The user profile thing – sliced, diced and reconstitutedby Kate on 10/11/2012
In my time of working with agencies and creatives I’ve come to find that one of the first things that people want to create is audience profiles. At the best of times they’re woven out of formal research but they’re often (in my experience) conjured up using information from business owners or employees, perhaps some personal familiarity with the user group, and a dose of assumption.
They can be fun to create; I’ve also personally not met a client who didn’t enjoy a characterfully written profile. However, and I stick my neck out here, I can’t say I’ve found them to be all that useful. Except for perhaps at the start of a project when you need to loosen up mental juices and familiarise yourself with just who you’ll be creating for – this is a very useful use for audience profiles. Then again, they’re pretty nice to refer to when you feel you’re losing your way or can’t remember why you decided on something. They’re also good as an easy-to-assimilate record for anyone else who later comes to work on the site.
So perhaps I’m eating my words here; perhaps I do find them useful; perhaps it’s just the way they’re typically created that needs some slicing, dicing and reconstituting.
Here’s my first concern
If profiles are not based on solid research and/or data, their assumptive’ness and therefore ability to lead down the wrong path, makes them more risky than rewarding.
Considering the amount of work and budget that goes into creating a really decent and trustworthy set of profiles, for the few instances in which they might be useful, are they really worth it? Perhaps ‘yes’ if your client has riches to squander, but my clients are most often start-ups, agencies, educators, small businesses, and not-for-profits. I don’t want to waste their money on fancy-looking but perhaps inaccurate bumph that makes me seem professional but which is not just wasteful but perhaps even wrong. At the end of the day, my only aim is to help clients in offering their audiences a website that is easy and pleasant to use – a site that allows people to almost effortlessly complete the task/s they came to complete (or were inspired to complete while visiting).
It’s worth noting, that I am muddling and grappling with these thoughts from an IA and UX point of view. If I were thinking of tone of voice and engagement, I would be interested in knowing a lot more about what makes my audience tick and, in this case, a set of well-researched audience profiles might be just the thing.
I’m finding that…
I’m finding for the most part that, when working on a website that is pre-existing and has loads of data (in the form of analytics or surveys) that it is far more useful to initially focus my attention on tasks than on users and profiles. In other words: until I know what I should focus on, how do I know if I need to care whether my user is married, 33, has five kids and prefers to browse my website at night? I could spend time gathering all this data and find that it really has no bearing whatsoever on how the website or app etc. should be designed. So, in terms of creating profiles, I don’t start with the nitty gritty details of ‘Who are the people who come here?’, I start with a list of the most common user types and then research, ‘What are the most important tasks these people want to complete on this site?’ and build up profiles from there. The distinction might seem subtle but the result is quite different.
How I go about this
I find out what tasks are most commonly completed on the site and how regularly users come to complete those tasks. Considering each task, do users come to complete it all the time, once every now and again, or both? What are the circumstances in which they complete the task? Do those circumstances affect how something should be designed and where it should be available? The people who complete the task, who are they? Are there any special details which affect how they complete this task?
It’s out of these details i.e. out of details surrounding the tasks that a kind-of audience profile starts to emerge which, in my mind and experience, is of far more use than cooking up an audience profile simply by looking generally at characteristics and lifestyle like ‘Polly, 35, unmarried’. Unless it really matters to that task, who cares if Polly is unmarried?
I can leave all the detail which doesn’t offer useful insight regarding UXD, content or whatever out. It also means that you don’t corral users under user groups but rather under task groups. So, I could have three different types of users who all complete the same task and perhaps have different needs of that task. By describing user needs through the lens of a single task, you really end up building not just a profile of that task group but essentially a documentation of the kinds of things the design should address, a task specification in other words. As I mentioned, I’m experimenting (grappling was the word I used earlier) with this and haven’t got it all entirely figured out, but I reckon I’m close.
A real-life example
In a current research project, I’ve got several primary users groups, which were defined by way of general research, a basic top tasks study (I didn’t have remit for the full process), an in-depth study of Google Analytics, and research regarding the organisations remit. User testing, interviews and a survey were done before my time and had been well documented; the survey was especially useful.
