|
|
you are here:
home >
>
>
Raise Conversions Higher
Do you want to raise conversion rates of your website?
In order to get higher conversion rates, you must follow a disciplined action plan:
- Understand your business objectives - Be clear about the desired conversion action(s) on your website and their dollar value. A conversion action can be a purchase, form-fill, download, or clickthrough to another page. Understand the economics of your business and be able to quantify the profit impact of a conversion improvement (try our free
).
- Decide exactly what to test - Understand which parts of your website are "mission critical" to achieving the desired conversion actions. Decompose your pages into individual variables to test and decide what alternatives to your current website you want to include.
- Select a tuning method - Based on the size of your test and other factors you can select a tuning method. Common methods include A-B Split Testing, Multivariate Testing (Taguchi Method), and the SiteTuners TuningEngineSM (see a
here).
- Create the alternative versions of your pages - For each of possible landing page versions in your test you will need to create new designs, graphical elements, and text copy.
- Select your audience and filters - You may not want to use all of the traffic hitting your website for testing purposes. You may want to exclude e-mail blasts to your in-house mailing list or other visitors who already have a preexisting relationship with you. Many companies test only on fresh "acquisition" traffic from online advertising such as pay-per-click (PPC).
- Install tracking software - You will need to track which version of your page a particular visitor sees, and whether they eventually "convert" by taking the desired action on your site. Many web analytics packages are available to support these kinds of tests.
- Collect the data - Aim your test traffic at your new testing landing pages until you have collected enough data to be very certain of the results. The data rate is not based on the number of unique visitors; it is based on the number of conversion actions. If your want to raise your conversion rate but have a low data rate, you may have to spend many weeks or months collecting enough data.
- Analyze the results - Make sure that your results are statistically significant and are not just due to random chance. Bigger, more obvious improvements will be easier to detect. Small improvements may not be "real" unless you collect much more additional data.
- Verify the improvement - Your test analysis may predict that a certain alternative version of your site may be the best. Unfortunately most conversion tuning methods assume that there are no
. So your predicted higher conversion rate may not hold up. You must verify any predictions in an actual head-to-head test of your best "challenger" version against your original "baseline" version.
|
|