The Long View: Why E-Mail Metrics Are Different
Are you in the process of moving some of your direct mail dollars to e-mail? Here’s a tip: the metrics are different.
And we’re not talking about open rates.
“The trick is to measure not only the results of a specific campaign or tactics, but also measure the cumulative impact of many contacts and multiple tactics,” writes Katie Cole, vice president of analytics and research for Merkle/Quris, in a white paper titled, “Analyzing for True Email ROI.”
Generally, ROI is calculated for snail mail during the entire promotion period and for a short time thereafter. This enables mailers to determine whether the promotional lift was “truly incremental or whether sales that would have occurred anyway were simply shifted forward to take advantage of the promotion,” Cole writes.
Yes, direct mail ROI can be gauged on an ongoing basis—by “deducting any post-promotion deficits from (or adding post promotion benefits to the campaign ROI,” Cole notes. “E-mail ROI models, on the other hand, are continuous—showing period specific as well as cumulative lifts. Aggregate ROI programs for e-mail programs should show the month-to-month (or other relevant time period) variation in ROI as well as the cumulative lift across months.”
That’s not the only difference. In an ongoing e-mail program, the marketer will probably be in constant touch with the customer, compared with the “variable, sporadic contact” in a direct mail campaign. Thus, tests are likely to take the long view rather than consist of mere snapshots.
In direct mail, Cole writes, test controls are “randomly selected for each campaign. This approach is employed in a one-shot manner (i.e., controls are randomly selected for each campaign from a list specifically defined for the campaign).”
But e-mailers face “the ethical and financial considerations of withholding communication from a consumer group that requested to receive them via opt-in,” Cole continues. “Also, since the marketer typically defines the list in traditional direct marketing through some list selection criteria, the distribution is tighter (variability lower). Therefore, the required holdout from the list can be smaller to be representative. In e-mail, “they” (the list member) select “you” (the marketer), meaning that the list is much more variable—and as a consequence and holdout group would need to be significantly larger in order to be representative.”
And the metrics? They’re similar, except that direct mail is likely to include coupon redemption, whereas e-mail won’t. And lifetime value will be added to the e-mail measurement.
Cole concludes: “The measurement of e-mail programs must reveal not only the results of a specific campaign or tactic, but also measure the cumulative impact of many contacts and multiple tactics to determine if the intended effects are sustained over the lifetime of the customer or product.”
To request a copy of the full white paper by Cole, click here. http://www.merklequris.com/contact/wp_index.jsp/
