Measure Your Customer Lifecycle
For example, a consumer-focused e-commerce vendor recently approached I/PRO to help analyze behavioral data from its Web site and recommend ways to lift its look-to-book ratio (the percentage of visits that end in a purchase). After all, the company was spending a lot of money on getting people to its site and on analyzing what banners worked on which Web sites. It felt more could be done to drive people to purchase once they were at the site. I asked a series of questions about its customers:
* How many people come to the site once and never return?
* How many people place items in the shopping basket and then abandon the site?
* Why do people come to the site?
* What percentage of people who arrive intending to purchase don't?
Most of the responses were about its site:
* We are concerned about the level of leakage at the site.
* When the site is slow, people abandon a partially filled shopping baskets.
* Our site attracts active buyers.
* Our initial path analysis research is inconclusive as to whether our site layout is good or bad.
I was struck by the contrast in our mindsets. People use the Web for more than buying. They window shop, they seek product information, they hunt for bargains. The fact that 100 percent of visits don't end in purchases isn't bad. Even if there is no purchase, if the visit ends with a person who is one step closer to becoming a long-term customer, it's a success.
Conversely, even if there is a purchase, if the visit ends with a disgruntled buyer, it is a lost opportunity.
Online advertising measurement has evolved to a point where marketers recognize it is not just about the click-throughs, but also about the brand impact of banners. Web site measurement is beginning to take a similar step. It not just about this visit, but about long-term relationships with customers.
Measuring whether you are successfully developing a strong and loyal relationship with an online customer is not easy. It has much more to do with meeting a customer's needs and expectations, than completing a specific transaction. To date, most Web measurement has been stateless. For example, it has concentrated on measuring "gateway events." Measuring gateway events can be misleading -- and even dangerous -- if your goal is to build customer loyalty.
Currently, the most successful way we have found to measure whether a single visit contributed to a positive customer relationship in a way that provides actionable information has been through a combination of attitudinal (survey) research and behavioral (click-stream) research.
When someone arrives at the site, intercept him and ask a simple question to find out why he is visiting the site. Ask for his e-mail address and ask permission to conduct some follow-up research. While people are at the site, track their clicks. Similarly to a physical store, some people will go straight for what they want, others will browse. When reconstructing their visit for analysis purposes, a good way to identify whether someone is a browser or a driven shopper is to look at their time per page and compare it to the average for the site. If possible, compare to people connecting via a similar speed.
Soon after they leave, send a follow-up survey. Ask them to rank their experience. Do they feel they accomplished what they set out to achieve? What is their overall perception of the site? How would they compare it to other e-commerce experiences?
While analyzing the click-stream look to see if you think they met their goal. If they wanted product information about a palm pilot, did they find it? By cross-tabulating the goal of the visit, the feedback from the survey and the behavioral information, you will get a strong sense of whether this interaction with the customer lived up to expectations.
Then wait for them to return. For the next several visits, observe their behavior and see whether they meet their stated goal in any of those visits. Many people will make several visits before purchasing. If you have a content site, many people need to learn
where to find content before they become a regular user.
If you want to develop profitable customer relationships there are three things to remember:
* Different people are at different stages in the customer lifecycle; do not set a single goal for all visits.
* Ensure your site is easy to use for people at all stages of the lifecycle.
* To assess the success or failure of your Web site, measure customer interaction across a series of visits, not within a single session.
Well thought-out metrics can help you improve your Web site's performance significantly. Don't settle for less.
Tim Reed is vice president of marketing at Web traffic measurement and research firm I/Pro, Redwood City, CA. His e-mail address is firstname.lastname@example.org.