Defining the Truly Useful Metrics for E-TailersA good friend once told me that marketing is half art, half science. The science part largely boils down to measurement. If you can accurately gauge what works and what does not, you can make smart, informed decisions. Otherwise, you are throwing darts blindfolded.
If ever there was a marketing domain in desperate need of good measurements, the Web would be it -- especially for online storefronts. This might surprise you because, on the surface, Web sites generate a plethora of data and reports. You probably have a stack of reports in your inbox right now. The problem is that much of this information does not directly tell you what you need to know.
Why not? Most Web reports originated either from technical people ("Isn't it cool that you can see every hit on your site grouped by IP address?") or traditional retail/catalog marketers ("Give me sell-through by SKU."). While this information is useful in its own way, it fails to illuminate the real activity of people browsing your store and it does not give you accurate feedback about the incremental marketing tactics you want to use online.
Instead of a report that simply showed you the net sales for each product, imagine if it also told you the following:
• How many times the product's details were viewed (for example, someone clicked to read more about this product from a category list or search results page).
• How many times the product was added to someone's shopping cart.
• And, out of those, how many times a product sale was completed.
Though each of these numbers is significant, the real answers are revealed through their ratios to each other. For each product, what is its details-to-cart ratio? Its cart-to-checkout ratio? How do these ratios differ across your best sellers and your worst sellers? How do they compare within specific categories or price ranges?
This gives you insight into the appeal of your products beyond raw sales. Perhaps some high-ticket items have great pull -- visitors regularly click-through to read more, maybe even add one to their cart -- but have a low close rate. Is it the price or other concerns? Where are you and the customer diverging?
With the right metrics, you can experiment intelligently and track the results. If a product is very popular, but is not selling well, that may be a great opportunity for a targeted promotion. You can alter your pitch as well as measure and tweak some more until you hone in on the winning presentation.
It is worth pointing out that this is not a report organized by page hits, though that is what is happening behind the scenes. Instead, this describes Web-based customer activity at an application level, organized by product.
The next level of market feedback is achieved by recording and reporting how a customer arrives at a certain page on your site. Consider the implications of these different ways in which someone reaches a product page:
• The customer clicked through an intrasite banner advertisement.
• He picked it off a list of featured products on your home page.
• He selected it from a category listing of products through your site's navigation.
• He jumped to it from a search result list.
• He chose it as an upsell or cross-sell while looking at another product.
• He acted on a link included in an e-mail promotion.
• He reached it via some external link from another site.
Imagine that your report has a breakdown under each SKU that shows the percentage of each of these pathways to that product. This tells you a tremendous amount about your audience and the effectiveness of your site's organization and merchandising.
With just a little more data collection, you can flush it out even further. If the product was picked from a category listing -- and the product is listed in more than one category -- which category did it get picked from? Does the product sell better in one category or another? If the product was picked from search results, what was the customer searching for?
A short list of the top 10 keywords that generated results that led the customer to that product page is a marketer's dream.
Match this report with the product pull report to learn how different navigation paths on your site affect your details-to-cart and cart-to-checkout ratios.
Do yourself a favor. Forget about those cryptic Webtrends reports. Have one of your developers add in the code to collect the data described above. These two application-level reports will tell you almost everything you could want to know about the state of your store.
These reports can be implemented without infringing upon an individual user's privacy. There is no need to stalk a user's session with cookies. By aggregating data collected at strategic points and analyzing it through the right lens, you will know what is going on.