Pat’s Point of View

Today I gave a talk entitled Prioritising user experience at Ark Group’s Information Architecture – Designing and managing information structures for improved web access and usability conference.

I tried to make the topic a bit more interesting (read controversial) and I think it went quite well.

I covered two main topics, firstly I outlined why I think user experience (UX) should be prioritised over information architecture (IA), and then I discussed prioritising UX within the organisation.

IA vs UX

They’re the same thing aren’t they? I have certainly said as much in the past. I’ve also made distinctions between them. And depending on the level you’re looking at, both of these views hold true.

At a very high level, both IA and UX are forms of user-centred design (UCD) and would be more similar to each other than either would be to, say, cooking. And of course many people consider IA and UX to be synonymous (as well as a whole host of other terms such as IxD, ID etc). But for the purposes of this discussion, let’s define them as being quite different from each other.

What is IA?

A fairly narrow definition of IA might be something like the design of structure, navigation and labelling of a website or other information system. In other words, it’s very product-centric; it’s about the user interface, the system, the thing we’re designing and making usable. No doubt many people would argue with this definition or want to offer their own definition, and that’s ok. The point I’m trying to make is that if one was to take quite a low level definition of IA—one that is focussed on the interface—then you’re leaving out quite a bit!

What is UX?

The holistic user experience is made up of many different factors, and many different components. It’s more than just the IA, it’s more than just the UI, it’s more than even the product itself. It’s all the things that might make up the experience someone has with, or around, a product, service or organisation. A good UX will satisfy your audience, and keep them coming back.

If we expand this definition just slightly, you can see why UX is starting to merge into the area of Service Design and Customer Experience. We’re talking about multiple touch points and multiple media, not just the technical elements of UX in the Jesse James Garrett sense.

The contrast with the above definition of IA should be pretty clear.

Why UX over IA?

The purpose of this discussion is not to criticise IA (as defined above) or those that do not distinguish it from UX, but rather to highlight what we should be focussing our attention on. We should be focussing on the overall experience, not the individual aspects of the design of the a product. This includes the IA, but also the visual design, technical design and usability. Yes that’s right, usability is not the most important factor. I believe usefulness, appropriateness and the overall experience are much more important than usability.

So when we talk about prioritising UX, we’re talking about making sure what we’re building is useful for the people we are building it for. Does it make some aspect of their life easier/faster/better? Does it fit in with the rest of their lives in an appropriate and beneficial way?

When I say appropriateness I mean fitting a user’s “ecosystem”. You ecosystem is the (typically large) variety of sources from which you gather information that allows you to perform some function or role (Sonnenwald uses the term information horizons to describe a similar concept). These sources are often offline and used with little loyalty, killing the myth of the “one stop shop” that users will come back to like lemmings. This is a very useful perspective to take as a sense check for usefulness and appropriateness of proposed solutions. For instance, we might be tasked with designing a website, yet does that make sense? Not just ‘can we build a website’ but how realistic is it to propose the audience goes to a website to perform the task or action? Sometimes it just doesn’t add up and you need to re-think how you can sensibly be part of their ecosystem.

Of course there are times when a new product or service—even something as intangible as a website—can have a marked impact on people’s lives; change their ecosystem, change their habits, change the way they see the world. Good examples include Google, the iPod, Tivo and mobile phones; people do do things differently now because of the impact of those things. But I tend to think of these successes as serendipitous rather than calculated. Over time we have changed in response to products/services/technologies that could probably not have been predicted, let alone planned.

So to expect to design and launch a life changing product/service/technology is probably being a tad optimistic. Yet that is so often the reason given for ignoring the audience’s ecosystem: “this will be a market killer!”. It’s not that this doesn’t happen by design, just that only a small percentage of us are in this league. And it takes a hell of a lot of hard work to do this, it’s not as easy as saying your pride and joy will be a “one stop shop”. If you build it, they won’t necessarily come.

Ok back to the point, which is that in almost all cases you need to fit into your audience’s ecosystem and provide something useful, otherwise it doesn’t matter how pretty, clever or usable your product/service/technology is.

Hence I believe that we (those that might be willing to undertake this IA thing) can easily miss the boat if we’re not careful, and forget to meet our true, if perhaps implicit, objective which is to meet the needs of real people in the real world. If you’re just making it usable you’re not going far enough. If you’re just doing IA then you’re not doing enough. Similarly, focussing just on usability will ultimately fail.

Crucially you must understand who your audience is and what they need, want and do, before you can have any hope of creating something that is useful and that fits into their ecosystem in an appropriate way. You need to undertake research to find out these things.

Testing vs research

There is “usability testing” and then there is “user research”. These two activities are often confused and before we continue, let’s clarify.

User research is generative, exploratory and formative. It’s aim is to gain insight and inform the design process. That said, it doesn’t have to happen prior to or disconnected from the design process, just that it’s uncovering information that goes into design. In fact, this kind of research can continue well into the design process, especially if an iterative approach is being taken.

