5 ways data sets the best design teams apart

As UX design and product thinking become increasingly mainstream, more and more companies consider design essential to their success.

It’s not uncommon these days to hear CEOs mention affordances or find engineers on Twitter discussing the mental models of their users. As designers, we should applaud this progress. If everyone is a designer, we’re much better off if they’re informed designers.

But while this increased design fluency is inspiring, it also means it’s more challenging than ever to claim design as a strategic advantage. Good design is now table stakes.

Related: Getting started with data-driven design

So how can design leaders give their teams a leg up?

The top design and product teams build a strategic advantage by finding more ways to integrate data into their day-to-day. Let’s take a look at five ways using data can set the best teams apart from the rest.

1. Data shapes culture

Every morning at my startup Swish, we start the day by reviewing key metrics.

Two TVs dominate the back wall of our office, shamelessly broadcasting our progress—and failures—to team members and guests alike. The team huddles in front of the screens, and we dissect the numbers, trade ideas for how to improve them, and criticize ourselves and each other with a humility I haven’t seen in any of my previous stops.

Everyone lives and breathes those numbers. It’s hard to overstate the impact this has had on our day-to-day interactions.

When every team member knows that we test and measure every product change thoroughly, it becomes far easier to avoid the type of religious debate so endemic in toxic work cultures. Nobody clings to their ideas, because they know that if another, worse idea gets implemented, we’ll tear it out just as quickly when the results come in.

Design teams that embrace data and integrate it into their day-to-day interactions have a huge strategic advantage over those that don’t.

If you lead a design or product team, look for ways to insert more data into your process. Your team will thank you for it. If you’re an individual contributor, bring data to the table at every meeting. Basing your ideas on data makes them far more likely to succeed.

Here are a few changes that have worked well for my team at Swish:

• Pick One Metric That Matters for your team based on your product’s current growth stage; focus solely on that metric until you’re ready to move on
• Keep a dashboard charting your One Metric over time as well as the current performance of the individual funnel steps for that metric
• Give everyone access to this dashboard and display it prominently
• Tie all roadmap items to your One Metric with specific improvement targets (e.g. new onboarding flow will improve user activation by 25%)
• A/B test every new feature and measure its performance against the current baseline
• Discuss performance of current tests with your team daily

2. Data makes design reviews better

Brian Pullen’s article on running productive design reviews captures the typical design review perfectly:

“Feedback plays a critical role in the design process. But instead of being productive, many design reviews merely conjure the ghost of collaboration, devolving into firing lines where everyone takes potshots at the design, or pitch sessions, where a designer gives an elaborate presentation to win over their peers.”

But imagine a design review free from subjective criques and domination by loud-mouthed team members.

The best design teams I know have adopted the simple practice of conducting design reviews after testing the feature with users. They consider user testing a prerequisite to conducting design reviews.

This encourages the team to get feedback from the only people who really matter—actual users—before discussing the merits of various changes. It also means that team members can credibly represent users with quotes and anecdotes from the interviews themselves.

3. Data helps you assess prospective team members

A few evenings every week, I coach aspiring UX designers over Skype for Springboard. It’s a fun way to help people new to the craft, and I highly recommend it to any senior designer who’s looking to give back.

The designers I coach often struggle with how to best present their work in their portfolios and resumes.

The first thing I emphasize is that—regardless of the role or the company—hiring managers only care about three things:

1. Can you do the job?
2. Do you fit in with the team?
3. Do you really want to be there?

While the last two concerns are largely subjective, the first is actually measurable. If design is about solving problems, then data can show clearly how well a prospective candidate has solved them.

In addition to including fantastic work samples their portfolios, the best designers tie their work back to real business results.

When faced with the choice between two interaction designers with comparable experience and portfolios, savvy design leaders hire the candidate who can quantify the results their work produced. This is especially true if they need buy-in from non-design team members to make the hire.

Next time you’re evaluating candidates, look for the ones who tie their work to concrete numbers.

Related: Funsize’s Anthony Armendariz on designing a design team

I trust the designer who knows that their onboarding redesign effort helped increase activation by 26% far more than the one whose designs led to some nebulous “improvement.” If a candidate can creditably say that their new interaction design increased conversions by 58%, hire them quickly before somebody else does.

4. Data-based personas keep teams grounded

In her article on predictive personas, Laura Klein, author of the fantastic UX for Lean Startups, argues that most design teams fail to create personas that are actually useful.

The question they should be asking themselves isn’t, “If I interviewed a user, would this describe her?” The question should be, “If I found a person like this, would she become a user?”

She’s absolutely right.

Personas are only useful if team members can draw meaningful, accurate conclusions from them. So how do you create predictive personas?

Use actual, real-life data.

Coryndon Luxmore has an excellent presentation on the exact steps his team took at Buildium, but it’s by no means exhaustive.

At a minimum, I recommend including the following to make sure your personas are actually tied to real data:

• The percentage of your current user base this persona represents
• The percentage of the potential user base the persona represents
• Activation, retention, and referral rates for this persona
• Quotes from at least 3 actual users that describe the jobs-to-be-done

Including these data points will make your personas far more useful, which helps keep discussions about user needs grounded in reality.

5. Data aligns priorities

In most organizations, the product roadmap is the result of an unholy amalgamation of horse-trading, hand-waving, and coercion.

Entrenched groups cherry-pick anecdotes that justify their proposals and ignore those that suggest an alternative course. The debate feels endless. The product, its users, and the business suffer.

Data-driven product teams, in contrast, often avoid these arguments altogether.

When you know the breakdown of your user base by persona, it’s easy to decide which groups are priorities. When you know which phases of your funnel are underperforming, it’s easy to decide where to focus. When your model shows the extent to which a feature could potentially drive results, it’s obvious if the bet is worth making.

Having this data easily accessible helps product teams ditch their prioritization process. Priorities become self-evident.

If you’ve ever sat through an hours-long priority meeting, you know how freeing this can be.

None of this will happen overnight

If all of these improvements feel like a fantasy to you, know that it doesn’t happen overnight.

As design leaders, our job is to find subtle, bite-sized ways to improve the the processes, culture, and results of our teams. Find more ways to integrate data into your routines. Make a data-first approach habitual.

The results will speak for themselves.

Header image by #WOCinTech Chat. CC BY 2.0.