By ignoring non-click conversions in the calculations, most online campaigns understate ROI. Learn how to calculate them and what implications they have for marketers.

Most online marketers are using Website conversions (sales, registrations) to gauge the effectiveness of their Web media campaigns. However, with very few exceptions, these marketers are significantly undercounting the total conversions that result from those campaigns. By neglecting non-click conversions, they arrive at an artificially low estimate of the return on their advertising dollar.

By crediting only those conversions that follow from a click, most advertisers ignore what often proves to be the lion’s share of campaign response: non-click conversions. This article defines non-click conversions, describes how to quantify them, and discusses their implications for savvy marketers.

Non-Click Conversions Defined

A non-click conversion occurs when a user views an online advertisement, does not click, but converts on the advertiser’s Website within a specified window of time (e.g. 10 days after viewing the ad). Recognizing non-click conversions requires the ability to associate actions that occur on your Website back to the Web media impressions that may have stimulated them.

However, it’s obvious that not all non-click conversions are attributable to an online campaign: some of them are the result of TV advertising, word-of-mouth, or loyal customers doing what they would have done anyway. The trick is to separate causal non-click conversions (i.e., those that resulted from the online campaign) from incidental non-click conversions (i.e., those that would have happened anyway).

Before delving into that thorny question, it’s worth asking, “How big of a problem is it simply to ignore non-click conversions?”

Based on experience with a variety of different clients – larger brands, smaller brands, online and multi-channel – the non-click factor may range anywhere from 1x to 5x. In other words, for every click-based conversion from a given Web media campaign, there are between one and five non-click conversions. At a minimum, these numbers imply a two-fold improvement in online ROI; in the most extreme case, a six-fold improvement.

The non-click factor tends to be higher or lower depending on several variables:

• Brand awareness – better-known brands typically have a higher non-click factor.
• Weight of offline vs. online advertising – companies that advertise across media often have a higher non-click factor.
• Degree of targeting – more targeted placements tend to generate a higher non-click factor than does run-of-site inventory.

Quantifying the Non-Click Factor

Let’s return to the question of how to tell causal non-click conversions from non-click conversions that were driven by other means (TV advertising, word-of-mouth, newspaper promotions). In other words, how is it possible to know which non-click conversions were due to the Web media advertising and which ones would have happened anyway? This can determined through either a highly opinionated debate or a fairly straightforward bit of research.

Using the TrueLift™ methodology shown below, one group of viewers (test group) is exposed to the advertiser’s banners online; the other group (control) is instead served a “dummy” banner. The control group – which did not see the online campaign – reflects the Website visits and purchases that would have happened anyway. By this methodology, any incremental Website visits and purchases observed within the test group can be attributed to the online campaign.

Example A: Non-click factor = 4x

Interpreting the Data

Again, the control group establishes the baseline. Because online advertising was withheld from the control group, these users had zero click-based conversions. At the same time, the control group had 120 non-click conversions. By contrast, the test group had five click conversions and 140 non-click conversions. Since the only statistical difference between these two groups is whether they were exposed to the online campaign, this leads to the conclusion that the 25 additional conversions among the test group were causal.

The non-click factor is calculated by dividing the incremental non-click conversions by the incremental click-based conversions. In this case, 20 non-click conversions divided by five click conversions yields a non-click factor of four. Based on the results of this TrueLift test, a marketer could safely deem that in a similar future campaign, every click-based conversion represented four incremental non-click conversions.

Implications of the Non-Click Factor

The price of not accounting for non-click conversions is money left on the table. Advertisers get into trouble when they count all of the costs of their online campaign, but only a portion of the benefits (i.e., only the click-based conversions). This flawed view results in an artificially low ROI for Web media, and the elimination of a large volume of sales that are actually cost-effective.

Let’s say a company has a willingness to pay $25 per sale. By crediting only click-based conversions, the Web media campaign is optimized to drive online sales for $25 or less. As all of the $25 sales on the Internet are captured, the online sales growth flattens.

By proving causal non-click conversions, online sales could be instantly doubled or tripled without eroding the margins. By conducting a TrueLift™ analysis, statistical confidence is gained about which of the sales are rejected – when they appeared to cost $50 – marketers can now harvest at a true cost of $20 (cheaper than some Search results).

What Now?

I recommend conducting a TrueLift test and re-calibrating results annually. However, generic factors should be avoided. By arbitrarily using the wrong non-click factor, marketers can just as easily overpay for online conversions. Properly measured, the non-click factor can be a powerful ally in evaluating and optimizing Web media campaigns.

Peter Greb is an account director at Avenue A, an interactive agency and operating unit of aQuantive, Inc. (NASDAQ: AQNT), a digital marketing services and technology company.