Insights for Design

Anecdotes and perspective from a seasoned analytics and user research sleuth

The part I really care about is these anecdotes. A person may google “design insights” or “user research” and find themselves sifting through a bunch of process stuff. Instead, I wanted to just give some color to the outcomes I’ve had over the years as a design researcher.

How I feel on process. I believe in a loose definition of insights and that “data-driven” work is also largely creative work. To be clear — I am not proposing some type of insights making process or framework (yawn).


A simple observation that does not require your own research

You have it, just bring it to the forefront.

e.g. Visitors are more likely to navigate in-page then navigate to a new one > Put every product in “one page”

As a digital analyst, I had often been shocked by the lack of engagement on web properties. I cajoled my client to get on board with the “one page” product page. And the numbers are staggering.

We didn’t have the luxury of a/b testing at this stage. 1: boy would I have loved to, 2: based on the results I’m guesstimating we made a great first choice.

e.g. The first song on Spotify’s “discover weekly” playlist must be amazing to achieve traction on that playlist > Ensure that first song passes as an “amazing song”

My first time trying “discover weekly” I heard a song so obscure and incredible that I found myself very encouraged by the promise of the feature. This is a really good idea…I don’t care if you have “data”, because that is a darn good hypothesis.

e.g. Nielsen’s usability heuristics. > Yeah buddy

I still get insane value from these and am fine with working them into an insight I can throw up on the board.


Inspired by a great question, not the result

Ask and you shall receive.

e.g. Visitors spend very little time outside of one specific program > Sell the programs, not the school*

I don’t recall being particularly surprised by the result, but I’m sure glad I took the time to build 10 Google Analytics segments and put this chart together.

The client was surprised by the lack of overlap and just as important, the imagery was memorable and powerful for our stakeholders.

What did surprise me was the insight that came from the result. A lot of the content effort on the site was geared to the overarching program. It became clear that the site had very clear personas that needed to be targeted individually, not addressed in mass.

*Note: You could also have as a takeaway that the observed lack of overlap was an insufficiency of the design to afford exploring the diverse offering of programs. This would be in contrast to taking the observation at face — as a user need that applicants are only focused on specific programs.


An inherited fact you are making too much of

Because you just haven’t inherited enough legacy.

e.g. Students and Faculty listed “leadership” as the least important item they would use to sell prospective students on the school > Get leadership content out of the highest level in the navigation and de-prioritize it as a message

Despite the result of a survey, the creative director for this project studied the competition and believed that 1 on 1 “leadership coaching” was a unique and compelling offering for a program in their market. Go with your gut, a/b test, recognize that the you don’t know, and most of all, always be looking to triangulate your data sources.


A clever observation that was in plain site

Refocus in hopes for that special epiphany

e.g. Users will go out of their way to preserve context in their space > Design a window management system where staying in context is at the forefront

I noticed a very subtle behavior of users dragging a window simply to obscure less of the other windows in their space. This is similar to what is accomplished by the feature of “cascading windows.” On this project, we found out early that many users were averse to maximizing (by asking them), but through a keen observation, the insight that users want context gained significant more weight.

e.g. Knowing the weather is relevant to tasks users accomplish in their Maps app > Lets put a “67 and cloudy” in the map?!

If your maintaining an open mind, you would have noticed how often weather was part of the total cx happening in maps and this is exactly what apple maps did.


A surprising finding that is strong evidence for a particular strategy

For the lucky ones only.

….and what most people think they are going to do when they get their hands on that “data” as they start to slow on the design front.

e.g. Those who visited Doctor and Affordability pages were more likely to convert then visitors to other pages > A worhy hypothesis: Funnel users through doctor-specific, not clinic-specific conversion flows. Consider ramping up the doctor-centric strategy for content.


A simple tactic that bubbles up as a key element of your new design strategy

An initial straightforward observation may make a big comeback.

e.g. Those who apply to the MBA program spend a majority of time in admissions pages and do not visit alumni/careers/news content > Sell the program on admissions pages where prospectives are spending a lot of time

This was an early observation that users were not visiting these other pages, but as I worked with the creative director, the way to solve for that became front and center of our pitch.


The info is in plain sight and the insight is obvious

Don’t overthink it.

e.g. You have as many Facebook visitors as actual visits on the year on your television site > When I started to look at the numbers for the Natgeochannel.com redesign, I advised them to focus on the Facebook snippet designs and specific strategies that start the CX journey in the Facebook-verse.

e.g. When I surveyed small biz owners, they reacted well to “A Professional Email” over “Private Cloud.” > Message that way

What was typically being advertised as a feature of Google Apps for Business (which you could get set up on your personal domain), started to message itself as Google Domains. I also found that words like “Domains” didn’t resonate either compared to “Professional Email.” My own research with 20 users actually supported such a strategy.


You ran an A/B test, so you feel confident

e.g. Landing page “A” beat landing page “B.” > Use landing page “B”

“A > B” is “true” in a pretty specific context. This result is forever tied to the specific circumstances of that test, which will always have limited relevance to the real product. There

A/B test I managed for Alamo.com

Let’s Recap

Basically, free your mind about where your insights can come from and what your takeaways can be.

Was the insight driven by…

  • A hard number (or an a/b test) or guesstimate?
  • Your own (primary) or someone else’s (secondary) research?
  • A fancy or a darn simple metric?

Was the takeaway…

  • Perceptive or pretty obvious?
  • Top-down/design-related or more abstract?
  • Controversial or not controversial?

My point is…

  1. So much of insights work is a creative process. The outcomes are a result of the researchers choice to ask a good question, to point something out and prioritize it above other facts, and finally to sell it with appealing actionability.
  2. So don’t let your work suffer by limiting your “research” to hard numbers or your own primary research. Try and reduce your bias towards reco‘s that you found yourself or the data sources you’ve been given access to. General truths, not necessarily “data” carry so much of the weight of good design. Step back. If you are doing this work you are a designer, dude. So start acting like one.

Visit my site to learn more about me: http://rickswette.info