How to accelerate conversion rate optimization

If you had infinite time, money, and traffic, you could test every possible way to improve conversion rates on your site. But you don’t.

There are millions of things you can try in order to improve conversion rates. Most companies are lucky to get between 10 and 100 optimizations per year on their website.

With so few shots available, growth leaders can’t afford to just try stuff and hope it works. Each iteration is precious and you don't want to spend months on trial and error. Leaders want to win as early in the year as possible.

Sadly, most optimizations fail to move the needle. There is no shame in failing. Losing does help you discover what actually works.

But no one wants to lose. You only deal with losses so that you can find wins.

The most important thing is to learn as quickly as possible so you can discover what works fast. Losing tests reveal what customers value (is speed more important than reliability), what UX elements to include (should we invest in personalization), and methods of persuasion (what kind of social proof works).

Each loss can get you closer to a win if you learn from it.

Learn from everyone

The only thing better than learning from your losses is proactively learning from someone else’s losses. You can improve your win rate by looking beyond the four walls of your organization.

Scientists leverage results from other scientists - businesses should too.

Until now, you haven't been able to learn from all the ongoing split tests happening online in your industry. We are changing that.

We built DoWhatWorks to help people learn from everyone else’s trial and error. We detect and analyze tens of thousands of split tests from all the major brands so you don't need to waste time chasing losing strategies.

You can now look at what works before you plan, design and execute changes on your experience.

When you generate ideas, start by understanding what has worked and failed for your competitors and companies you admire. When prioritizing, evaluate how often the approach you are taking wins and loses on similar pages in your industry. Look at what the winners and losers have in common to execute your designs in ways that are most likely to work.

Don't just scan competitor websites for ideas. Look at what wins and loses - not just what is live. Our engine sees many brands copy the losing side of split tests run by their competitors. Don't rely on conventional wisdom. You will be shocked when you see how often tests prove it wrong.

Harvest the rewards

Optimizing is hard, but worth it. Even small improvements to conversion rates make a huge impact.

If you can run split tests, you should run as many as you can. Period. Many will fail, but the more you test, the more you learn. The more results you see, the better you will be at steering towards bets that are more likely to work.

The formula is simple: Conversion Growth = (# of optimizations) * (% that win) * (impact of winners)

Most leaders focus on velocity. If you push to get 20% more shots, you'll get more chances to win. But, there is only so much you can do to increase the number of optimizations you run each year. Testing tools like Adobe Target, Optimizely, Amplitude, etc. have made it easier than ever to launch tests, but you still need to let tests run long enough to get significant results, meaning that your test volume will be capped by your traffic levels.

While you can work to get more tests launched, the larger opportunity for most is getting more of them to win. Only 11% of split tests move the needle. That's surprisingly low, but the good news is that you can do significantly better. Our customers are getting consistent wins by using results from other major brands.

If you can double your win rates, your volume goes up while your cost of acquisition (CAC) goes down. It's like doubling the impact of your team and marketing budget.

‍

Photo credit: Marc-Olivier Jodoin