Wishpond added its blog overlay about 9 months ago, and immediately saw some serious results.
But we know, after years of testing, that an early positive just means there’s potential for more growth. We’ve fallen victim to the “If it ain’t broke, don’t fix it” mentality too many times.
Just cause it ain’t broke, that don’t mean it’s optimized.
This article will take a look at one of Wishpond’s most effective tests in recent months, break down our hypothesis, why it worked, and a few ways you can implement the same lessons in your own business today.
The test I’ll be looking at in this article is a multivariate test – which technically goes against testing best practice. It was run after an initial, “proof-of-concept” campaign we did to see if the addition of a blog overlay was a good idea. Once proven, we applied design best practices alongside a test related to telling our visitor what to expect. In order to reach our monthly traffic KPIs we swung for the fences with a multivariate test. I’m happy to take questions in the comment section.
The A/B Test Hypothesis
Before I get down and dirty into this, let me show you what the original “proof-of-concept”-style blog overlay looked like:
Okay? So we knew we were going to apply design best practices (as I mentioned above) but there was still something missing…
A/B Testing Hypothesis Step 1: The Conversion Problem (Why Aren’t People Converting?) –
When looking at ways to optimize this overlay, we thought, “while people are converting, they might not be sure of what happens after they click on the CTA button.”
And that’s a problem. You always want your lead generation/conversion optimized pages and popups to make it extremely clear what a visitor can expect if they click any given button. No surprises.
A/B Testing Hypothesis Step 2: The Proposed Solution –
In order to better communicate what happens when a visitor converts on our blog overlay, we should give them a three-step process, ending with a desirable outcome:
We would then number these steps to make the expected process super simple:
1. Pick your Plan
2. Sign in with Google
3. Generate Leads
A/B Testing Hypothesis Step 3: The Complete Statement) –
Letting visitors know what to expect when they convert on our blog overlay will ameliorate trepidation and increase the number of people who sign up for a free trial.
To learn how to create your own A/B testing hypotheses, check out Kevin’s article The 3 Step Formula for Creating an A/B Testing Hypothesis.
Running the Test
Firstly we added the design change:
* A blue background
* “Get started” CTA copy
* “Not today” button option
All of these design changes had been proven to have positive conversion influences in previous A/B tests.
Then we added the three-step “what to expect” section. Here’s the final product
There are many testing tools out there. I highly recommend you employ one as you’ll get your money’s worth in the first couple tests. A/B testing is just that impactful.
Here’s a screenshot of our test results, showing percentage of blog overlay viewers who clicked “Get Started” and went to our pricing page:
There’s no better sight for a growth hacker than an orange line well above a blue one. The final test results were a 82% improvement at a 99% statistical significance.
How You Can Implement The Same Strategy
Whenever you’re asking people to convert, be sure they’re prepared for what’s to come…
- Be sure your links are labelled clearly so people know where they’re going when they click them
- Be sure your pages look traditional. If you have a pricing page, don’t put your prices in paragraph format. People are more likely to engage if they feel sure of their surroundings.
- If you have a complicated signup or sales process, test adding a simple “step-by-step” walkthrough in your landing page.
- If you are asking for blog subscribers, be sure people know the frequency and subject of your newsletters. Otherwise they’ll unsubscribe faster than Usain Bolt with a jetpack.
Wrapping it Up
If our blog traffic sees about 200,000 viewers a month, this overlay (converting at 10%) will drive about 20,000 people to the pricing page.
A/B testing/monetization is the third pillar on which an inbound marketing strategy stands.
Follow up Note on the Test we Ran:
This test worked not only because it was a good idea, but because the “proof-of-concept” original was un-optimized and it was predicated on previous tests we’ve run. It’s essential you keep detailed spreadsheets of the tests you run and what their results were, where and why. This informs every test you run thereafter, and allows you to run higher-impact/multivariate tests with more reliability.
As I mentioned above, if you have any questions about when to run an A/B test and when to swing for the fences (and why it can work), don’t hesitate to reach out in the comment section below.
- The 3 Step Formula for Creating an A/B Testing Hypothesis
- 7 Reasons Why Your A/B Split Tests Aren’t Working
- 50 A/B Split Test Conversion Optimization Case Studies
- 10 Easy A/B Tests to Help Increase Conversions
- The Art of A/B Testing – 9 Tests You’ve Never Tried
To create your own blog overlay, check out Wishpond’s overlay creator. 14-day Free Trial available.