May 18, 2011: Sweat the important things

I've used the following diagram in many workshop presentations (like this one). It's homely, but it always seems to resonate well with at least a few of my workshop attendees:

Just got asked to explain what's going on here, and realized I'd never done so outside the workshop context. So here goes:

One reason sites suck is that so many of us have forgotten why our sites exist. We get distracted, lose sight of priorities, and end up with sites that don't do the most important things users want. Such a site is kind of like Swiss Army knife without the, um, knife.

This simple little report card forces you to (re)prioritize what your site's there for in seven simple steps:

1. Identify critical audience segments
2. Determine what each really, really wants
3. Determine what stakeholders really, really want for those segments
4. Combine #2 (wants) and #3 (needs) through negotiation
5. Evaluate performance
6. Fix what's not working
7. Repeat regularly

Simple, eh? Yet a frightening proportion of organizations I've worked with can't:

1. Identify critical audience segments
2. Determine what each really, really wants
3. Determine what stakeholders really, really want for those segments

...and so, they're completely screwed when it comes to doing these things:

1. Combine #2 (wants) and #3 (needs) through negotiation
2. Evaluate performance
3. Fix what's not working
4. Repeat regularly

Really, how can you operate a site if you don't know who the primary audiences are, what their critical needs are, and what stakeholders want for each audience? Yes, I'm talking to you; don't try to hide.

And, buster, if you're not evaluating your performance on major tasks, how the hell can you know what to fix?

By the way, repeating this regimen regularly (step #7) is to costly, pointless redesigns as garlic is to vampires.

Some elaboration:

• Users' wants come from (drum roll, please) user research! I'm a big fan of site search analytics as one way of coming up with a priority-ordered list of wants, but you should use other approaches, like clickstream traffic analysis and inbound call analysis. Put another way, what methods do you use to determine which tasks to include in task analyses? Inject them into this column.
• Users' most common wants will account for a huge proportion of all their wants. See Zipf Distribution. Put differently, a little inevitably goes a long way, whether we're talking search queries, document usage, you name it. Use this strange law of nature (and IA) to your advantage.
• Users' needs are what stakeholders think users need. E.g., Stakeholder: "We think all applicants will want to know about our alumni profiles". You: <sarcasm>"Sure..."</sarcasm> So populate this column by asking stakeholders.
• What's this negotiation stuff in step #4? It's where you show your user data to stakeholders, compare notes, and combine what they've come up with anecdotally with your own evidence-driven approach. This is the hard part, but this is the stuff that separates UX mice from UX men. Uncomfortable negotiating? Congratulations: you've just hit your career ceiling.
• It's an academic example; so I used letter grades. Score things however you want; just score them, baby! That's how you'll know which need to be addressed.

Like this stuff? Then I've got a workshop for you (one more this spring--in Chicago, June 3).

email this entry

Comment: Victor (May 18, 2011)

That's an awesome table. Feel free to Creative Commons that :-)

Comment: mpt (May 19, 2011)

Looks like this slide is http://xkcd.com/773/ taken to the next step.

Comment: Lou (May 19, 2011)

I love that slide. It's many times more enjoyable than my table. But the table has been around for ever (well, about five years at least)...

Comment: lynn boyden (May 19, 2011)

That is a terrific matrix, Lou. We are going to use it next week on a project that we're working on, but rather than present it as a whole fait accompli, we're going to make it more of a process tool. We'll present the first two columns as a way of showing the research we've captured, and then have a meeting dedicated to completing the third column. Then we'll take it back and do user research, heuristic analysis, etc., to fill out the fourth column.

You keep rocking my world, dude, and I for one am grateful. --lynn

Comment: Lou (May 19, 2011)

Cool; glad you like it! And definitely do this as a sequential exercise. Very few things in IA should be done all at once. And just once.

Comment: Mark Thristan (May 21, 2011)

I like this Lou - I keep banging on with my "clients" about the difference between a) expressed wants or desires and b) actual needs and then between user mental models and stakeholder business models. This is a lovely, simple approach (sort of a mini-Kano analysis in a way...)

Comment: Dey Alexander (May 29, 2011)

This idea also works for content on a given topic/page. And people are just as difficult to deal with when it comes to getting them to focus pages on what users want/need.

I tend to call column 2 'What you wish users wanted' and then column 3 becomes 'Wants/wishes'.

Comment: Alberto Mucignat (Jun 14, 2011)

Hi Lou,

we work with mental models, whose idea seems to me very close to your slide.

so why don't you use them instead? what do you think about it?

thanks

bye, Alberto.

Comment: MBT (Jun 17, 2011)

Way to focus and straight to your point, i love it. Keep up the work people. Dont let anyone stop us bloggers.

Comment: Tory Burch Flats (Jun 17, 2011)

This was a very usefull post. A great big thank you.

Comment spam has forced me to close comment functionality for older entries. However, if you have something vital to add concerning this entry (or its associated comments), please email your sage insights to me (lou [at] louisrosenfeld dot com). I'll make sure your comments are added to the conversation. Sorry for the inconvenience.