Usability testing on the other hand is evaluative, validitory and summative. It’s aim is to assess the outputs of the design process. That said, it doesn’t have to happen after the design process is finished, just that you’re taking stuff you’ve created and seeing how well it works, on many different levels. In fact, testing can take place from any point when there is something produced by the design process that you want to check or validate. Again, this is especially the case if an iterative process is being taken. (Oh and usability testing is not UAT)

The main similarity between the two is that they involve, nay require, user involvement.

On a practical level, there are many methods or techniques that are the same or similar between research and testing. The way you use them, or the value you extract from them, might be different, however. For example, interviews and focus groups are commonplace research methods, but they can also be used to “test” concepts, ideas and an understanding of the audience’s needs. Likewise, the “talk aloud” methodology typical of usability testing can be used as a research tool to learn about the audience and gain insight into how they use information systems, almost disregarding feedback on the actual user interface users are shown.

But what we’re talking about here is the research side of things; generating insight that informs design, and strategy.

Understanding users

A large proportion of the insight you’re gaining when you undertake user research is an understanding of users, or the audience as I prefer to call them. For example the kinds of question we might aim to answer include:

• Who is our audience?
• What are their goals, attitudes and behaviours?
• What are their information needs?

How do you discover all of this? Well there are a variety of methods you can use, ranging from face-to-face interviews and focus groups, through to “virtual ethnography” where you monitor the digital footprints of your audience to see what they’re doing online. And everything in between. There’s plenty of information on all these methods, just Google it, or you could attend one of my user research methods workshops.

Informing IA

Once we’ve got this understanding of our users, how do we use it to inform our UX strategy? And how does this trickle down to structuring the IA? Both excellent questions.

Let’s start with the strategy. Putting it simply, strategy is deciding “what to do” at a big picture level. User research can help with this in two ways: ensuring what you’re going to do is what people want, need and will use, but also ensuring that you going to do it to the right people. Once you have that sorted you would be able to go on to the IA.

Whilst at one level we want to ensure we focus on the UX, the useful, if we’re designing a website, intranet or other kind of information system, we will need to create an IA at some point. This is where you can get more specific and look at those three basics of IA I mentioned before: structure, navigation and labelling. Your user research can inform each one of these.

The structure of the information is something you can easily gain insight into through research. You might use card sorting to uncover how your audience see the content fitting together or you might get feedback on an existing categorisation scheme.

Navigation is closely related to both structure and labelling, however, a key research finding for navigation might be the outcomes of a task analysis that tells you the process users follow and on which you may want to model your navigation.

Labelling involves naming parts of your IA in a way which will allow your users to correctly understand what content or functionality is identified by that label. To make the IA as usable and meaningful as possible, the labelling used should reflect the words/terminology/jargon that the audience uses. A simple example can be found in most corporate organisations where the official vernacular might include the term “personal leave” but staff refer to it simply as “taking holidays”.

As such, going hand in hand with learning the language of users, you may need to unlearn the official language of the organisation for whom you are producing the IA. Or at least bridge the gap by somehow using both terms and establishing a process for migrating (“educating”) the audience, if that is appropriate.

I find that this quickly falls out of the research, because in order to speak with people effectively you have to arrive at a point where there is a common language between the researcher and the researchee. Ideally it is the researcher who should be actively bridging the gap and adapting their terminology to match the user. By doing this you need to recognise what words and phrases users use to refer to the problem space. If the subject you are researching has specific jargon associated with it, then this is something you need to pick up, and in most situations it would make sense to use that jargon when labelling the IA (assuming that the audience is familiar with that subject). The bottom line is: speak your audience’s language.

While research can inform the IA design process, it won’t necessarily give you the answers. The hard part is still to come up with a solution, which through no fault of your own might not be perfect. A great deal of compromise is often involved, whether it be regarding what can be implemented within the budget available, or a lack of support for taking on board the results of the research.

Encouraging user involvement

Ok user research is what we need. How do we get them involved? Encouraging users to get involved is relatively easy. Costs can be high if you outsource recruitment to a market research firm, but there are ways to cut this almost to zero if you’re willing to do some of the heavy lifting yourself. Such as, recruiting participants yourself through internal connections or perhaps using a survey or poll on your website (eg Ethnio).

You can even just grab people in the hallway, run concepts or sketches past them, rather than go for full on usability testing in a lab. The approach really depends on your objectives and available resources.

To help encourage this user involvement you should offer a form of incentive, either cash or some other compensation. But often the audience are keen to be involved, either because they want to help improve the product or service, or because they want to let you know how good/bad it is. Of course if you’re recruiting non-users or non-customers then this won’t be a factor, hence it’s usually a good idea to include both users and non-users in your research to see both sides of the fence.

Even before recruiting for a specific piece of research or testing, you should be open to feedback and suggestions from your audience. This might mean asking for feedback on your website, through an email address or a tool such as UserVoice. Or you could sign up to Twitter and listen to you users/customers. Or maybe you can just stop hiding your customer service phone number in the bowels of your website (in 3pt text).