Referring to the top tasks study, the most important task (by a long way) completed by several of the user groups is checking a set of registers. I called this groups of users, ‘Register Checkers’.
Other top tasks described other user groups. For instance, a prevalence of training related tasks showed that Trainees are a primary audience requiring a specific series of information, which could be prioritised using the top tasks study. Then there were a few specific tasks, which related to overseas enquirers.
I’ve also defined a public group who were not prominent in either the top tasks or Google Analytics study as they visit the site rarely and irregularly. Though the tasks they complete are not common, their tasks are very important in terms of the remit of the organisation. Also, as the users are generally first-time and once-off users, these tasks need to be extremely easy to find and complete.
So, ‘Register Checkers’ are generally very, very regular users of the site (specifically of the register checking tool) and, so long as the register checker stayed more or less as is, they’d know where to find it and how to use it. This group of users have educated themselves as to how the site works and, if any changes to the register checking process were made, special attention would need to be taken to introduce them to the changes.
The Trainees are a large user group and have very specific tasks to complete at certain times of the year. I would need to make sure that they have a unique user journey with a clear focus on their completion of certain tasks at certain times of the year. They’re perhaps the only group I’d want to, at least initially, profile with more detail.
The overseas user group are usually first-time or highly irregular users and need to be able to easily find, from the homepage, the very specific information they’re looking for. Top tasks made it possible to define the information they’re looking for and, whilst further research regarding their specific information needs would be useful, I wouldn’t spend too long in developing an extensive profile of them.
As mentioned, the general public have certain tasks which are uncommonly completed but which are important in terms of fulfilling a primary remit of the organisation. I would need to make sure that these first-time or irregular users would have no difficulty in completing their tasks starting from the homepage or from Google search. In aiming for excellent usability regardless of the audience, I’d not spend time or budget profiling this audience but rather put effort into specifying their exact tasks and developing a design that is as simple and fool-proof as possible for even the most techno-silly of people.
You may have noticed that I’m mixing methodologies – I’ve got a profile which is defined exclusively by a single task i.e. Register Checkers, and I’ve got other profiles which are more conventional – they are definable user groups or types of people – though they have been defined by the tasks which are unique to them and not a set of arbitrary lifestyle characteristics.
If you like things in straight lines this mixed method approach might annoy you but I feel, why not? If I approach each instance of tasks and users with a method that works best in defining the real needs of users regarding the website and the things they want to do, does it matter that some are defined as a group under tasks and others as a group defined by tasks? I don’t think so. It’s as if I’m approaching each group of people and their tasks/needs as a unique project, or mini-project, but still maintaining a vision of the wider design.
In the convoluted world of digital design where sometimes it’s content that’s the problem, sometimes it’s architecture, sometimes (and most times) it’s a bit of both, it’s hard to know when profiles are needed or when you should ditch the profiles and view things through the telescope of tasks (or do a bit of both).
When I’m working on content and seeking to emote audiences, then personas, with all the ticks and tastes of a knowable human character, offer me the inspiration I need to produce to who the audience is. However, when I’m focusing on user experience, a design-oriented and mixed-method approach may be better than a process that aims for consistent and pretty-to-look-at profiles that are very client presentable but which perhaps aren’t the best was to express or deal with the unique needs of each task and/or user group. Some tasks after all will be more content focused and others more design focused, others a bit of both. Shouldn’t I respond to this?
I guess it’s the always-fascination of working on the border of new-world digital and age-old human interaction – nothing is black and white or absolute. I don’t think it’s useful or progress-orientated to get bogged down in ritual and process just because it’s how things have been done in the past or it’s what you’ve read in a book, because it’s what’s worked in the past, or because you’re scared to change things up a bit and try another way.
Perhaps my irreverence and chop suey approach to process will burn me in the end but, for now, I’m going to hold out and say: beyond profiles, never stop slicing, dicing and reconstituting the processes you hold dear.
Besides, isn’t it precisely this that makes a techno-career so mind-sizzlingly, seat-of-your-pants, never-the-same not-boring?