specialising in socially beneficial design solutions for both public- and private-sector clients. The feedback – from exactly the kind of designers I envisage being the ultimate users of the method – was extremely useful, and resulted in a significant redesign of the way the method is presented, moving from a tree structure to a series of concentric rings which allow easier creative exploration of ‘related’ design techniques and target behaviours. This redesigned method, along with some revised (simplified) terminology, is what will be tested.

The testing programme is intended to involve both design students and design consultancies: this is the best way of assessing its usefulness both to existing designers in the context of commercial constraints, and the next generation of designers in an academic setting. The method will be refined as a result of the testing.

First, a pilot study is being arranged with individual design students/recent graduates, using a think-aloud protocol, with all guidance and assistance recorded, primarily to identify points that need clarification or potential problems that may arise. The plan for this study is being written at present (August 2008) and, subject to approval, should be reasonably quick to undertake.

The full study will take the form of workshop sessions in the Autumn term, probably with Level 3 Design students. Participants will be introduced to the method, and, in separate groups, assigned ‘sustainable user behaviour’ problems, with the method there to guide them in generating solutions. (The control will not have the method.) The group interactions and creative process will be recorded and assessed, as will all the output; the specifics of this study have not yet been decided.

A possibility has also arisen to apply the method to one of a consultancy’s client projects, in due course, which has significant potential for testing the method’s worth under more market-based constraints, in a real design consultancy. Other consultancies will also be approached.

Research phase 3: Application of the method

The usefulness of the method will best be tested by the quality of the concepts it generates, so the aim of this phase of the research is to build (prototype) and run user trials comparing products developed by applying the method to a particular problem (users overfilling kettles is a favourite, but there are many possibilities).

This will allow quantitative assessment of the actual energy used by different products, by representative users, in use, over a period, to provide some information about the effectiveness of different techniques in that context, as well as qualitative feedback on usability and other issues. This information can then be used to refine the method further, so that, for example, details of the relative effectiveness of different techniques can be incorporated.

Contributions to knowledge

The project will address these questions, reformulated as appropriate:

• How can users’ behaviour be changed, through redesign of systems, to reduce environmental impact?
• How significant are the impact reductions, and what technology and human factors issues affect the implementations?

It’s hoped that the process of investigating and answering these questions, will constitute an original, distinct and useful contribution to knowledge, and that the Design with Intent method — however it evolves — will prove useful to designers working in the field of behaviour change in society in general. Since the ‘suggestion engine’ of the method is effectively an ‘innovation engine’, it is envisaged that worthwhile intellectual property may also result.

Research output and academic development

Two papers (one journal article, one published conference paper) have so far resulted from the research, and thanks primarily to visitors from the blog, have achieved significant visibility on BURA (top paper and 3rd highest number of views in June, and still currently the highest average views per author):

Lockton, D., Harrison, D.J., Stanton, N.A. ‘Making the user more efficient: Design for sustainable behaviour’. International Journal of Sustainable Engineering Vol.1 No. 1, pp. 3-8, March 2008 [3]

Lockton, D., Harrison, D.J., Stanton, N.A. ‘Design With Intent: Persuasive Technology in a Wider Context’. in H. Oinas-Kukkonen et al. (Eds.): Persuasive 2008, Lecture Notes in Computer Science 5033, pp. 274-278, Berlin: Springer-Verlag, 2008 [4]

As a result of the IJSE paper, I was asked to become a reviewer for the journal and have so far reviewed one submission.

I have presented at two external conferences, Persuasive 2008: The Third International Conference on Persuasive Technology, in Oulu, Finland, in June, the costs of which were partially funded by receiving a Vice-Chancellor’s Travel Prize (presentation: ‘Design with Intent: Persuasive Technology in a wider context’ [7]) and New Sciences of Protection: Designing Safe Living at the Institute for Advanced Studies, Lancaster University, in July (presentation: ‘Design with Intent: behaviour-shaping through design’ [8]). I also presented a poster [5] at Brunel’s ReSCon (and the Graduate School poster competition) which provided a good opportunity to try explaining the research to more engineering-focused visitors, and has (I hope) helped me understand how to improve the clarity needed to present research in poster form.

The invitation to present at Lancaster came as one of the organisers has been following the research via my blog; it’s hoped that this kind of visibility can help even further as the research progresses. At present I have an invitation to present at Design|Behaviour: Making it Happen at Loughborough in October.

