Activity-Centered Design

I couldn’t agree more. I’ve been using the “Customer Activity Cycle” by Sandra Vandermerwe for years now, and it always delivers.

With regards to your points about framing, I was wondering if you feel it is possible/sensible to ‘re-frame’ during an interaction; somehow taking the user along. I made an analogy to Japanese doors to clarify this point.

My wish is to create an experience during the interaction, without losing track of the main activity of course.

An academic POV: the body of work on Activity Theory is terribly helpful in this line of thinking. Not only does it keep the orientation of design work on the right thing (user’s motives & goals rather than systems’ features), but there is a host of other useful apparatus for designers too.

Just one tiny example: Activity Theorists understand the term “Activity” in a specific way…as the thing that folks are motivated to do. A smart guy named Leontev proposed a hierarchical relationship wherein “Actions” and “Operations” fall underneath “Activity.” Actions are goal-oriented, but not necessarily motivated. Options are neither, they are just the situtionally-constrained ways that folks have to actually care out an activity.

So…it might cause some revisions to Josh’s list. I think, for example, that many people do use Flickr to share photos. That is, they are motivated to do that. I use Flickr for something else: to coordinate experience with people I care about, but who live far away. I still share photos with them, but I don’t take high-quality photos or necessarily care about those shots…expect as they help me to share with my friends and family what we are doing, feeling, seeing, etc.

One thing that explains Flickr’s success is the way it supports “sharing photos” so well at both the Activity and Action (goal-oriented) levels.

ugh…typos galore…sorry; no way to edit now

What I like most about activity-centered design is that the focus on usability starts from Day 1, not right before launch.

I just posted an article on my blog about Prototype Driven Development which is actually a perfect complement to activity-centered design. I think I’ll have to write a follow-up post on how the two could work together. If, that is, Prototype Driven Development actually can work.

David, I’ve been doing prototype-driven development with the companies I work with for a long time, without calling it that. I can’t wait to read your article on it. I also want to check out that “Customer Activity Cycle” - this is a great topic. I think I do something more visual and customer/marketing-centric than people in the role of “information architect” do, but that doesn’t mean I don’t do detailed design. I think with all the tools available, there are definitely different ways of reaching the same goal, and there is NOT one, single model that is THE only model that works. Thanks for a great article Joshua… I will be exploring the activity-based concept more. Great mental food to chew on from all of you.

Hi Josh,
You might want to search the archives of the IxDA list for ACD vs. UCD vs. design or any combination there of. There was even a basecamp created for the purpose of defining ACD separate from UCD, blah blah blah.

But the way you are framing ACD is really similar to and corroborates what Alan Cooper has been professing ever since “Inmates …” when he developed and started teaching his methods of “Goal Directed Design”. I would also think that “Contextual Design” could also be thought of as ACD. And both of these are WAY beyond academic at this point.

For me the issue feel moot. Why? B/c whether you focus on the noun (user) or the verb (actions), it is still really the same approach and framing, IMO. Verbs w/o actors are quite useless. I.e. a “shopping” site for diamonds is going to be way different than a “shopping” site for vintage clothing. The PEOPLE and the MERCHANDISE both change the very nature from which you will design from. So say that Amazon supports “shopping” is true, but it supports a specific type of “Walmart-esque” type shopping.

– dave

I think I’m with David here. “Goal directed design” still makes the most sense to me. All activity happens because people try to accomplish some goal, even if it’s as loose as “relieving boredom” or “being entertained”. Successful design for activity still requires understanding your users, their goals and their context.

@Josh: Thanks so much for continuing the conversation on ACD. I’m so glad I’m not the only one advocating it.

@zephyr said: “Successful design for activity still requires understanding your users, their goals and their context.”

ACD, and indeed any approach to design, certainly requires a very strong understanding of users—how they think, how they work, etc. The difference with ACD is that the focus is on how they perform activities rather than on their goals.

By focusing on activities, a user’s larger goals become irrelevant. The only goals you need to consider are the ones involved with the activity itself. It doesn’t matter, for example, that Jenny the single mom wants to be a nurse in 3 years—what matters is how she performs the particular activity you’re trying to help her perform, and how you can support the much smaller, much more specific and contextual goals that are related directly to your application.

