Back in 2006, Google launched their “Website Optimiser” testing platform which made A/B and multivariate testing available to the masses.Prior to this, website testing platforms were very expensive and subsequently exclusive to companies with big budgets.
At the time, I was lucky enough to work for one of the first UK agencies to be awarded “Website Optimiser Certified Partner” status by Google and have since been involved in the planning and deployment of many tests.
Below you’ll find what I consider to be the three most important lessons I have learnt from these experiences:-
1) “Conversion Rate Optimisation” Is More Than Just Page Design
Conversion Rate Optimisation is often referred to as the process of increasing a web pages conversion rate by testing different versions of a page or its individual elements.
..And if you’re just testing web pages designed to generate leads the above definition is relevant.
But, if you’re working on eCommerce sites, the process becomes far more complicated.
In fact, based on what I’ve seen you can completely redesign a page on an eCommerce site, drastically changing each and every element and still fail to significantly influence the conversion rate of that page if you’ve neglected other pivotal commercial factors.
These factors typically include:-
- Pricing– Never underestimate the importance of getting your pricing right, in fact if you’re running an eCommerce site you should be checking your pricing as often as possible be it manually or via an automated system. If you’re struggling to compete on price, focus on finding out why. I’ve worked with clients who’ve run dozens of tests, attempting to improve sales conversions rates only to fail miserably because they’ve neglected to check their pricing on a regular basis. Put simply, if your pricing is wrong nothing you test will significantly improve your conversion rates.
- Product Descriptions/Images– Unlike bricks and mortar stores, visitors to eCommerce sites can’t physically interact with products. As such, it’s imperative that you take the time to write thorough, detailed descriptions and combined them with good quality photography.
- Stock Levels– The web is a very fickle sales environment, after all, your competitors are only a click away. Therefore, if your stock management is poor and popular products are frequently unavailable your conversion rates will suffer.Nine times out of ten, if you don’t have it in stock, someone else will.
- Bundles, offers, sales, promos – I can’t stress strongly enough the importance of updating your site with new and genuine offers. This gives people a reason to return to your site and helps combat the current bargain hunter mindset that’s been amplified by recent financial pressures.
2) “Conversion Rate Optimisation” Can’t Fix A Broken Business
Due to the very nature of Conversion Rate Optimisation I’ve found that it attracts a lot of struggling businesses that are afflicted with more serious issues than their page designs.
For this reason, I’m quick to stress that whilst Conversion Rate Optimisation is a powerful way to improve performance it’s important to understand that this improvement is typically incremental to a business that is already working.
A couple of things Conversion Rate Optimisation can’t fix:-
- A Flawed Business Model – When broken down to its core components, a business is a series of equations involving costs, sales, profits and margins. A business that is struggling because it has failed to get these equations right cannot be fixed by Conversion Rate Optimisation alone.
- A Poor Product/Service – You can dress it up, put makeup on it and put it in a frock but if the product or service you are offering is badly designed or poorly conceived Conversion Rate Optimisation isn’t likely to help.
3) Don’t Let Testing Slow You Down
Only websites with vast amounts of traffic and lots of conversions can run complex tests and generate answers quickly.
..the remaining majority have to restrict the number of elements and variations they test at once to avoid long delays in determining a conclusive winner. Test that take too long to reach a conclusion can potentially restrict the agility of your business and do more harm than good.
To combat this, consider viewing most of your tests as what I refer to as “check tests”. Or in other words, running a test until you’re confident that your changes won’t harm existing performance and then move onto your next test.
I’m sure the purist amongst you will be groaning right now and stating that tests should always run until a statistically significant conclusion has been reached. But, the reality is most business simply can’t wait that long.
None of the above is intended to sound negative. I’m a completely committed to A/B and multivariate testing and have personally witnessed the positive results it can produce.
But remember.. 1) don’t neglect other areas, 2) be realistic about who it can help and 3) don’t let it slow your business down. Keep those three things in mind and you’re off to a good start.