Enterprise UX personas - on strategic models and design mistakes

Profile picture for user jreed By Jon Reed August 14, 2017
The value of personas in enterprise UX remains a subject of debate. At Enterprise UX 2017, Robert Reimann explained the problem with personas, and why enterprise UX needs strategic models.

I've been simmering on meaty pressos from Enterprise UX 2017 ever since I left San Fran. One keeper: Robert Reimann's session on Taming Design Complexity with UX Models (see YouTube replay).

After his talk, Reimann explained why enterprise software is still catching up to UX hype. He also explained why he provokes designers to think beyond tactics - and look at modeling frameworks.

A Senior Manager of Experience Design at athenahealth, Reimann is also the co-author of About Face: The Essentials of Interaction Design. He's got over twenty years of design craft under his belt; little wonder it took me more than twenty days to grasp his thinking.

We're still passing too much enterprise UX complexity onto users

Reimann's first UX projects involved ERP and enterprise software. Though he sees UX progress - including the impact of consumer UX - we aren't there yet. As he told Enterprise UX 2017 attendees:

Enterprise software, remains, well, let’s call it a work in progress. I bet if you think about it, even with the great work you’ve all done yourselves, most of you can point to at least a few areas of your products that seem stuck in the past, don’t deliver on promises, or just plain don’t work for users. And then there’s all the enterprise products and services out there that still get precious little UX love.

Then the kicker:

So why, after a decade or two of UX thinking and awareness, is that still the case?

Reimann cited three main reasons:

  • Enterprise software still predominantly serves corporate goals, but rarely serves people goals well.
  • Enterprise software doesn’t usually scale in all the right ways for any given company.
  • Enterprise software is really complicated, not just to use, but to design. Maybe even more so. And that complexity ends up getting passed on to users. (emphasis mine).

The question so many UX professionals - and project managers - obsess about:

So how do we eliminate that feeling of dread most users experience every time they launch the enterprise system they are required to use?

Taming design complexity with strategic models

To answer that, Reimann referred back to the prior speaker, Tricia Wang, who delivered a memorable talk on Solving Sexy but Important Problems. Wang spoke about how companies get blindsided by their "unknowns." Reimann's method is about making those unknowns known- "particularly those regarding what users really need."

"Taming design complexity" with models sounds appealing - but Reimann is not talking about the wireframes and flow diagrams UX designers are familiar with. He laid out a strategic UX model, which follows from a "model of models", proposed by Scott Berinato of the Harvard Business Review. You can see Berinato's image here, but he defines four models on two axes:

  • Concept Driven versus Data Driven
  • Declarative vesrus Exploratory

Along the "Declarative" axis are "Illustrative" models. Reimann's talk hit on five:

  1. Enterprise user models (personas)
  2. Data ecology and sequence models
  3. Organizational models
  4. Concept maps
  5. Systems models

Obstacles for enterprise personas

I find personas are either overrated or under-utilized. Often, their uses are misunderstood. So what is Reimann's take?

Personas have been around for a long time. They have their ups and down in terms of popularity.

Reimann says the history of personas were driven by enterprise work. Then he surprised me:

The enterprise is the sweet spot for personas.

OK, I'll bite - why?

Enterprise personas are sort of clear in one aspect, because they tend to be based on roles. You have different roles in an organization. Those map pretty well to personas.

Alas, it's not that simple. Reimann sees several obstacles to persona effectiveness:

  • User research is needed - "There's no way that somebody has the knowledge to know what's going on for a technician at a hospital. The research is really required." Note: for more tips on user research, check my piece from Enterprise UX 2017, Designing for policy change - Ariel Kennan on applying design to New York City's homeless problem, and my 2015 article, You can't build a good enterprise UX without user research.
  • Persona proliferation - "In enterprise software, because of its breadth and depth of complexity, you can run into issues where you're managing, 'How many personas do I really need? Do I need a small set that sort of broadly describes everything, but doesn't get into the details?' That's not totally useful."
  • Complexity results in abandoned personas - "[These issues can] cause some people to abandon them. Some people make reams and reams of personas for everything and then, well, how do you manage them?"

Daunting problems - so what do you do about it? Reimann advises a "tiered approach" to personas. The top level personas allow a company to map out its role on the marketplace and what the major functional roles are. The second tier defines personas for each functional area.

Reimann also stressed another areas companies overlook: the need for non-user personas.

There's always the customer who is buying the system, and you want to make sure that they're happy, because otherwise they won't buy the system.

A good persona model incorporates both:

You don't want to short change the users eithers because we know that when the users of an enterprise system are shortchanged, they circumvent the system.

Personas in action - field examples

Reimann brought these points home with field examples. From healthcare, a nursing home example where the nursing home resident (a non-user) is a crucial persona:

A resident at a nursing home never touches the clinical or financial system there, but if something goes awry there, or if it's not designed with the long view in mind that the patient needs to be healthy and happy and not overcharged and not mischarged - there's gonna be problems, both financial and also ethical.

Companies can also use the other models Reimann covered to extend the utility of personas. He cited a data ecology model from an "e-discovery" system, legal software for use by corporate lawyers. This e-discovery can literally involve millions of documents, from email to scanned contracts. Once all those documents are scanned, they must be reviewed by lawyers and tagged and redacted. An entire industry has grown to serve these needs, including industrial scanning services. The role of modern software? To automate as much of the process as possible.

Mapping out that data ecology pays off: "When I did the design work, we mapped this out so that we really understood, where is everything going, and what's the process, and how are we doing it?"

If you're moving ahead with personas, Reimann's views fuse nicely into the 2015 piece I did on Everett McKay, The virtues and perils of simplicity. McKay's persona advice moves us from strategic to practical:

  • Keep persona definitions simple – a page or two of bullet points at most for most projects. “At Microsoft, the Windows Vista team had 20+ page documents for each persona,” says McKay. “That was just way too much detail, so few people bothered to read them.”
  • Focus the persona on the specific feature you are designing. “The only purpose of personas is to help your team make better design decisions. If an attribute isn’t relevant to making decisions, get rid of it—it will only get in the way."
  • Don’t bog your personas down with hobbies and off-topic interests.

My take

Reimann's take on enterprise UX aspires to a higher level of strategy. This tone helped set the Enterprise UX show apart from a tactical design show. Those with no expertise in UX can learn from these models. I recommend checking out Reimann's talk in its entirety to get a bigger view, beyond personas.

Exposing designers to strategic models is part of Reimann's mission. Delivering his message, he is part theorist, part UX provocateur:

This is out of the sphere of UX. It's coming from a different world. I'm basically trying to introduce tools that could get people to think about the UX domain in a bit of a different way.

If we think we've got enterprise UX figured out, we are making a big mistake. But Reimann believes that system thinking can help:

Designing for enterprise systems is a daunting challenge... What I'm provoking people to do is think of it more at the systems level.

That should keep Reimann busy until Enterprise UX 2018.

End note – this article is part of my ongoing diginomica Enteprise UX series. You can check out a number of the speaker slide decks from the conference on the Enterprise UX 2017 web site. Reimann's video presentation from the show is available for replay.

Image credit - Businessman searching puzzle solution on a rooftop 3D rendering © sdecoret - Fotolia.com.

Disclosure - Disclosure - Disclosure - Enterprise UX provided me with press access to Enterprise UX 2017.