Identify the local and global maxima in your optimization efforts and focus on the site and app changes that matter most.

Have you ever spent days or even weeks optimizing a product, only to find that your changes are having less and less of an impact? How do you escape this frustrating cycle and refocus on making changes that really matter? The solution, much to the disbelief of our high school selves, may take its cue from calculus.

Don't worry; this will be painless.

In calculus, a local maximum is the highest point in a small subset of data. A global maximum, on the other hand, is the highest point in the entire data set. Take an example from real life: A roller coaster has several local maximums (the highest point of each hill), but only one global maximum (the highest point of the entire ride.)

In UX terms, reaching a local maximum means you've optimized the current iteration of your design to its fullest potential. To continue moving forward, you have to take a step back and look at your UX more holistically. You've got to zoom out from the specific page or experience you're optimizing for, and look for the big-picture opportunities higher up in the marketing funnel.

From Joshua Porter's presentation on Metrics Driven Design

In other words, optimizing for the local maximum will only get you so far. To take your user experience to its greatest heights, you've got to find the path to the global maximum and scale it all the way to the top.

Let's look at how Go Pollock used FullStory to do exactly that.

First homework assignment: the demo page.

Go Pollock is an online tool that helps teachers gain immediate feedback regarding their students' level of comprehension during a lesson. Educators create questions and students answer on their own devices.

The Go Pollock team's challenge: Address the low number of site visitors engaging with the product demo, which required selecting a set of questions to see how the tool works.

The original Go Pollock demo

Watching a few FullStory session replays quickly revealed the source of the problem—visitors were unclear how to start the demo—and the specific trouble spots in the design that were causing confusion.

Based on these findings, Go Pollock's designers made the following changes:

The revised Go Pollock demo

After making these changes, Go Pollock saw their conversion rate on the demo page climb from 45% to 70%. Case closed. Mission accomplished.

Or was it?

While the Go Pollock team successfully used session replay to optimize conversions on the demo page, did other opportunities exist higher up the marketing funnel? What about the users who landed on the homepage but didn't make it to the demo?

Perhaps this initial victory was just a local maximum, meaning the bigger prize of the global maximum was still up for grabs.

Go Pollock goes big.

Go Pollock pivoted their optimization toward initial engagement with the site—the homepage. As the most visited page on the site, the homepage sits at the top of the conversion funnel and is key to driving far more conversions through the funnel than the demo page could alone.

To optimize for this global maximum, Go Pollock turned to session replay once again, this time filtering sessions to watch only the ones where users dropped off before reaching the demo page.[1] Watching these sessions led the team to surmise that users were bouncing because they didn't want to register or fill out a form before starting the demo.

The solution? Go Pollock removed registration requirements and issued a clear, strong call to action to access the demo right away. As you can see on the new homepage,[2] visitors no longer have to spend time in detention filling out a form. Now they can simply click on the button to launch the demo and go straight to the head of the class.

The revised Go Pollock homepage

By zooming out from optimizing the local maximum to the global maximum on their site, Go Pollock opened the floodgates to conversions downstream.

Make the changes that really matter.

The next time you think you've maxed out your design or UX opportunities, consider if it's because you've successfully conquered a local maximum and need to aim for the global maximum instead.

Keeping this principle in mind will help you leave one success for an even bigger one, avoid the path of diminishing returns, and reach the rarefied air of first-class user experiences.


  1. If you've never used FullStory's "Has Never" search to find and play back sessions of users who don't do [something], you don't know what you're missing—or what they're missing, either. ↩︎

  2. Go Pollock's homepage deftly uses whitespace—as well as a consistent blue highlight color—to draw the eye to the call-to- action button. ↩︎