Design

Revolutions don’t come out of A/B tests

4 min read
Eli Woolery
  •  Nov 27, 2019
Link copied to clipboard

We learned a lot from our conversation with Abigail Hart Gray, Director of UX at Google, but this was one of the top insights she shared during our interview for the Design Better Podcast.

We’d heard her say something similar when we spoke to her last year for the Design Genome Project when she held the role of VP Head of Design at Northwestern Mutual. She mentioned that A/B testing is great for small optimizations.

But if you need a revolution—an experience for your customer that’s exponentially, not incrementally, better—you’re not going to find them there, even if you run 100 at a time.

She elaborated during our recent chat, saying,“A lot of companies are coming up against the fact that they started with an experience at X point in time, and they’ve been iterating on it and adding to it, and there are new features or market conditions change, and all of a sudden, you’ve got an experience that’s a little bit Frankensteined.”

And if you want to make a 10x improvement, you can’t do that with small A/B tests.

Instead, Abigail advocates for finding out what is important to key stakeholders and learning where you can make an impact on the business with design in an area that’s currently being overlooked.

As an example, she shares a story from her first year at Northwestern Mutual when she asked their EVP of client digital experience what he cared about, and he said he wanted to increase client adoption.

She saw an opportunity with the client portal, the dashboard customers landed on as soon as they logged in. Abigail decided “that was a fantastic test case because one, metrics wise, there was nowhere to go but up, and two, nobody really thought that much of it. So if I did something that people didn’t like, it wasn’t going to negatively impact any major company metrics in any way.”

Abigail took a small group of product people and engineers and did a workshop—the first design workshop the group had ever done—and within six weeks they had redesigned, rebuilt, and launched the dashboard. The results were almost immediate:

“…within six weeks, we were getting some really good indicators that things were going well. We had an over 300% increase in account aggregation. We had a 29% increase in client adoption. We saw an 86% reduction in the bounce rate. And I’ve found that when you take something really low value, you can turn it into something that really works for you because people see you as almost like a miracle maker, right?”

A/B tests are great for making incremental improvements to your products. But if you’re looking to make a 10x change, and something that will show big returns on an investment in design, follow Abigail’s lead. Learn what is important to your business, and find an overlooked area of the product where you can make an impact. You just might find the revolution you’re looking for.

If you want more great interviews with design leaders like Abigail, subscribe to the Design Better Podcast on Apple Podcasts, Spotify, or wherever you get your audio, so you never miss an episode.