Everyone says they’re doing it but few are doing it well.
Many have danced with conversion optimization, flirted with testing, even committed to a test or two, then either given up or lost support (and friends!)
Companies who try out conversion optimization testing often fall into common traps.
I want to help you avoid those.
You may have heard some conversion advocates say that, if you’re not testing yet, you should “Just start testing! Just start with testing anything.”
Well, that’s bad advice. If you start with a poor process or mistaken assumptions, you can waste a lot of time and organizational support.
At WiderFunnel, we often hear from marketers who’ve been frustrated by their early attempts when they haven’t seen the big conversion rate lifts they think they should. Unfortunately, getting their colleagues excited about testing the second time around takes a lot more work. Sometimes, the conversation is just crushed, like a rose under the foot of a jilted lover.
Here are the 13 most common ways you can screw up your conversion optimization
- Implementing the latest “conversion optimization tips & tricks” they see on blog posts. This won’t give you the results you need. You may get some sporadic wins, but tips and tricks ignore your unique context and can’t be generalized.
How to screw your tests #1- chase after tips
- Consensus testing. Developing test plans is an area where democracy doesn’t work. You need to test bold new ideas that groupthink will kill. You need a Conversion Champion who is empowered to test counter-intuitive concepts. I encourage you to collect ideas from anywhere they can be found, within and outside the organization, but control of the test plan should be centralized with your conversion champion or outsourced conversion rate optimization agency.
How to screw your tests #2- consensus testing
- Facing organizational barriers. As a conversion optimization champion, if you haven’t gotten buy-in from all stakeholders, you can run into roadblocks in unexpected places. I’ve seen turf battles, silos and competing priorities hamstring some very promising tests. This is especially important for enterprise organizations where the champion has to navigate a complex political environment. One tip for overcoming this is to run a “Skunkworks Test” by picking a few target pages with low political visibility to gain some quick wins. In a few weeks, I will share 7 more tips for being an effective conversion champion. Make sure you’re subscribed to WiderFunnel’s blog updates to get it in your inbox.
How to screw your tests #3- hitting organizational barriers
- Relying on usability testing to make website changes. I’ve covered before, the many reasons that usability testing alone is not reliable. Usability testing can be an effective part of your marketing optimization strategy, but its role is to generate hypothesis, not to direct website changes. Small numbers of strong opinions lead to dangerous decisions.
How to screw your tests #4- Rely on usability tests to make decisions
- Testing too conservatively. Meek tests produce weak results. You must test boldly to make big improvements. It’s ok to test things that make you uncomfortable if you remember this: It’s just a test. You can always test something else. The data will teach you something.
How to screw your tests #5- testing timidly
- Prioritizing incorrectly. Starting with the wrong pages can waste tremendous effort, even if the result is positive. You can make prioritizing website tests as easy as PIE.
How to screw your tests #6- Prioritizing incorrectly
- Deciding too early. Often, it’s difficult for the conversion champion to hold off the pressure to make decisions without enough data. If you’re not waiting for the 95% statistical confidence level before making conclusions, you can hurt your business. The only thing worse than not testing is drawing bad conclusions from tests.
How to screw your tests #7- Deciding on a result too early
Also, there are no losing test results if they’re planned properly.
- Relying on so-called “best practices”. Look no further than the most common home page design best practice to see this: rotating offer sliders don’t work. At WiderFunnel, we love destroying these commonly-used elements with our cross-industry testing program.
How to screw your tests #8- Rely on so-called best-practices
- Using before & after testing. Many are still testing their websites using the old “Pre & Post” testing method. They’ll make a change to their landing page or ads and then compare their results before and after the change. There are so many outside factors influencing your conversion rate that these types of tests are meaningless. Seasonality, competitive activity, your marketing offers, stock levels, etc.; they all conspire to muddy your results. Only scientifically-controlled testing will tell you the truth.
How to screw your tests #9- Using before & after testing
- Taking advice from “experts” who don’t do a ton of testing. If their primary business isn’t testing, testing and more testing, where is their advice coming from? You might be surprised at how little testing some of the industry’s pre-eminent figures actually do. Many of the assumptions that are baked into “experts” recommendations simply don’t work. Never trust the advice of people who don’t do a lot of A/B testing.
How to screw your tests #10- Take advice from ill-informed experts
- Using tool vendors’ services to plan their tests. We’ve found that the tool vendors services groups are great at supporting technical implementation challenges or tool bugs, but tool developers don’t deliver the best testing plans. They can’t avoid the pressure to create tests to use the capabilities of the tool rather than to get the best results. No company can be an expert in both software development and marketing services. That’s why, at WiderFunnel, we decided early to focus on the thing we can be best in the world at. By remaining unbiased and technology agnostic, we can recommend the best conversion tools for our clients and deliver tests that get the best results, regardless of the tool.
How to screw your tests #11- Letting tool vendors plan your CRO
- Not bringing in fresh eyes. Staring at the same business every day has a tendency to atrophy idea-generation. It’s not just you; everybody has the same challenge. You can exponentially multiply your pool of ideas by hiring a conversion expert with experience in many businesses and industries.
How to screw your tests #12- Doing everything in-house
- Confusing quantity of tests with quality of ideas. All tests are not created equal. Running a whole bunch of tests rapidly is not hard. All of today’s testing tools enable quick testing. But, are you using the tools to test ideas that will dramatically improve your results? One good test can give better results than dozens of weak ideas. One great test will give better results plus marketing insights that will produce even greater results in the next tests. Two great tests will… well, you get the idea. Great testing is exponentially better than pure volume testing.
How to screw your tests #13- confusing quantity with quality
There’s a Gulf Between Conversion Rate Optimization and Great Conversion Rate Optimization
There are two main reasons companies settle for mediocre testing:
- They lack the right methodology (like the Kaizen Method, for instance) that identifies goals, prioritizes the work and lays out a roadmap.
- They lack the unique skillset in-house of a Conversion Strategist (or whatever you want to call it). Conversion optimizers have a unique balance of left and right brain, and disciplined decision-making, which is difficult to find if you don’t know how to train them.
Companies getting the best results have developed a structured process for conversion optimization and have built a cross-functional team of talented conversion champions.
By building your marketing optimization strategy properly, you’ll have a chance to make sweet love to your revenue and profit!
What challenges have you seen?
Can you related to these problems? How have you overcome them? What are your tips for doing Great CRO?