Basically, you can’t really design an app that helps Jenny become a nurse, but you can certainly design one that enables her to take an online course related to her pursuit to become a nurse. And as such, you can also design an app that enables all kinds of other people to take online courses to become all kinds of other things.

hey Joshua -

great insights. i would agree that at least for consumer-focused web design (which is most of the web i guess?), user activity is the primary focus.

however, this segments into 2 basic things:
1) what the user wants to do
2) what the business / website wants the user to do

the (hopefully complementary, not antagonistic) combination of these two objectives is what should drive design, IMHO.

over the past year or two, i’ve been working with a lot of startups to put together a simple 5-step model for how to think about user actions, and how that drives product & marketing decisions. it’s called Startup Metrics for Pirates, and the 5 steps are:
* A: Acquisition: channels where users come from
* A: Activation: the “happy” first experience; target actions taken by users
* R: Retention: how you get them to come back
* R: Referral: do they like it enough to refer others
* R: Revenue: how to make money

it might be a bit contrived (AARRR! = “Pirates”), but the idea is to come up with relatively simple, actionable ways to measure progress.

i’m also putting together a conference on these concepts called STARTonomics.com next week in San Francisco. if there’s any chance you’re around, please drop by

in any case, i’m a big fan of your book, and i think it’s great.

- dave mcclure

Great post! And I’m always glad to see judicious definite-article choice — “a” vs “the”
Frames are tools too, as you say/imply, and having a number of them available means we don’t fall into blinkered orthodoxy. Having more frames for looking at design problems is essential, especially (as Lakoff says) for new kinds of problems that we don’t have a language for yet.

Framing does provide you with the ability to get down deeper with the stuff you choose to spend your time with. I very much like the idea since it simply lets your mind get in order concerning a topic that previously could have consisted of 1000 firefly-like ideas.

David & Zephyr - You’re absolutely right. Activity Theory goes a bit deeper than Joshua described here, and says that all activities are situated within a specific social and historical context. A Thrift Shop is a perfect example of context significantly impacting how the activity of shopping actually occurs.

Activity Theory doesn’t really throw out goals so much as change the scale.

Now, I love Activity Theory as a theory, and “Activity Centered Design” as a frame for design, but in practice ACD is pretty vague. I don’t think I’ve ever found a good explanation of how exactly ACD significantly differs from UCD in practice.

I agree with David here, that UCD and ACD are different sides of the same coin. The frame you choose in each engagement affects how well you’ll be able to affect the organization and affect the product/service you’re designing.

Along those same lines, I think each frame makes different kinds of design more possible.

For example, ACD is really fantastic at creating useful products. You frame the problem around an activity, and create a useful facilitating product.

On the other hand, UCD might be better at “innovating” *new* products because it looks for gaps in what the user does, or how the user thinks, and their environment, a gap a new product might fill.

I think one could argue that when it came out, the iPod was user-centered, whereas other MP3 players were activity-centered.

This isn’t to suggest UCD is some kind of bonanza machine. Good design is the bonanza, not the frame. Basecamp is a good example of a very successful, activity-centered software.

Do you agree with the examples?

@David M. I would actually disagree with your comment about focusing on either the noun “person” or the verb “action”.

The point I’m trying to make is that the *other* noun to focus on is the activity…

@zephyr, @David M., @Jackson, @Austin I agree completely that this needs to be fleshed out more…perhaps it is not far from goal-directed design, but I’m not sure yet. I know that Cooper focuses a lot of energy on personas, and I’m not sure those get as much focus in ACD.

Another, perhaps more pragmatic concern of mine is that I feel that design often has too many deliverables and artifacts…especially UCD. Part of what I want to discover out of this is what artifacts are really needed to do design as opposed to what are needed/used to communicate value to stakeholders…

Thanks for your pushback here…that’s exactly why I published this unfinished thought…

Hi Josh,

Maybe I’m missing something here (and I always feel like I am when we start on this topic), but I’m still not clear on what problem we’re trying to solve. What is it that ACD is doing different from approaches, whatever they might be called?

Isn’t what you’re talking about here just a narrowing of the design scope to the specific activity the design is trying to support? So, the design focus for a fork would be focused on getting food into the mouth.

This works for many design projects, but doesn’t really help when the broader scope is important.

Recently, we’ve been working with the Home Shopping Network. It’s an ecommerce site, just like Best Buy or Amazon. But, it’s not.

Shoppers who come to HSN aren’t on the same shopping missions we see at Amazon or BBY. Their mission more closely match those of the HSN TV viewers — wanting to see what deals and bargains are appearing on the network. They may or may not buy, but it’s clear that for many of these shoppers, the shopping is not always the same as buying.

Maybe you could help me understand how ACD helps creating a design for HSN that would be different from Amazon or Best Buy?

Jared

@Jared, good question…it sounds like you’re saying that shopping on HSN is a different activity than on Amazon or Best Buy.

I don’t know how the activity differs, but I think activity-centered design would say those differences are going to drive the design decisions. Maybe it’s that there is more impulse shopping and less product choosing, as HSN has defined inventory for a particular time period. Or perhaps they’re supporting impulse buying from the TV and so customer reviews become either too time-consuming or impossible to gather. I don’t know what the specific differences would be (I’m just guessing here), but those difference in activity will reveal the way forward.