I also attended a doctoral consortium organised by the Universities of Oulu (Finland) and Aalborg (Denmark) prior to the Persuasive conference, and the networking and discussion with other researchers working in similar areas of design, computer science and psychology were extremely useful and have dramatically expanded and sharpened the focus of my thinking. I now have contacts at a number of institutions and companies internationally who are interested in the research and some of whom may be, in time, potential collaborators. During the year I’ve tried hard to attend and participate in as many relevant events as possible, both to meet other researchers involved in related fields, and to learn more about how academia and practising designers work together – a partial list:

• Anthrodesign & UX Meetup, London, Sept 2007
• BSI Manufacture, Assembly, Disassembly, and End-of-life Processing standards meeting, Sept 2007
• EPSRC Network on Product Life Spans seminar, Sheffield Hallam, Sept 2007, with Alex Plant
• Sustainable Design Network seminar ‘Envisioning a Sustainable Future’, Nottingham, Dec 2007
• Meeting at University of Bath with Dr Elies Dekoninck and Ed Elias to discuss similar research areas, June 2008
• Attended meeting with Staffan Davidsson (Volvo Cars), Dr Mark Young and Stewart Birrell, June 2008
• Interviewed by Jamie Young (Imperial College) for behavioural change policy research, July 2008
• OpenTech open innovation & technology conference, London, July 2008
• The Affective in Sustainable Design, seminar, Central St Martins, July 2008
• RSA lecture by Richard Thaler, author of Nudge, London, July 2008

At Brunel, I also gave a seminar in June 2008 as preparation for presenting in Finland, and received some very useful feedback.

In terms of parallel activities at Brunel, as well as the Graduate School and SED induction training modules, I’ve completed the Graduate Training Assistant training, and the Graduate School’s Entrepreneurship Masterclass, and have helped assess Level 3 Environmentally Sensitive Design group projects. During the Spring term I assisted with the weekly Level 2 Electronics labs and also marked some of the final assignments, which has given me a good insight into how all this works. I’d welcome the opportunity to be involved further with Design teaching in the next couple of years.

I am excited and enthusiastic about the years ahead, and the opportunities they present, and would like to thank everyone who’s helped me so far.

[1] Available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=908493
[2] Architectures of Control / Design with Intent blog: http://danlockton.co.uk
[3] Available at http://hdl.handle.net/2438/2138
[4] Available at http://hdl.handle.net/2438/2137
[5] Available at http://danlockton.co.uk/research/poster_DL.ai.pdf
[6] Live|Work website: http://www.livework.co.uk
[7] Available at http://architectures.danlockton.co.uk/2008/06/09/design-with-intent-presentation-slide/
[8] Not yet available online

I passed the review OK, but it was made clear that I really ought to have a more formal, critical literature review, at least in draft, done by now, pertinent to the actual intended contributions to knowledge, and explaining the ‘hole’ in current knowledge and previous work that I’m aiming to fill. Of course, I’ve done plenty of reviewing what’s out there, but given the amount of new avenues and relevant theories I seem to come across weekly, it’s been difficult to draw it all together coherently, and I’ll admit I’ve been putting it off. Perhaps now it’s time to do it properly, along with a ‘contents page’ for the thesis, alongside organising the pilot studies of the DwI method (more on which on the blog in the near future). Yes, deciding what to leave out is going to be hard, but that’s part of the point.

Thanks again to everyone who’s helped this year: having the collective experience of hundreds of intelligent blog readers from many disciplines to draw on and inspire the research has really made the whole thing so much more dynamic, somehow.

As longer-term readers may remember, back in 2005, I initially approached this whole subject from a very anti-control point of view. Products, systems, environments which seek to control or coerce the user into particular behaviours are, ultimately, not desirable, and the user mentality which seeks to avoid that control, or circumvent it, is something to be applauded, and very much necessary for the advance of society. Designing ‘for’ emergence is difficult; but emergent behaviour – at the very least, weak emergence – will emerge anyway.

Now, I still very much hold the anti-control belief – the proliferation of fundamentally anti-user artefacts in public space (as we see every couple of weeks on the blog) still astonishes me – but the major insight that led to taking on the PhD and the direction since the end of 2007, has been that, applied in a different context, some of the same techniques can actually help the user, improving efficiency and helping society at the same time. All design is persuasive, perhaps: any design technique can be used for ‘good’ or ‘evil’, just like any other tool. Guiding/persuasion/coercion may all be part of a continuum, depending on your point of view, but if something is (transparently) helping you to achieve something which benefits you, you’re less likely to try and find a way round it.

Rest assured, then, I will attempt to include recognition of this in the method that’s being developed, even if it’s a simple step that asks the designer to consider that the particular technique under consideration “has been shown to provoke user resistance/hostility/reaction” and hence maybe isn’t ideal. We’re all users, even designers: even supercilious councillors were kids once, and we mustn’t lose sight of that.

