Are you really improving your site with this recommendation?

Across the Internet in forums and indeed on this site there is plenty of advice regarding how to improve your ecommerce business.

Whilst all advice is offered with the best intentions, you have to realize that it may not work for you. The bottom line for every ecommerce site should be how much profit it is making. There is no point “improving” your site if it reduces your bottom line.

To this end, before you undertake any major change to your site, measure the before and after effects of the change. Sometimes a change that everyone expects will improve your site can actually damage it.

Metrics are essentially boring. But they are so important that you must not ignore them. The key metrics are (a) how many people visited your site, (b) how many converted into paying customers, and (c) how much money did you make from them. As your site grows, and as your metric data likewise grows, you can start to collect more information, such as the best landing pages, the bounce rate of the pages, and customers’ navigation through the site. All this data can help you find the strong and weak points in your site. Analyzing this data can give you ideas of what areas you need to change and what you can leave for now.

So whatever your source of ideas, either from here or analyzing your own data, the important fact still remains: You must test the change, monitor the metrics, and make sure the change has actually improved your site.

If you believe that your site desperately needs a live chat service, or a proper one-page checkout, or a super advanced search, then test those features. Sometimes you may be surprised by the results. Sometimes what is the perceived good idea may turn out to be a lemon.

For example, I am testing the implementation of bar codes on my ecommerce site. When people are shopping, there is a growing trend to compare prices. They increasingly look up on the Internet what they see in a brick-and-mortar shop and see if it is cheaper online. They tend to do this with their smartphones, by scanning in the bar code and searching on that. I do not have enough data on this yet, but for my market I see a growing number of searches using just these codes. This has a significant impact on my organic search visitors. I am reviewing my search engine optimization and on-page optimization to see if I can improve my results for such searches.

Do I put the bar code in the page title? Do I actually show it on the page? Do I add it to the product title? Do I add it to the alt text on the image?

If I increase the search visibility of the bar code, I will naturally reduce the visibility of some other keyword. Does this matter? Is someone searching on the bar code alone much more likely to buy, as he is already in the buying frame of mind? Will I get a better conversion rate on this search term?

I am guessing that concentrating on the bar code in this way will improve my site. However, I do not know for certain. Using good metrics takes out all this uncertainty and gives answers.

The final caveat is to implement one change at a time. If you do two changes and one doubles your visitors and the other halves them, you will see no overall change. You will not realize that one change is crippling you.

So next time you change your site, make sure you know that it was a change for the good.