“Data-Driven Thinking” is written by members of the media community and contains fresh ideas on the digital revolution in media.
Today’s column is written by Robin Opie, vice president of data science at Oracle Data Cloud.
Digital marketing offers a seemingly limitless number of different measurement metrics to evaluate campaigns. Some are confusing and worthless, others offer useful insights, but only one gets a unanimous nod of approval from marketers of all stripes: return on investment (ROI), the causal lift showing advertisers how much more consumers spent because they were exposed to the campaign.
When evaluating best practices to drive ROI, however, the answer often varies (predictably) depending on the person answering the question. A publisher is likely to explain the virtues of high-impact units and viewability rates, while a creative agency may stress the importance of design and aesthetics. Data companies and ad networks will tout the importance of machine learning and programmatic optimization.
All of these will improve campaign performance against traditional performance-oriented metrics. To increase sales, however, advertisers should focus on four fundamental factors.
This isn’t just corporate jargon. The primary reason that campaigns don’t meet their ROI goals is misalignment between the campaign’s goals and its substance, including elements like creative, channels, targeting and execution.
For example, an advertiser may implement a campaign strategy that drives upper-funnel activity, such as increased awareness of the brand, but express surprise when subsequent ROI measurement shows limited or no short-term sales lift. You wouldn’t evaluate marathon runners by their 100-yard dash times. So why structure a campaign to achieve one objective and then judge its effectiveness with another?
As another example, a campaign designed to convert users of competing brands should focus on the number of new or lapsed buyers added – and their expected lifetime value – rather than pure ROI.
Targeting The Right Audience
A beautiful creative, compelling call to action and optimized delivery will fail every time with the wrong audience. For example, the smartest “Buy one, get 10 free” mascara ad on Earth just isn’t compelling to me, while a simple no-graphics ad for a Snickers bar at 5:30 p.m. on a workday could influence my behavior.
Brands that can segment their audiences to reach new buyers, existing but infrequent buyers and frequent buyers – at both the brand and category level – are able to measure and optimize their campaigns to drive sales lift for each group. Significant performance improvements are available to marketers who are more thoughtful when defining their digital audience targets.
Knowing How A Product Is Consumed Should Determine How It Is Sold
How customers purchase a product should drive the design for its campaign. A relatively inexpensive product with less than 5% market penetration and a 12-month average purchase cycle should have a very different campaign – and different ROI expectations – than a highly discretionary product with 40% penetration and a monthly purchase cycle.
Rather than creating cookie-cutter campaigns, marketers should use product nuances to their advantage, both in expectation setting within their organizations and in designing their campaigns.
Measurement Makes Better Creative
It may sound counterintuitive, but great creative goes hand in hand with great statistical analysis. Rather than following a tired aphorism – “humor works best” or “ads need offers” – marketers should start with a strong creative strategy, then rigorously test it.
It’s possible to measure sales lift down to the audience and creative level. The real key to success in this area is to create an intelligent learning agenda, measure audience response as much as possible and continue to adjust.
Dodge The Execution Torpedoes That Can Sink Success
While strategic alignment, audience targeting, product consumption analysis and creative optimization greatly improve the odds of driving sales lift, there are a number of campaign execution issues that can torpedo an otherwise healthy program.
One issue: impression dumping. Publishers that have been unable to exhaust a budget may sometimes open the frequency cap near the end of the campaign, dumping a significant amount of your budget in a short time to users. Advertisers should evaluate impression frequency per user at the distribution level and not settle for an average, which could hide the underlying drivers.
Another problem is audience broadening. Audience targeting specificity is sometimes “relaxed” if a campaign is underpacing. Advertisers should monitor the “within target audience” impression rate and compare publishers to ensure their “in-market chocolate bar” audience doesn’t become “all US consumers who shop at a grocery store.”
There is also viewability and pub quality. Enough has been said on this issue that it doesn’t need rehashing, except to emphasize how important it is.
Finally, there is geotargeting accuracy. If a campaign is geospecific, in terms of story proximity or region-level applicability, advertisers need to monitor impression distribution by geography.
Marketing is never completely predictable, but a well-designed, rigorously analyzed and continuously improved campaign can help unlock the sales mystery originally posed by John Wanamaker, of which half of marketing actually drives sales.
Marketers who follow the fundamentals while proactively identifying and addressing execution issues will improve their odds of driving sales lift.