Lean Analytics

Chapter 8. Model One: E-commerce

In an e-commerce company, a visitor buys something from a web-based retailer. This is perhaps the most common kind of online business, and it’s certainly the one that the majority of traditional analytics tools are aimed at. Big retailers like Amazon, Walmart.com, and Expedia are all e-commerce companies.

If the e-commerce model most closely matches your business, this chapter will show you some of the most important metrics you need to watch, as well as some “wrinkles” that make the analytics more complex.

Early e-commerce models consisted of a relatively simple “funnel”: a visitor arrived at the site, navigated through a series of pages to get to a particular item, clicked “buy,” provided some payment information, and completed a purchase. This is the traditional “conversion funnel” from which mainstream analytics packages like Omniture and Google Analytics emerged.

But modern e-commerce is seldom this simple:

• The majority of buyers find what they’re looking for through search rather than by navigating across a series of pages. Shoppers start with an external search and then bounce back and forth from sites they visit to their search results, seeking the scent of what they’re after. Once they find it, on-site navigation becomes more important. This means on-site funnels are somewhat outdated; keywords are more important.

• Retailers use recommendation engines to anticipate what else a buyer might want, basing their suggestions on past buyers and other ...