Wednesday, June 1, 2016

How to Create a Winning A/B Testing Strategy

Sometimes, despite your best efforts, your A/B testing just doesn’t seem to work. You run test after test and nothing seems to improve your conversion rate in a meaningful way.

Or, worse, your new pages actually decrease your conversion rate.

In your search for A/B testing success, you’ve read countless case studies and incorporated their findings into your variants…and still, nothing.

Maybe you’ve had a few token successes, but those either didn’t reach statistical significance or simply didn’t improve your conversion rate enough to make a real difference to your company.

What are you doing wrong?

The good news is, you’re not the only one with this problem. If you’ve tried A/B testing, you’ve almost certainly had that moment when you look at your results and think, why isn’t this working?

It’s a frustrating situation, but the fact of the matter is that the majority of A/B tests don’t improve conversion rates. If you’re looking for a big win with every test, you’re guaranteed to be disappointed.

Instead, the key to successful A/B testing is strategy.

Strategize Your Way to Success

Recently, we improved a client’s conversion rate by 22%.

Seeing results like that, your first impulse might be to say, “Wow, that’s an awesome case study! What did you do to get those kinds of results?”

The answer to that question is actually fairly simple: we got rid of the testimonials.

landing-page-ab-test

Exciting, right? The moral of the story is, testimonials kill conversion rates!

Except…that’s not really the moral of the story.

In fact, testimonials often improve conversion rates. To be honest, when we started working with this client, I never would have guessed that their testimonials were hurting their conversion rate.

So, if we thought this client’s testimonials were helping their conversion rate, why did we decide to test a page without testimonials?

Well, we didn’t pull this idea out of thin air. As you can see below, we tested a lot of other hypotheses first.

landing-page-ab-testing-450x600

Although not every test improved the client’s conversion rate, every test taught us something about our audience and what they were looking for in a landing page.

Building an Effective Testing Strategy

The secret to successful A/B testing is creating a great testing strategy—one that teaches you something with every test. This takes planning and documentation, but it saves time and significantly increases the effectiveness of your tests.

Essentially, an effective A/B testing strategy has 4 parts:

1. Buyer Persona

Before you even build your landing page, you should create a detailed buyer persona.

Basically, your buyer persona is your testing blueprint. It gives you the information you need to come up with testing hypotheses and a framework for understanding your results.

i-know-all-about-you

At a minimum, you should know the following about your audience:

  • How old are they?
  • Are they mostly male or female? Or evenly split across both?
  • What are their responsibilities?
  • What is their budget?
  • What are their goals?
  • Why are they interested in your product or service?
  • What is their pain point?
  • How does your product or offer fix their problem?

To get at this information, it’s often a good idea to talk to your sales team or even interview some of your current customers. It’s a bit of extra upfront effort, but it can cut a lot of time off of your testing learning curve.

In the case of this client, we knew a lot about their audience. In general, we were targeting middle-aged men and women with money to invest. They knew they needed to be smart with their money, but they also knew they needed someone to help them through the process.

2. Define Your Goals

To test effectively, you need to understand what you’re really trying to accomplish with your test.

After all, if you don’t understand what you are trying to achieve with your test, it will be very difficult to create variants that produce meaningful results.

Consider the following:

  • What is your overall goal? (if you answered, “To increase sales,” give yourself a gold star).
  • What does your ideal customer look like?
  • How does a visitor become a customer? (what steps are involved?)
  • How will your test help you to achieve your overall goal? (increase form submissions? purchases? email signups? phone calls?).
  • Are you trying to increase conversion volume or conversion quality?

If you want your page to be successful, you need to understand what your company needs to succeed and be able to track the metrics that matter most. Otherwise, you’ll never know if your test was a success!

For this particular client, the overall goal was—of course—sales, but we were testing this page to try and drive more qualified leads for their sales team. Since the client had a lot of different options, our page also needed to identify which option the leads were interested in.

3. Create Your Hypotheses

Based on your understanding of your audience (ie, your buyer persona) and your business objectives, it’s time to start coming up with testing hypotheses.

The primary goal here is to identify the points of friction between what your buyer persona wants and what your business wants and come up with ways to reduce that friction.

this-is-good

Here are some potential reasons why your audience might not be converting:

  • The offer is wrong. Your audience is looking for something specific and you aren’t giving it to them.
  • The next step is unclear. Your call-to-action may be hard to find, ambiguous or uninteresting.
  • You have a copy-offer mismatch. If your audience is looking for a simple solution and ends up on a complex page, that will create friction and confusion. On the other hand, if your offer is complicated and you don’t address an important point, that can create unnecessary doubt.
  • You aren’t evoking the right emotion. A funeral services page filled with images of laughing people at a party probably doesn’t match the audience’s emotion. If you don’t stir the right emotion in your audience, they won’t feel like you are a good fit for their need.
  • Your content is confusing. Poorly written, hard-to-read content or awkwardly designed pages can make your page too difficult to bother with.
  • Your page doesn’t seem trustworthy. People are very wary of marketing, so over the top claims, poor design, lack of social signals or trust seals can seriously damage your conversion rate.
  • You have the wrong traffic. Even the most effective site won’t convert uninterested traffic. This isn’t exactly a CRO problem, but improving your traffic quality will dramatically improve the effectiveness of your A/B tests.

Understanding your potential friction points can help you set up an effective testing strategy. By testing variations that address your friction points in different ways, you can figure out exactly what works for your audience.

4. Document and Learn

The last step is to actually run your tests, document your results and then use what you learned to create new hypotheses and tests.

Your documentation can be as simple or as complex as you want, but it is critical to document what you were testing and what your results taught you.

Here is a simple example of how you might document a CTA test series:

cta-test-documentation

If you don’t document and learn, you might as well not use A/B testing. However, if you are documenting your results, you can greatly improve your conversion rate.

For example, while working with this client, we progressively discovered that our audience responded best when they felt like they were in the driver’s seat. The less we sold ourselves, the more effective our page was.

Ultimately, this led us to use a simple CTA that invited our audience to identify what they were looking for before we asked for any personal information.

In addition, we discovered that our users wanted a simple and straightforward experience. So, we spoke to that in our headline and body copy.

But, we didn’t see our best results until we applied all that knowledge to our overall page design.

In an effort to create as focused of an experience as possible, we got rid of the testimonials and created a streamlined landing page that focused on how we could help our potential clients succeed.

landing-page-1b-test-2

As a result, the new page outperformed every previous variant. But, at this point, that was hardly a surprise.

We knew our audience and we knew what they were looking for in a page. And, as we discovered increasingly better ways to meet their needs…our conversion rate improved.

Conclusion

Most successful landing page A/B tests aren’t lucky one offs—they are the result of a great, methodical testing strategy.

It might take a few tests to yield the kind of results you’re looking for, but with the right testing strategy in place, you’ll get there eventually.

You’ve heard my two cents, now I want to hear yours. What sort of results have you seen with your testing efforts. Has a strategic approach helped you improve your conversion rate?

About the Author: Jacob Baadsgaard is the CEO and fearless leader of Disruptive Advertising, an online marketing agency dedicated to using PPC advertising and website optimization to drive sales. His face is as big as his heart and he loves to help businesses achieve their online potential. Connect with him on LinkedIn or Twitter.

from The Kissmetrics Marketing Blog http://ift.tt/1sMjY3G




from WordPress http://ift.tt/1VwQsKK

No comments:

Post a Comment