19 March 2015

Doing business right

A lot of my time is spent fixing broken designs, or creating new (and hopefully great) designs, and we do this through many different ways, from expert reviews and user testing to workshops and mentoring and many more.

That's all fine. That's what UX companies do.

But it's kinda wrong.

It's wrong in the sense that what is really on trial here is the business model and how we interact with it. There's no point to a great user interface that doesn't end up in some goal, be it money, or transferred message, or some other reaction. This latter compound of goals is called business objectives; it's the reason we're doing this thing in the first place.

In order to reach those business objectives we create ... a something; user interfaces, communication, a narrative. Maybe we use computers, or phones, or pamphlets, carrier pigeons. Maybe we click a button after ended call. Or we make a call after the bird lands. We have Julie talk to Mark. We have a daily meeting. Or we give out some piece of paper with words on them once dollars enter our coffers. There's a myriad of interconnecting objects, services and events that together *is* your organisation. Whatever you're hoping to achieve, it's done through this cacophony that we call the business model.

The business model is what makes or breaks your organisation, so it's very important to make it as good as possible, or as fit for purpose as we can. But how do we normally make sure it's a good one? Or the right one?

I fear the right answer is that we mostly don't, and when we do, we measure things are are only seemingly related to it, like "do we still get a salary here?" or "what is this year's profit margin?" If the organisation keeps floating, keeps staying alive, the model seems good. If it's struggling, or going under, it's probably bad. You've probably been told about KPI's, and unless you're living under a sponge, they are supposed to indicate that the organisation is, well, performing to certain standards. Key standards.

Internal KPI's are reasonably easy to find and collect (although interpreting them is not so straight forward, although that's a future blog post) To get to external KPI's, we quite often use use user surveys (and some times user group testing) to try to verify some of the basic perimeters of organisations, like brand awareness, product recognition, advertising uptake and so on, and, well, if you don't think too hard about it might give you some valuable feedback for your BI (business intelligence) . This is the bread and butter of marketing people everywhere.


In my years of doing UX, process and semantic modeling, I'm struck by how poor design is reflected in poor business models. And I have a suspicion that poor design comes directly from the models it is supposed to support. And I suspect that changing business models are more important than fixing little cogs here and there, polishing them to look new and shiny when the cogs around them are rusty and old. There's squeaking and creaking in the machinery, and new cogs aren't the answer. The old cogs need replacing before the machinery comes to an unrecoverable halt.

How can we test business models, then? I'll leave the actual organisational change to the side for now (but there's a blog post on that one brewing as well) to focus on testing the model itself. Here's an unadulterated brain-splatter of ideas;


  • Hold more regular workshops to test your ideas with people who do the actual work
  • Map your current processes (on a whiteboard, and take pictures)
  • Do internal surveys where meetings can't be help
  • Workshop : If you didn't have to worry about current constraints, how would you really do it?
  • Followup workshop : how to remove those constraints


  • Do user groups which goes beyond the superficial (for example, what price do people think is right, as opposed to what you think is right?)
  • Do user testing of core interaction points with customers
  • User third-party professionals to conceptualize your business model
  • Have someone re-design something, even the stuff that you think works fine
  • Do user testing on conceptual designs (information architecture, processes, business values, etc.)

The key for me is to sanitize your assumptions. Whatever you already know internally to be a problem, well that's yours to fix (or hire in someone to help you do that). However, are you sure the things "that work" really work? Or are working the best they could? Do you know for certain how many business goals you're missing out on by the simplest modeling snag or, at worst, error?

I wished more organisations would spend their time verifying internal assumptions; that is where the gold is. We put too much faith in people's "gut feelings" and "being savvy". Sure, we might have skills and experience, but the truth is that even the best of people fail the onslaught of all the complexities coming our way.

Test your assumptions. Test your business model. Make sure you're doing the right thing, before some old cog goes ping, gets stuck in some gear, and kills the whole machine.

No comments:

Post a Comment