I remember speaking to a PM from QVC a while back, and he said that it’s all about the home page, where they update content *all the time*, like every 15 minutes or so. Folks are coming to the site from the TV, so their concern was to change the homepage to keep up with the TV deals, and it was all about impulse buying…time remaining, stock remaining, etc. All of their feature set revolved around impulse buying, as that was the primary activity they were supporting.

So the problem for QVC was to map out the activity of moving from the TV to the web, which is a quite different activity than what Amazon is supporting…and this would provide them with a different feature set as a result. Perhaps it’s similar with HSN?

Josh,

As promised via twitter, I’ve published a comment/response at http://docholdsfourth.blogspot.com/2008/09/activity-centred-design-response-to.html

I’ll happily migrate the response in whole or part here if you’d prefer.

Steve

What an awesome thread this post spawned!

I echo some of the other commenters in that ACD and Coopers Goal Direct Design as described in his About Face book have a lot of the same properties. But oh man do I hear you on the deliverable bloat Josh.

Since this is the social design blog I cant help but bring up the connection between Social Objects and ACD,UCD,GDD. Over the past year we have been creating a lot of strategies with social objects as the kernel. I have found that framing activity, features, and the experience in general around social objects allows things like feature parity and bloat to melt away. So first defining what the social object is, then envisioning what activities grow out of that object. This obviously works better with smaller focused applications.

But when you look at this in terms of Amazon Vs. HSN the nature of the objects should change the activity/design. Amazon is no nonsense shopping. The social object is the product review. HSN is shopping with a twist. The HSN social object is the bargain. What kind of activities and designs grow from bargain objects?

The frame of user-centered design isn’t just the specific people who will use the product, but also their goals, motivations, and activities. I think you’re just fragmenting the field with this article. User centred design already captures activities.

I think the difference of what you’re talking about here (ACD vs. whatever) is really just about good-quality design vs. poor-quality design. Poor-quality design ignores important factors and focuses on unimportant factors. Good-quality design isolates the important factors and prioritizes their consideration.

The reason I brought of HSN was that, in isolation, it is the same activity (as per your list above) as any other e-commerce site: shopping. But, the goals and context of the users are very different than those shopping at Amazon or Best Buy.

Those different goals will change the way they interact with the design elements. So, as @justin suggests, if you’re doing good-quality design, you’ll end up with different design results for HSN than for Amazon. If you’re doing poor-quality design, you’ll create similar designs, which, for one audience, will inevitably fail them.

I really don’t see a distinction between Activity-, User-, or Goal-, when they are done well. Or I’m just dense on this.

Josh, thanks for writing such a compelling posting. I’m with you on the notion that the frame/paradigm we choose for thinking about products is important.

I’d like to pile onto the discussion with an argument in favor of “needs” being the key concept for conceptualizing products, rather than “activities” or “goals”: Picking a frame: Activities vs. Goals vs. Needs .

@David - well I would worry about fragmenting the field except that it’s already pretty fragmented…but I hear you, we don’t need another way to talk about things just for the sake of it.

@Jared, yes, that’s a definite possibility. I’m currently reading a bunch of activity theory research, which tends to be extremely dry and difficult, and I’m trying to figure out that question…is this frame useful, or is it too close to one we already have?

I can’t help but wonder if focusing first on the activities of the user doesn’t bias the design to meet only a limited number of uses. If you start with “I want to send email” then you build an email application, you’ve already implicitly scoped your solution. However if you focus first on managing email data, then you can give your users the freedom to manipulate it in any large number of ways, they can decide how they want to make the tool work for them.

Paul.

Hi Steve, you bring up an incredibly important point. Are methodologies not as important as we think, and the people using them much moreso?

Josh,

I think it’s important to clarify that I refer not only to how good are your team members, but also what it is they’re good at. Depending on their skillsets you might find it more appropriate to choose one methodology over another. Or at least to emphasise those tasks in which your team has the greatest strength.

I wonder whether those projects used as poster-children for each of the different methodologies were the result of making a choice that best matched the project requirements & team members rather than the superiority of the methodology itself.

Steve

Hi Josh,
Thanks for the great post. Here’re some points i could bring up
1. Alan Cooper’s GDD and ACD is not conflict, you can say Goal driven and Activity Centered, makes sense
2. The original activity theory is dry and seems not bring pragmatic to the design domain ( especially after study some piece of it.
3. The activity could also be considered as Scenario or use case, they have the common kernel from my view.
4. UCD already bring many useful practice, but agree with you, the name brings to much trouble to designers ( for design thinking and practice), that’s the reason why it’s not so good

maybe something like, goal driven, scenario/activity Centered, prototyping&Testing iterated process makes some sense here.