A recent report from The Data Warehousing Institute (TDWI) estimated that poor data quality costs U.S. businesses up to $600 billion a year. TDWI found that company leaders are often unaware of the underlying issues driving the loss, and reported that inaccuracy remains a core problem because it prevents company executives from making effective decisions.
With fierce competition and small advertising budgets, retailers must be extremely cautious with online marketing expenditures. Poor data quality is the number one obstacle for marketers who are trying to make informed decisions about customer acquisition and retention marketing activities – both of which are a top priority for marketers today.
The TDWI report brought to mind a very important issue that I have been evangelizing for years – the difference between session and lifetime-based revenue analyses for measurement of online customer acquisition efforts.
Virtually all Fortune-class companies with an Internet strategy are familiar with session-based revenue analysis for online customer acquisition. The concept of lifetime-based analysis, however, is often an entirely new way of looking at revenue for these fortune-class companies.
Largely, this is because lifetime analyses are hard to perform, requiring extensive data warehousing expertise that is both costly and cumbersome to manage.
To illustrate this difference, I'll walk you through a real-world example.
Let's say that a company chose to advertise on three different sources – Yahoo!, MSN, and AOL – by displaying a single banner ad on each throughout the week of August 11th.
In a session-based analysis, a marketer would calculate whether or not a visitor made a purchase (or made another “conversion” action if the Web site is non-commerce) within the session in which the visitor clicked through on the ad.
The results may look like those below, which would lead one to conclude they should invest more heavily in Yahoo! and cut spending in MSN.
Session-based analysis: (8/11/02 – 8/17/02)
Source | Click-Thru | New Customers | Revenue | Cost | Profit |
| Yahoo! | 40,000 | 1,000 | 50,000 | 20,000 | 30,000 |
| MSN | 40,000 | 200 | 10,000 | 20,000 | -10,000 |
| AOL | 40,000 | 500 | 25,000 | 20,000 | 5,000 |
By comparison, lifetime-based revenue analysis considers both the immediate and long-term actions of the customer.
Lifetime-based analyses assume that visitors may not purchase the very first time they click through on a banner ad – in fact, they might return to the Web site several times before they make a purchase or take an identifying action.
For example, a visitor might close the browser window after the banner ad click-thru, and directly access the Web site the next week (by typing the URL in their browser instead of clicking on an ad) and complete a purchase. A session-based analysis would not correctly measure this activity – the visitor didn't purchase in the session that they clicked on the ad. But the ad clearly played a role in getting the visitor familiar with the Web site to make a future purchase.
By comparison, in a lifetime-based analysis a company would not only analyze the conversions during the week that the ad ran, but could also analyze how the visitors who responded to the banner ad behaved over their lifetime of interactions with their business online, or in this case the next three weeks (during which the ad was no longer running). The results may look like those below, which would lead to a very different conclusion – discovering that AOL is driving much more valuable long-term customer leads than Yahoo.
Lifetime-based analysis: (8/11/02 – 8/17/02 and 8/18/02 – 9/7/02) Source Click-Thru New Customers Revenue Cost Profit Yahoo! 40,000 1,500 75,000 20,000 55,000 MSN 40,000 400 20,000 20,000 0 AOL 40,000 2,500 125,000 20,000 105,000
Assuming the results above stay constant, by spending on AOL over Yahoo! (the obvious marketing allocation decision), the company would generate an incremental profit of $50,000 each month.
More importantly, by analyzing the effects of these ads over time, a marketer can determine where their best customer are acquired and commit more resources to those customer acquisition sources.
Lifetime-based analyses are a better gauge of actual customer behavior because behavior is ongoing, and every marketer understands that advertising doesn't always lead to an immediate purchase response. Over time, lifetime revenue-based analyses would lead to a rapid decrease in the cost of acquiring customers online, greater profitability, and, most importantly, a fundamental competitive edge.
In Forrester Research's September 2001 report, “How To Measure What Matters,” Forrester interviewed Global 3,500 companies and found that 98% of companies still use the most basic measures of site traffic – hit, page view, and unique visitor counts – to determine success online. This remains largely the case today due to current resource constraints and the relative simplicity and lower cost of implementing session-based analyses.
However, dollars spent on session-based analyses yield a lower ROI than lifetime-based analyses. Whereas use of lifetime customer data warehousing and analysis lead to an accurate picture of customer reaction to advertising over time, use of session-based data capture and analyses evokes the old saying from John Wanamaker, the famous advertiser, that “half of my advertising is wasted, I just don't know which half.”
There is no better time than now to realize the advantages of data warehousing all online visitor interactions to conduct lifetime-based revenue analyses. Not only does it allow a company to build an important online data asset and a distinctive competitive edge, it also enables marketers to analyze a customer's lifetime-revenue response to a number of different advertising and merchandising initiatives.
Companies can muddle along with session-based analysis, but it will likely lead to flawed decisions and another statistic for the next TDWI study.
