Read these 18 split testing Tips tips to make your life smarter, better, faster and wiser. Each tip is approved by our Editors and created by expert writers so great we call them Gurus. LifeTips is the place to go when you need to know about Shopping Cart tips and hundreds of other topics.
It's actually easier to perform split testing with email marketing campaigns, because you can often do it without special software. You can divide your mailing lists into as many different groups as you want to test. Send a different email -- best to make the really, even radically different. You can even set them up to filter any replies or return mail into different folders. Then check the results in each folder. It's not all that sophisticated at this base level, but it can work. As you use more sophisticated email marketing software, some include an A/B split testing button, which automatically segments mailing lists. But you can do it on your own easily enough.
One of the best ways to evaluate new software is to find out who else is using it and how they like it. In the case of split testing software leader Offermatica, some large and successful online retailers are listed on the company site as successful case studies: Timberland, Columbia House, Zale's and Petco. So split testing software can work for large and successful online retailers. How about for smaller sites and start-ups? It's easy to find split testing software makers with successful case studies from smaller, lesser-known merchants on their sites. But the real test is to download trial software and see for yourself how it might work for you and your business.
About split testing: Split testing is a fancy way to explain live online market testing. By creating multiple versions of your sales page and sending customers to pages set in a rotation schedule, you can seamlessly split test your customer base on your e-store. It is important to remember that this is very scientific, so each split test page should only modify a small item, whether it is a graphic call-to-action or an ordering link. Examples of what you can offer to increase sales range from mentioning benefits of the product to free samples. How you use the split testing script from a marketing standpoint is your decision. However, split testing arms you with knowledge of what works and doesn't.
The cost of split testing appears to be somewhat negotiable, as most split testing software providers do not list prices on their web sites. But split testing is available in both locally and remotely hosted offerings. Expect to pay at least a few hundred dollars for an out-of-the-box software solution. Hosted split testing solutions can cost anywhere from under $30 to hundreds of dollars per month, depending on the software. If you're thinking of purchasing split testing software, especially software hosted online, make sure you understand exactly what is being offered, since services vary widely. Tracking A/B pages is going to be a lot cheaper than services that can do more in-depth tracking and reporting.
Bringing web site traffic in is only part of the equation when it comes to online business. Converting those leads into paying customers is what it's all about from your angle. One reason you may not be converting as high as you'd like is that your web site may not be intuitive for the purchaser. The solution: consider split testing software. When you run a split test, the user experience is changed slightly for a percentage of your traffic. As you guide the customer through your e-shop, some will take the existing path, and others will go through the new split test path. When you check your web stats, you can see how the modifications worked out to cold cash conversions, and continue tweaking from there. All of the pathing takes place automatically if you have the right split testing software, so choose wisely.
The dilemma is this: Whether you want to make a better customer experience get a better click through rate, you definitely do not want to disrupt a user experience. Try split testing on your webstore. Split testing allows users to follow different paths you want to test. Say, for example, you want to try a new navigation scheme to get better site click through. You can split test your traffic down two paths, the old navigation and the new one, at a 50/50 random split. Any idea is worth running a split test on. It costs very little in time and effort and can save you a lot of mistakes.
You can find downloadable and out-of-the-box split testing software in common programming languages like ASP and PHP. The best part of this is that you are going to pay once for the software up front and have no ongoing payments. Also, boxed split testing software is more easily customizable. However, like any locally hosted software solution, split testing software will require someone experienced with web programming to set up and manage, as least to start.
One multivariable web user testing company is making the claim that split testing is already obsolete. Optimost, a 5-year-old company based in New York, claims it has technology that allows real-time comparison testing of limitless numbers of pages and variables with the overall idea of increasing customer response rates. Does it work like the company says it does?
As a careful consumer, make sure you always check the demo's and pricing of any company you hire for your shopping cart needs, including split testing. While new technologies and innovations are constant and appealing, it's a good idea to review the needs of your business and your budget before signing on a future trend.
There aren't quite as many split testing providers as there are, say, ecommerce or merchant account providers. But there's no shortage of companies offering professional split-testing services, either. Some companies manage split testing packages at the corporate level - meaning it's good, but probably too expensive for most smaller ebusinesses. Other companies get good reviews with highly detailed software for smaller businesses. Both of these are hosted plans.
Say you want to try out some customer deals on your online store but you want to see if, for example, your customers respond better to one proposed marketing deal rather than another. Split testing, also known as A/B Testing, is software that allows a web site owner to send site customers randomly to two different test pages and then test and record the response rates against each other. The customers don't know the difference as they shop. They all click on the same link, but the split testing software sends them to different, pre-chosen pages and allows the site owner to monitor specific behaviors on those pages.
You can find split testing software as part of an integrated ecommerce hosting package, but it's not all that common. But be careful when you do find an ecommerce shopping cart host that advertises split testing services included. Often, this service is only included as part of ad tracking software and does not include the type of detailed A/B testing that can be done by commercial split testing providers. Still, if it comes as part of an overall plan, it's nice to have.
Shopping for split testing is easy. Google "split testing" and you'll get your choice of dozens of providers, most of which offer free trial downloads or trial periods. You can pick and choose and see which ones are best for you. You should know by now if you prefer a locally hosted or remotely hosted option and how you want to use it with your site. If you're just starting out, look to see if split testing is included in your overall ecommerce package. Finally, finding accurate pricing information on split testing isn't easy for casual shoppers. But this means prices are likely negotiable, so don't be afraid to ask for a lower price once you find split-testing software that otherwise works for you.
While split testing advocates suggest you try major, even radical changes while split testing, most of the time your results are going to be incremental. Once in a while, users may react strongly to a change in navigation, or a new email call to action. But most of the time, the most successful changes that come after split testing happen in small but steady increments.
With good split testing software, you can do all kinds of testing of how visitors to your site actually respond to everything from site design, to calls to action, to email marketing. No matter what you do with your split testing software, here are some of the basic procedures you need to follow:
* Get good tools.
* Determine your control page.
* Establish your testing goals and parameters.
* Determine your testing intervals.
* Come up with some radical redesigns.
* Evaluate the redesigns with split testing.
* Determine your new control page.
* Optimize your site according to your testing results.
Why does anyone have to have specific split testing software? Isn't it possible to do it yourself? Good questions. The answer: Yes and no. You can certainly do A|B testing yourself between pages or even with different content items on pages. If you have real-time reporting, you can test pages under specific time periods and look at your web tracking software. If not, you can test in 24-hour periods. However, if you want real time testing, or if you want detailed reporting of that testing without having to go back to your server logs, you really need split testing software.
Getting good split testing software is critical to the success of your split testing efforts. But unless you understand how to use that software, and have some patience while you come up to speed with it, it's not going to deliver the results you want. Get yourself a good instruction manual, check out web sites, ask others who have done split testing. Make sure you have a solid testing procedure down and some idea of what you are going to do with the results.
|Sheri Ann Richerson|