No matter which approach to recruitment and research you take, finding or making opportunities to get in front of users is relatively easy. Encouraging your designers, developers and managers to allow user involvement is somewhat harder.

Many people (including Dan Szuc, Joel Flom, Jared Spool and Bruce Temkin) have talked about the idea of an organisation requiring a certain level of maturity before things like user centred design (or “service design” or “design thinking”) can take place. And user involvement is definitely one area in which this is quite evident. Sometimes an organisation or a team just aren’t ready to have user involvement.

It could be that they’re too focussed on establishing the business and getting products out the door; refining the user experience just isn’t a priority.

It could be because they are scared of what users might say (this is also a very common reason for the resistance to social media that many companies have).

It could be that the team(s) involved do not want to relinquish authority over the product or its design. They may feel threatened by users having input, because their colleagues might start to question their worth.

It could be that reward and recognition in the organisation comes not through creating the best product or service, but from cost-cutting or just doing what the boss says.

It could simply be because of the time and cost. It’s a popular view that everything slows down if we have to go and do user research. Or my favourite teeth-clencher: “this is an Agile project, so we don’t have time for research”.

The solution? IF you can sell the idea of user experience as a crucial part of the business, then the user involvement becomes a no brainer. You have to research and you have to test if you want to nail the user experience, and IF the organisation sees that bad user experience means loss of revenue, lower loyalty and higher costs then they’re going to want to make time and budget for those activities. Even in an Agile world.

Those are, of course, rather large IFs. Selling the idea of UX being crucial to business success is much easier said than done. Even in companies whose sole business is products and services that rely fundamentally on a digital user interface (websites, software, mobile etc), where you would think it’s be an easier sell. Alas, it isn’t.

This has become a hot topic in recent years, with excellent advice coming from such wise people as Dan Szuc and John S. Rhodes, to name but two. I’m not going to repeat all of their great suggestions for prioritising UX, but rather I’ll discuss a few tactics that I have seen work.

• Gain their belief – don’t expect respect and understanding, rather “gain their belief” (paraphrased from Mark Schenk’s great post on successful leadership). You’ll do this by showing you’re effective; make some quick wins, get a few runs on the board, prove why you should be given the time/budget/resources you need. This requires choosing the right projects to demonstrate value and impact.
• Communicate clearly – as I said before with reference to labelling, ditch the pseudo-scientific jargon and geeky terminology and speak the language of your audience, which in the case of selling UX is likely to be business and management types. This will probably require you to drop the ego (something which will work wonders with all the suggestion I’m putting forward here).
• Listen to the business – they might just tell you what they’re looking to get out of your relationship, thus giving you something to aim for and hopefully exceed. You may need to study up a bit in order to understand them and what they do; get your business groove on!
• Make deliverables visible – many products of UX work are very useful for (some might even say their sole purpose is for) attracting attention and generating discussion around the work you’re doing. Personas and concept models are two types of deliverables that immediately some to mind. Stick them on the wall in a high traffic part of the office. Make sure it’s clear who created them and that they welcome feedback. You don’t need to go so far as the life-size cardboard cut-out personas some organisations have made, but a nicely presented A3 poster that clearly communicates an idea or concept will do wonders for the visibility of the UX team and the work they do.
• Tell stories – they are a great way of breathing life into what can be a rather dry subject, tell stories of your success but also of UX challenges. The latter works in the same way as that old usability cliche: show management a video of usability testing where the user becomes frustrated by the product and they will probably get on board the usability train pretty quickly. Better yet invite key stakeholders to research and testing sessions. It can be tough letting them watch but you’ll need to get past that if you really want them to take you and your work seriously.
• Collaborate – break down the silos, drop the ego (again) and look for other parts of the business who have overlapping or complimentary skills and capabilities. A UX “Community of Practice” can be a good, low cost way to kick off this co-operation between different parties with an interest in UX. In many ways such a community approach can be much more effective than a dedicated UX team going it alone. And the cross-pollination effect can be a real bonus, not only in terms of gaining knowledge and skills from other teams but also because their reputation, legitimacy and respect can rub off on you, hopefully in a positive way.
• Be passionate – if you’re boring and act like the situation is dire, then why on earth would anyone want to encourage or support you?
• Find executive champions – I put this one last because ultimately what all these ideas are pointing towards is getting someone with authority on board and help you change the way the organisation functions (to a certain extent). Look for like-minded or sympathetic people who “get it” throughout the organisation, use them to help spread the word and build a case for doing things the right way. Usually these allies will be the result of demonstrating value through a project (see the first point above) but they can also be the result of networking performed “off the clock” so to speak. A born networker is thus a valuable asset for the UX team. If nothing else they can give the UX “elevator pitch” when an executive is more susceptible down at the pub.

This entry was posted on Wednesday, September 30th, 2009 at 2:24 pm and is filed under Conferences, Design research, IA, Speaking, User experience. If you enjoyed this post, you might also like to leave a comment or follow @patrickkennedy on Twitter.

You can also trackback from your own site or follow any comments made on this post through the RSS 2.0 feed.