When you read case studies about a big traffic website that managed to increase email signups by some insane percentage, your brain probably thinks: "That's cool but it's impossible for me to get the same results".
That's why I'm thrilled to feature Chris in today's case study.
Read on to discover exactly what he did and how you can do the same on your website.
What's Your Hypothesis?
When you start testing opt-in forms (or any other content for that matter) it is important to first create a hypothesis.
A good hypothesis looks something like this:
Stating a clear hypothesis to why you think your test will bring in the results you're looking for will allow you to do tests that will teach you something about your audience.
Let's take button colors as an example, the hypothesis would read something like this:
Changing the color of the submit button from green to red will result in an increase in signups due to... people liking red more than green?
That's not a good hypothesis! Even if you would get a significant result (which is highly doubtful) you will not be able to apply this to anything else in your marketing material.
Compare this to the following hypothesis:
Changing the main benefit in the title of my opt-in offer from "more productive" to "more free time" will increase opt-in rates because my audience is more interested in improving their lifestyle than getting more work done.
If this hypothesis turns out to be true, you'll be able to change your sales page, your emails, your blog content... to speak to this exact benefit and increase overall conversions, not just signups.
From Hypothesis to Real Test
Chris also started out with a hypothesis:
Focusing on the relief of negative feelings will get people to opt-in because people want to get rid of these.
His testing hypothesis, that he started at the same time was:
Changing the title of the opt-in form from "feelings" to "3 powerful remedies" will boost opt-in rates because people will be curious to find out what the remedies are.
Now there are a few things Chris did that were very smart:
- He made BIG changes. Instead of just fiddling with the text of the submit button or some other small design detail, he decided to change the value proposition for the opt-in offer. Getting your value proposition right is one of the most important things you can do for your online business. You'll be able to use the insights you learned in all of your marketing materials because you know what makes your audience tick.
- He immediately launched 2 versions. You do not have to wait to get results on the first opt-in form to start testing. You can start A/B tests from day one and boost your signups. This will accelerate your learning and your list growth!
- He kept the settings the same. You should decide what you're testing. Are you testing value propositions? Opt-in form triggers? Opt-in form types? Etc. All of these tests can be interesting, but you should choose which one you want to concentrate on and test one by one. Imagine changing the opt-in form trigger from exit-intent to showing up after 15 seconds while also changing the copy of your opt-in form. When choosing a winner, you wont know what actually made the difference in opt-in rate and you'll have to test again.
As you can see, Chris ran the test for a pretty long time (from August 2016 to mid-January 2017) to get significant results. And note how in the beginning, the variation opt-in form (the blue line) is actually out-performing the original one to later stabilize and under-perform consistently.
This is the perfect showcase of why you should be patient when running A/B tests.
One of the biggest mistakes people make is ending tests prematurely, which will result in false outcomes and will hurt your lead generation efforts. The problem is, there is no right answer when it comes to the perfect time frame for an A/B test.
As a guideline, we recommend:
- To wait at least two weeks. You've probably noticed that traffic on your website varies a lot from day to day. Waiting at least two weeks will give you a better base for decisions.
- If possible, wait for 100 conversions. If you can get at least 100 conversions over the span of more than two weeks, you can consider your test accurate. If you get 100 conversions before the two weeks period is over, still wait two weeks before jumping to any conclusions.
- Aim for a 95% accuracy (=chances to beat the original) minimum. If your test doesn't hit this accuracy after a certain period of time (more than 2 weeks and more than 100 conversions) your test is inconclusive.
For Chris, the winning opt-in form is converting people at a 28.77% rate and the variation at 21.02% with a 96.3% accuracy.
Now, this might not seem all that much, but this is actually an increase of 37%. For Chris it's the difference between 63 rather than 37 subscribers... Not bad for a few minutes of work!
Most of the tests you'll run will not result in a big, world-changing difference. But relentless testing will do so over time. That's exactly how we managed to increase opt-in rates by 268% for John Lee Dumas of Eofire.
Now that Chris tested and approved one of his hypothesis, he can start running a new test.
Because this opt-in form is running on a sales page, I would NOT change the trigger. The main goal of a sales page is to get people to buy your product. Getting them to sign-up is a secondary goal which is perfect for an exit-intent lightbox.
For the same reason, I would not test different opt-in form types against each other.
These two tests would be tests that we recommend running on just about any website because these are BIG changes (an overlay vs. a lightbox, showing a lightbox on page load vs. on exit intent, ...) but not when the form appears on a sales page.
Some things Chris could start experimenting with next are:
- Home in on the copy even more. Maybe Chris can find pain points that resonate even more with the audience. In this article you'll find question prompts to help you discover additional benefits and value propositions for your free opt-in offers in just a few minutes.
- Radically different design. Not just changing a button color but choosing a completely different opt-in form template without the ebook image or with a different ebook image.
- Test a new opt-in offer. This is a little bit more work, but maybe Chris already has some content that he could share as a new opt-in offer? To get inspiration for opt-in offers, check out this article with 13 examples of brilliantly effective opt-in offers.
- Test a multiple-choice opt-in form. Asking for a micro-commitment before asking to sign-up can be a very efficient way to boost opt-in rates and is definitely worth the try.
How Can You Apply This to Your Opt-in Form?
The very first thing I encourage you to do is to add a variation to your existing opt-in form. This will only take a few minutes! Not sure how? Here's a tutorial to guide you through it step by step.
For this variation, do not try anything fancy. Simply change the title.
If you need help improving the copy of your opt-in form, you can check out our free Thrive University course: How to Create Persuasive Opt-in Forms.
In general, make it a habit to always add a test variation to your opt-in forms. You've got nothing to loose!
And one last thing:
Set up the test and be patient, even better, forget about it and let the automatic winners settings do the job for you.
Now, I would like to know:
- Are you testing your opt-in forms?
- Have you seen any results?
- Did you get stuck or are you having trouble testing?
Share your experiences in the comments below!