Kissmetrics Blog

A blog about analytics, marketing and testing

Built to optimize growth. Track, analyze and engage to get more customers.

What to Do When Conversion Optimization Goes Bad

Conversion rate optimization isn’t always all kittens and rainbows. Sometimes you test things that you’re sure will send your conversions through the roof, but it ends up going over like a lead balloon. Sales plummet, sign ups slow to a trickle…

And you freeze.

The most important step you can take is to roll your site back to its pre-test glory. And you might be inclined to just keep it there because testing the wrong things (again) could send your traffic into a tailspin.

But before you swear off testing ever again, consider the following tips. Not only will you be able to recover more quickly, but you’ll also be able to create a testing and optimization plan that helps you pinpoint where your target audience is slipping through the cracks – and get them back.

Welcome to Testing.

You aren’t the first person who has gone through this – and you certainly won’t be the last. It’s completely natural to go into this with high expectations, but what you’re seeing is possibly a more down-to-earth result. Don’t look at this as if your idea is worthless or your site is ruined – this is why we test. Once you understand that this is a positive step forward, you can start thinking like a real conversion optimization scientist – crunching numbers and trying different changes to see what resonates with your unique audience.

Real data and hard numbers are preferable to gut feelings and instincts any day – especially when it comes to maximizing your sales and subscribers. So let’s get started.

Step One: Check Your Setup and Stats

check website and testing setup

Once you’ve gotten your site back from the brink of its conversion rate free-fall, it’s time to take a closer look at the tests themselves. Did you set up your analytics correctly so that it’s tracking the right pages? Are you accounting for possible seasonal differences?  Have you tested your own sales funnel to make sure people are going to the right pages throughout the process?

Simple things like redirection errors or running concurrent promotions that conflict with each other (such as promoting free shipping on the home page and 30% off on your landing pages) could be all that’s causing your conversion rate to sink. Get back up, dust yourself off and…

Step Two: Try Again

If you’ve checked and everything is humming along in your code and in your sales process, the next step is to conjure up the courage to test again – except this time, don’t test everything at once. Changes to one area of your site might improve conversions there (such as getting more signups to your newsletter), but they might also cause conversions in other (higher priority) areas of your site to drop. Split test each change – even if it seems minor. Small things can make a big difference.  Of course, in the process of all these little changes, you may have forgotten exactly why you’re testing in the first place, which leads us to…

Step Three:  What is Your Ultimate Goal?

what is your ultimate website goal?

Sometimes, losing sight of your ultimate goal can cause conversions to fall. For example, if your end goal is more daily sales, adding in other calls-to-action (i.e. “Learn More”, “Subscribe”, “Watch Video”, etc.) can cause you to lose sight of that end result. When this happens, check your stats and sales accordingly. Did you see an improvement in subscribers but a drop-off in sales?  Conversion testing is always a delicate balancing game, and concentrating on a temporary increase in one place can add to a loss in another.

Step Four:  Keep Testing

keep testing

Don’t let one setback literally set you back and prevent you from future tests.  So the results weren’t what you expected? Learn from it and go on. Sometimes, our conversion rates drop not because of something we’ve done specifically, but from a combination of outside factors. Economic worries, the rise in costs of everyday necessities (I’m looking at you, gas prices), and general uncertainty can shift the consumer buying trends the opposite direction.

In other cases, your conversion rates could be dropping because of what your competitors are doing. If you’re selling a product and a competitor starts hawking it at half price, you can bet your own respective sales will plummet. That’s why it’s always important to keep one eye on your own stats, the others on what your competition is doing.

Have Your Conversions Hit Rock Bottom and Started to Dig?

Share your conversion troubles below in the comments, or share what you’ve done to lift that all-important rate back up!

About the Author: Sherice Jacob helps blog owners improve website design and increase conversion rates with custom blog theme design and SEO copywriting through her company at iElectrify.  Get more from @sherice on Twitter

  1. Nicholas Black Aug 02, 2011 at 6:31 pm

    Brilliant article, too many times people mix up what they think should be going on with the real deal in terms of what the stats are really saying.

    I like keeping another set of metrics as a sanity check against what one platform is telling. Over reliance can be really dangerous if there is a technical glitch.

  2. Really liked your article. Your comment on recognizing potential errors like “Simple things like redirection errors or running concurrent promotions that conflict with each other…could be all that’s causing your conversion rate to sink,” was a good reminder to me to think things through more carefully when planning promotions and tracking its success.

  3. Great suggestions. Another thing that works is to do some qualitative testing where you observe people trying to use your product. Often you’ll see that your problem is really something simple like folks not finding a call to action or misunderstanding your messaging. Fast qualitative testing can get you to the “why” of your metrics much faster than randomly experimenting with changing things one at a time.

  4. Forgott some of them…
    -wrong testing idea
    -didnt figure out what users think..which is the most common misstake.
    -build on knowledge from previous testing…which many people dont
    -understand how much you can optimize..after that it is pointless

    I have written some things about this on my blog

  5. Great tips. I find conversion elements take a lot of trial and error. Try different things and if they don’t work, do a bit of tweaking – in the long run it can only benefit your site when you start to see how visitors are behaving once they hit your site :)

  6. The image Try and Try again is definitely the story of my life at the moment. I have performed so many tests it is unbelievable, for one minute I think I optimising perfectly and then all of a sudden we get the conversion rates back and they are lower than they were before. Hopefully these recent tests will prove positive.


Please use your real name and a corresponding social media profile when commenting. Otherwise, your comment may be deleted.

← Previous ArticleNext Article →