Heading north-east
I’m going to be away for a few days in Finland presenting at Persuasive 2008 (in fact, should be there already as this post appears on the blog), so the posts might get a bit slower/briefer for a while: apologies if I haven’t/don’t reply to your emails or comments (yet), but I hope to do so as soon as I can. It’d be great to do a bit of Jan Chipchase-style blogging while I’m experiencing these liminal spaces of departure lounges and interstitial connexions: I’ll see what I can do.

Seminar Announcement:
Using design to shape user behaviour:
Design with Intent and Persuasive Technology

Dan Lockton, Cleaner Electronics Research Group
27th May 2008, 11.00a.m. (Approx. length 45 mins)
Room TA049 – Tower A, Brunel University, Uxbridge, Middlesex UB8 3PH

Everyone welcome.

Design can be used to persuade, guide and shape users’ behaviour. Anything designed for user interaction can be designed to embody intended ‘target’ behaviours, whether for socially beneficial or purely commercial reasons.

The emerging interaction design field of Persuasive Technology incorporates some of these ideas, mainly applying them to software and motivational games which guide users to change their behaviour for socially beneficial reasons, e.g. keeping fit (Brunel’s Gillian Swan’s ‘Square-Eyes‘ project has been cited in some of the literature) or giving up smoking.

Taking a much broader look across product design, engineering, architecture and computer science, a range of approaches to designing for intended behaviour emerges, and a more general concept of ‘Design with Intent’ can be identified, with techniques from one discipline being applicable in others. My research specifically involves guiding more environmentally friendly product use – Design for Sustainable Behaviour – but the development of a general model for Design with Intent, matching target behaviours with suggested design solutions, is an important part of this.

The seminar comprises two short presentations I’ll be giving at Persuasive 2008, in Oulu, Finland, in the first week of June. It’s a practice run, if you like, but it will be extremely useful to get feedback and reaction from a Brunel audience, and I hope it’ll be interesting and inspirational.

1. Design with Intent: Persuasive Technology in a Wider Context
This is a paper to be presented at the conference, and appearing in H. Oinas-Kukkonen et al. (eds.): Persuasive 2008, LNCS 5033. Berlin Heidelberg: Springer-Verlag, 2008. pp. 274 – 278

2. Kairos: Just-in-Time Feedback as a Design with Intent technique
This is an invited presentation to be given at the doctoral consortium which forms part of the conference.

Dan Lockton (Year 1 PhD student)
Daniel.Lockton@brunel.ac.uk | http://architectures.danlockton.co.uk (research blog)

Supervisors: Professor David Harrison, Professor Neville Stanton

Thanks to Lyn Edgecock and everyone who’s helped set this up.

I’m pleased to say that I’ll be presenting a short paper, Design With Intent: Persuasive Technology in a Wider Context* at Persuasive 2008, the 3rd International Conference on Persuasive Technology, taking place from June 4th-6th in Oulu, Finland.

The paper’s a (very) brief introductory review of some of the different approaches to ‘Design with Intent‘ from various disciplines, many of which have been discussed to some extent on this website, with an attempt to relate them to persuasive technology, the field started by Stanford’s B J Fogg and his team and now rapidly developing worldwide at the intersection of interaction design and behaviour change. (The paper doesn’t get as far as the DwI Method on which I’m currently working and hoping to test in the next few months.)

This is my first stab at a conference paper, and I’m incredibly excited (and lucky) to have had it accepted; there are a lot of very helpful comments and suggested revisions from the reviewers which I will endeavour to incorporate. I’m not sure what the conference organisers’ position is on making the paper available here; certainly authors from previous Persuasive conferences have put papers on their own websites after the conference, so I expect I will do the same. The proceedings will be available as part of Springer’s Lecture Notes in Computer Science series.

Many thanks to everyone who’s helped with my research via this site, suggesting angles to investigate and helping to clarify my thinking in this area, and to my PhD supervisors at Brunel, Professors David Harrison and Neville Stanton, for their help and support.

*Lockton, D., Harrison, D.J., Stanton, N.A. ‘Design With Intent: Persuasive Technology in a Wider Context’.

Abstract: Persuasive technology can be considered part of a wider field of ‘Design with Intent’ (DwI) – design intended to result in certain user behaviour. This paper gives a very brief review of approaches to DwI from different disciplines, and looks at how persuasive technology sits within this space.

UPDATE (21 April): Following the precedent of some other Persuasive authors, I’ve uploaded a preprint version of the paper here: Design With Intent: Persuasive Technology in a Wider Context [PDF, 169kb]. As required to be stated, this is a self-archived preprint version of the paper, to be presented at Persuasive 2008, June 4-6, Oulu, Finland, and published in H. Oinas-Kukkonen et al. (Eds.): PERSUASIVE 2008, LNCS 5033, pp. 274 – 278, 2008.
© Springer-Verlag Berlin Heidelberg 2008