Identifying the Metrics of Good Design

Recently, we talked about the importance of figuring out the true value of the design work we do for our clients and our products. Far too few designers and UX professionals can say with certainty what impact their work has on a product or project, which makes identifying that data all the more important.

But how do we hold ourselves accountable as designers if we don’t know what metrics we should be looking at?

In this post, we’ll be looking at metric-driven design. We’ll be examining what metrics matter most to designers and UX professionals and how we can hold ourselves accountable to ourselves and our teams.

Why Should Designers Care About Performance Metrics?

Before we jump into the metrics we should be focusing on as designers, you might be wondering why we should care about metric-driven design at all.

There are three reasons why metrics are crucial to our success as UX professionals. Let’s take a look at each individually.

Failure is Awesome

Failure isn’t often seen as a positive thing. But on the contrary, failure is actually awesome. Why? Because until we fail, we don’t know which areas we should focus on in order to improve our work. What happens if we fail to hit a target? We now have the chance to improve and fix it. Nothing ventured, nothing gained.

There’s More to Product Than Pushing Pixels

Product managers and designers often have one near-fatal flaw in common: the pursuit of perfection. There’s nothing wrong with trying to design a “perfect” product in theory, but the problem with chasing perfection is that it’s unattainable. If we try to make perfection the standard by which our work is judged, we’ll never achieve it.

Setting specific, tangible targets allows us to quantify the success or failure of a feature or design. When you set a target, you can evaluate a design feature according to whether it accomplished its purpose and met that target. Another way to think about this is to focus on designing for specific outcomes, not impossible ideals.

Celebrate Real Accomplishments

Another reason why we should care about metrics is that doing so allows us to celebrate real wins. Just because something was built doesn’t necessarily mean that your users will appreciate that feature or that it will be of genuine use.

Don’t get excited about shipping products — get excited about hitting specific targets.

Metric-Driven Product Design

Now that we’re all firmly on board with the concept of metric-driven design, let’s take a look at a couple of examples of what this looks like in practice.

Twitter Boosts Sign-Ups by Leveraging Gradual Engagement

The first few years of Twitter’s existence were defined by rapid growth and frequent product updates. Although the aesthetic and visual continuity of Twitter as a platform also changed dramatically in the site’s first five years, Twitter didn’t just tweak how the site looked — it tweaked how the site worked.

During the course of its user research, Twitter learned that while its celebrity users were attracting new users, they weren’t being retained. Some user churn is inevitable, but Twitter decided to take a data-driven approach to refining its retention workflows by emphasizing the things that interested people. To do that, Twitter redesigned its sign-up process.

Here’s how it looked before the redesign. Notice how Twitter used to focus on importing contacts and helping new users find accounts to follow?

That approach was problematic in several ways. Not only did new users have to go through the sign-up process, but they also had to connect their mail provider to allow Twitter to search for contacts with Twitter profiles. After that, users were prompted to follow some of the most popular accounts on the platform before finally being permitted to start tweeting themselves.

The flow wasn’t just cumbersome — it actually created multiple points at which new users could be overwhelmed by the information asked of them and abandon the sign-up process altogether.

To combat that and drive engagement, Twitter redesigned its sign-up process to emphasize interests. The initial account-creation screen was left unchanged, but the next screen users were presented with gave them content suggestions organized by topic. That allowed new users to quickly and easily identify accounts that pertained to their personal interests, thus significantly increasing the likelihood that new users would become regular, engaged users.

What’s really fascinating about this redesign is that the change might never have happened if Twitter hadn’t adopted a metric-driven approach to product design.

The redesigned sign-up process actually increased the number of steps in the sign-up flow from three to four — something that few product managers would sign off on. But Twitter wasn’t relying on intuition or hunches. By redesigning the sign-up process in this way, Twitter actually increased sign-up completions by almost 30% — a perfect example of how focusing on certain metrics can, and sometimes should, take precedence over conventional wisdom about product design.

Facebook Decreases Deactivations by Leveraging Emotional Triggers

Facebook has always given users the option of deactivating their accounts as a way to discourage permanent deletion. However, while deactivation is certainly preferable to deletion, it’s still not ideal.

To reduce account deactivations, Facebook’s designers examined the site’s old deactivation form to see what was going on. On the old page, Facebook asked users why they were deactivating their account, but that information was of limited use. One designer suggested personalizing the deactivation page to make it a more emotionally charged experience by showing deactivating users images of the friends they’d be leaving behind:

Including images of deactivating users’ friends on the deactivation page itself had a significant impact — a reduction in account deactivations of approximately 7%, or around 1 million users per year at that time.

This is a great example of metric-driven design that still appeals to designers’ intuition. The designer who suggested personalizing the deactivation page likened his vision to that of a summer camp: the kind of place you don’t really want to leave because of the relationships you’ve built while there. A classic designer’s hunch, this had a demonstrable impact on tangible business objectives at Facebook (i.e., reducing account deactivations). The design team had a concrete, data-driven goal and was able to optimize their design work around that goal.

Designing for Specific Outcomes

As a discipline, design can be somewhat resistant to quantitative metrics. But thinking about designing for outcomes rather than designing by intuition can help you and your team identify not only areas in which your product can be improved but also tangible goals that can be optimized for. It’s far easier to base design decisions on hard data and quantifiable targets than it is to convince your executive team about a design choice based purely on aesthetic considerations or vague intuition and assumptions.

Thinking about product design with metrics in mind doesn’t diminish the creative qualities of your work — it makes it much easier to petition non-designers with compelling reasons why your designs can help the whole company succeed, not just the design team.


Originally published at blog.nomnominsights.com on January 8, 2019.