If you have selected CRO as your major marketing project for the year, there is no lack of articles providing you with advice on what change, the best analytics software to use, and how to set up your split testing. But, what happens if you set everything up and things don’t work as they should? Maybe your test results don’t make sense. Maybe you go live with what appears to be a clear winner, and then that winner is an absolute flop. If your CRO campaign is failing, there is probably a reason for it, and a solution that you can apply.
1. You Are Expecting Results to Come Too Quickly For The Amount of Traffic You Receive
A large online retailer that receives thousands of website hits every day can set up A/B testing and receive actionable measurements quickly. Quickly, meaning within a day, and quite possibly faster than that. If you are small business owners who only receive a couple of hundred hits on a given day, you aren’t going to get back useful test results that quickly. This means you will have to wait significantly longer to get the data you need.
2. You’ve Taken an Approach That is Too ‘By The Book’
There are plenty of articles that provide great advice on the steps you can take to optimize your conversion rates. Some of these are:
- Providing your visitors with shorter forms
- Relocating your CTA buttons
- Changing your CTA text or color
- Providing detailed contact information on your home page
- Modifying the content on your landing pages for branding and message continuity
Now, it may be that you need to apply every single one of these suggestions and more to get the conversion rates that you desire. Alternatively, you may only want to implement one or two of them.
In some cases, the standard CRO changes may not work for you at all, and you might need to think outside the box. Be sure you are adding your own insights and taking into consideration previous customer feedback before making adjustments.
3. Your Efforts Are Traffic Boosting But Not Conversion Boosting
More traffic is a great thing, but it won’t help your conversion rates unless that increased traffic is finding its way into your conversion funnel. Step away from optimization methods that drive traffic to your pages that don’t get conversions, and focus on methods that get website visitors to the pages where they can and will convert.
Remember, you are looking to make changes that directly impact your bottom line. For example, you can employ somebody with great writing skill for content optimization, but if that isn’t what you need to boost your conversions, your CRO campaign isn’t going to be successful.
4. You’re Testing For Traffic Levels and Bounce Rates When You Should Look Elsewhere
If you are making the mistake mentioned in point three, you will compound that mistake if you aren’t testing for the right things. For example, if you are simply using increased traffic and reduced bounce rates to as metrics to determine whether or not a CRO change has worked, you aren’t going to get an accurate picture. What you want to do is use metrics that indicate whether or not more people are converting.
5. You Aren’t Syncing Your Tests With Other Influencing Factors
Let’s say you operate an online business that sells a variety of gourmet foods, and that a large portion of your product offerings come in the form of gift packages. It is reasonable to assume that your sales numbers vary widely according to the season, and that the holidays are your high sales time.
If you aren’t testing your changes during these times, you are going to get skewed results. If you test when lead generation is at its highest, you might also avoid testing problems related to a lack of traffic.
6. You Are Making False Assumptions About Your Test Results
One of the difficulties about CRO is that the test results you receive may not tell the whole story. One thing that can really impact this is the use of too many variables in your testing. Consider this. If you test 15 different versions of CTA button text, and one of these variations happens to come out slightly ahead of the others in your testing, does this mean you have found the best text for going live? Maybe, but you will probably want to do further testing to insure that is truly the case.
7. You Aren’t Starting Big Enough
If your conversion rates aren’t nearly where you want them to be, chances are there are big changes that need to be made. While you may be tempted to start with a small change, then test, then make another change, and test again, the truth is these incremental aren’t going to make a measurable difference.
When it comes to CRO, your best bet is to begin with the biggest change, test that, and then tweak that change based on your results. After you have that change in place, you can start thinking about smaller changes.
8. You Aren’t Making Good Use of The Metrics Available to You
Before you set up your testing and decide which metrics will give you the best information, ask yourself what a conversion means to you. Are you looking for email list subscriptions? Do you want people to fill out their contact information for your sales staff? Do you want more followers on social media, or to generate shares? Is your goal to increase the number of people who actually purchase your products?
Depending on your goals, the numbers you use to determine the success of your changes can vary widely. For example, if you want to increase social shares and followers, you may want to look at the amount of time people are spending on your page and whether or not they are watching your videos.
If you have other goals, those metrics might not be meaningful to you.
9. You Are Making Technical Errors
If you are going to run successful split tests, you have to be sure that you set them up correctly, and that you understand how your analytics software works. If you get test results that seem to be completely off base, take a few moments to make sure that you haven’t made any errors in setting up your tests.
10. You Aren’t Running Your Tests Long Enough
If your test cycles are too short, you might be erroneously selecting your winners prematurely. Let your tests run longer and you may be surprised at how often an initially clear ‘winner’ isn’t the best choice after you wait for a hundred or more conversions to occur on each of your variations.
- 26 Jan, 2016
- Posted by Julie Ellis
- 0 Comments