And the answer is….Run a test
- Can’t decide which headline will convert more of your visitors
- Not sure if the Add to Cart button should be orange or green
- Can’t agree on whether you should have flash images on your home page or not
Want to increase your conversion rate; get more sales, more leads or more orders? Start a testing program.
Why Test?
- To increase the number of sales, orders or leads you get from your website.
- It’s an affordable way to improve your website. Tests can be designed around any budget. Even small improvements can yield big results.
- If your customers are important to your business, give them a voice in the design and functioning of your website. Start a testing program. Let them show you what they like and what they don’t like about your website or landing pages.
Testing is a Win Win for You and Your Website’s Visitors
During a test visitors “vote” with their behavior. They register higher conversion rates on the page or combination of items on a page that is:
- Easier for them to use
- More intuitive
- Does a better job of meeting their needs
When the test is over, you replace your old page with the page with the highest conversion rate. You win because you’ve increased your conversion rate and your visitors win with an easier to use, better designed website. Your visitors voted and you listened.
Large companies like Amazon and Best Buy are continuously testing their landing pages. They want their websites to convert as many people into customers as possible, and the best way to do that is to continuously test small website changes and see how their customers react.
Selecting and Designing Pages to Test
In the Website Optimizer jargon, each test of a page is called an experiment.
- Use Google Analytics to determine which pages on your website are converting poorly.
- The key for testing is a conversion event such as a sale, a form completion, a click or other action that can be recorded and measured. The winning test page is the one that records the most conversion events.
- Choose a page that gets a lot of traffic. To get good results, the test needs to be viewed by a lot of people. It’s easier and faster to test high traffic pages.
- Find a reasonable balance between what you want to test and the amount of traffic you need to test it. There are no limits on the number of items and combinations you can test. You are only limited by your imagination and budget. However, the more items you test, the more traffic is needed to obtain useful results. .
- Be fearless—go for persuasive titles and eye catching images. Use tests to see how your page can make a bigger impact.
How to test a page
Once you have selected a page for a testing, you need to decide whether you will be testing one page against another or several different combinations on two or more different pages.
If you decide to test several combinations, first choose sections of the page that will have the most impact on website conversion. If you are experimenting with a product page, you can test things like the headline, the product description, price, and/or product image. Avoid testing extraneous material that is not directly related to the product or conversion event.
Website optimizer allows you to test different variations of each section. You can also test different combinations of sections on each page. For example: if you have three variations each of the headline, product description, product image and call to action, you can test them in three combinations:
Combination 1: Headline 1, Description 2, Image 3, Call to Action 1
Combination 2: Headline 2, Description 3, Image 1, Call to Action 2
Combination 3: Headline 3, Description 1, Image 2, Call to Action 3
Seems straightforward, but with three variations of four sections, there are 81 possible combinations. For a small business, even with reasonable traffic, you might not get meaningful results for months, or even years.
Our best advice is to keep it simple. Be bold in what you test, but conservative in the number of items tested simultaneously. This keeps the clear winners obvious, and lets you build on what you learn from each successive test to improve your website’s conversion rate.
What About the Unsuccessful Test?
Not all tests will be successful. There is no rule that you must run a test for a certain period of time, so if a test is costing your sales, stop the test immediately. Every so often the new page doesn’t convert as well as the old one. Don’t be discouraged or take it personally, this isn’t about you or your test design skills. It’s about your visitors and what they like.
Maybe you’ve done a number of tests on a particular page, and you’ve reached the point of diminishing …pick a new page to start testing.
Maybe this is the only the second test you’ve run and there isn’t a clear test winner. Maybe you weren’t bold enough in your vision for the page. Step away or a while and look at competitors pages for inspiration. Have someone outside the company look at the page and make suggestions.
Visitors are just like the rest of us—they resist change, especially when it doesn’t represent unambiguous improvement. You need to present a real reason to vote for your new page design. That is why boldness and thinking outside the box are so essential to designing your tests.
Looking at the Results
If you are testing one page against another, the winning page will be the page registering the most conversion events.
If you are testing various combinations on a page, Website Optimizer shows you results for both individual sections and combinations. The results for combinations are of five types:
- Estimated website conversion rate change
- Chance to beat original
- Chance to beat all
- Observed Improvement
- Conversions/Impressions
The “Chance to beat all” probability rate is likely the most important statistic. This tells you how likely a particular combination of sections on the page are likely to beat all other combinations, and give you the best website conversion. When a page has a high probability in this statistic, you have found a winning combination.
If you are not getting appreciable results with the “Chance to beat all” rate, then you may be using too many combinations. The more combinations you have, the longer it will take to get good results with this statistic. You can either reduce the number of combinations, or wait for Optimizer to collect more data. Try experimenting with more radical variations of your page sections. Taking a risk can sometimes have surprisingly positive effects on your website conversion.
Start Testing Now
Successful websites are continually testing and making improvements to their site. One test will make some improvement to your website, but multiple tests will have a dramatic impact on your website’s conversion rate.
Using Google’s Website Optimizer is a great way to analyze your website traffic and increase your profit margins by eliminating ineffective web pages and replacing them with pages that really sell your product or service.
Let us help you start a testing program for your website. We can show you how to pick the pages to test, set up the test, and analyze the results. Start converting more of your visitors into customers, today.
