I don’t know about you, but I don’t build websites for the fun of it.
Sure, I enjoy the process of watching a new site come to life, but I’m not doing it out of the goodness of my heart. When I build a new website, I want results.
Because of this, I’m a big fan of conversion rate optimization (CRO), defined by Sean Ellis of Qualaroo as:
“The method of using analytics and user feedback to improve the performance of your website; or, finding why visitors aren’t converting and fixing it.”
CRO is a virtually unlimited toolbox of tips and tricks webmasters can use to improve the performance of their sites. Unfortunately, though, the wide-reaching scope of these practices can lead to challenges in their implementation.
Are you making any of the following mistakes? If so, correct them immediately to get the most out of your testing campaigns.
Your Site Isn’t Ready for CRO
Industry wisdom suggests that every website can benefit from conversion rate optimization, but in my experience, that’s not the case.
If you haven’t yet demonstrated true product-market fit, your attention is far better spent on improving your product and your offer before implementing CRO principles. Here’s why…
An analysis by online advertising provider WordStream demonstrates the median conversion rate of their existing clients’ landing pages is 2.35% (indeed, conventional wisdom puts a “good” conversion rate at between 2-5%).
Suppose that, using CRO strategies, you’re able to increase your landing page conversion rate from 2% to 4%. That’s a big difference – but it still doesn’t land you among the ranks of the top 10% of performers illustrated above who had achieved conversion rates of 11.45% and higher.
When Wordstream took a closer look at the “unicorns” in this top bracket, they made the following observation: “Across all of the high-performing landing pages, we saw massively creative and differentiated offers.”
Ultimately, it’s your product and offer that will have the biggest impact on your conversion rates. Focus on these elements first, before committing time, energy, and money to a CRO campaign.
Your Ads Don’t Match Your Experience
But there’s another way a missing product-market fit holds back your conversion rate potential.
You think your product is the best thing since sliced bread, so your landing page pumps it up to Kanye levels of self-congratulations. Your readers click-through, but the content and experience they find on your website doesn’t match the lofty promises of your ads.
They leave, you lose. The end. No number of button tweaks or color changes is going to make up for this fundamental disconnect.
In the same way that your product and offer need your attention before you implement CRO techniques, your ads and commercial messaging need to be on the same page as what you can actually deliver, both in terms of your product itself and the experience you provide in delivering it.
Don’t expect CRO to make up for a mediocre product and some overblown ads. Deliver true value and consistent messaging before you begin any kind of testing process.
You’re Testing Minor Details, Not Big Picture Processes
Oh yeah – those button tweaks I mentioned earlier? Let’s take a closer look at another mistake most CRO campaigns make at one point or another: expecting small changes to carry a site’s performance.
Neil Patel offers the following insight in his Definitive Guide to Conversion Optimization:
“In many cases, it’s best to test radical changes to see how conversions are impacted. Then, once you’ve come up with a variation that significantly increases conversions, you can continue tweaking to inch conversions up even higher. But if you only tweak your site, you’ll never dramatically increase conversions which means you’ll never find a new baseline you can work from before testing additional tweaks.”
Once you’re ready to begin conversion rate optimization, think big. Only after you’ve addressed major structural changes should you look to small tweaks like button text, color, and placement.
You Don’t Wait Until Reaching Statistical Significance
You start a new test, and a quick winner pulls ahead. You’re anxious to apply the changes to all your variations, so you stop the test after seeing a 65% increase in conversions on the winning page.
But hang on a second! If you haven’t proven statistical significance in your results, there’s a chance your data could be influenced by unaccounted variables, such as the day of the week or the time of day when your individual variations are served up.
Shuki Mann, writing for Pagewiz, offers up this helpful guideline:
“Never stop testing before seven days have passed (assuming you had enough traffic on those days), and before reaching 95% confirmation that the data you received is statistically meaningful.”
The only exception to this rule is to kill the test early if it’s clear that your test is resulting in a significant loss of revenue. Steven Macdonald of SuperOffice CRM and VWO shared two such examples in a guest post for Moz in which his split tests lost the company $3,000-4,000 and $8,000-$10,000 (though, despite these results, he still cautions against ending tests to early).
You Don’t Have Enough Traffic to Run Split Tests
Now, beyond the rules regarding test duration and statistical significance I mentioned above, there’s one more industry-accepted guideline I want to share with you: never stop testing until each of your variations has received 100 conversions.
You see the problem with that, right?
What if your site only averages 100 conversions total in an entire year? Are you supposed to sit around and let your tests run that long? Maybe.
If, on the other hand, you want to see results faster, consider the following limited-traffic testing suggestions from Alex Johnson of Think_Traffic:
- Make big changes to produce more dramatic changes between page versions, requiring fewer page views before a winner can be determined.
- Test goals that have better overall conversion rates like bounce rate, click-through rate, or average time on site (rather than sales or leads generated).
- Run fewer variations so that your traffic is divided into fewer segments.
- Run tests on your highest traffic pages (like your home page), rather than those that get less traffic (like your checkout page).
Your Split Tests are Hurting Your Site’s SEO
Conversion rate optimization is great, but the last thing you want to do is to implement split tests that have a negative impact on your site’s SEO standing.
Unfortunately, as Joshua Uebergang of OnlineVisions points out, there are five specific mistakes you’ll need to watch out for:
- Showing different versions of your pages to Google than to users
- Letting pages solely made for testing purposes be indexed
- Not adding the appropriate rel=”canonical” tags to your tested pages
- Blindly deleting test assets once a winner has been chosen
- Letting your tests run for too long
You can find more insight into each of these recommendations in his Moz article, “Are You Hurting SEO with These 5 Split Testing Mistakes?” Apply his wisdom today to prevent the loss of the organic SEO benefits you’ve accumulated.
You Don’t Take External Factors into Account
Finally, imagine that you decide to redesign your website, at which point conversion rates fall dramatically.
Is the problem with your new design? Maybe, but it’s also possible you’re suffering from the primacy effect. According to Unbounce, “The Primacy Effect occurs when you make major layout changes that require experienced visitors to “re-learn” how to use your site.”
Without taking this or other external factors into account, you could draw incorrect conclusions about your site’s performance.
To see this effect in practice, imagine how your conversion rates would be affected by all of the following scenarios:
- Your company scores a positive press release in a major trade publication, resulting in an influx of new website visitors.
- New technology enters the market that offers a substantial improvement upon the benefits of your product.
- A major disaster strikes the geographic area where the bulk of your target audience is based.
- Your sales cycle lasts roughly six months, but your split test only runs for two weeks.
In all instances, consider your CRO results in context. If a result seems off, determine whether unrelated influences could be affecting your performance apart from your testing protocols.
Certainly, these aren’t all the conversion rate optimization mistakes you could be making, but watching out for them could make a major difference in your site’s results.
Have another mistake you think should be added to this list? Leave a comment below sharing the mistake, as well as your solution for overcoming it:
Subscribe to SEJ
Get our daily newsletter from SEJ's Founder Loren Baker about the latest news in the industry!