Programmatic

People-Based Marketing’s Biggest Problem: The Retargeting Conspiracy

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 Andreas Reiffen, founder and CEO at Crealytics.

If there’s one indisputable truth about digital media, it is that digital is measurable. While that’s bona fide, debate still rages around what’s important to measure and how to measure it to understand if ad spend actually works.

For example, retail marketers can take two very different approaches to reaching customers: placing very high bids when consumers are most likely to convert, or placing high bids long after an initial site visit when consumers may need reminding.

There’s no clear answer to which strategy is the best, but the former often produces some of the best-looking KPIs. As a result, marketing teams double down on this approach to keep CMOs and CFOs off their backs. Even if the marketing department loses incremental revenue, they are held captive to the idea of great metrics.

Since there is no easy way to measure if ads drive incremental revenue, it is in the best interest of performance-marketing directors, retargeting companies, ad agencies and Google to aggressively target consumers who are highly likely to purchase anyway. It amounts to a retargeting conspiracy among willful participants, and it threatens to drag down digital people-based marketing’s potential long into the future.

This issue is the direct result of a measurability problem. Even seasoned performance marketers have difficulty telling if their campaign numbers just look good or if they are actually good.

Assume you are tasked with driving sales for the Banana Republic store at the Westfield World Trade Center in lower Manhattan. You have printed vouchers offering a 10% discount and will receive a commission for each redeemed voucher.

You spend the first day at Battery Park finding tourists who fit well with the brand, based on their appearance and inferred preferences. You realize that very few customers walk to the mall and make a purchase. The next day, you position yourself at the entrance of the mall and see more redemptions, resulting in a higher commission.

Your employer is excited about the performance. Eager to please, you walk into the Banana Republic store and hand the vouchers to people waiting in line for the register, and even more take advantage of the offer.

This is how online people-based marketing works today. Tech providers put retail ads in front of people who are already about to buy and credit themselves with the sale. But there’s a paradox at work here: The better the numbers look, the less impact you actually have.

The few people from Battery Park who eventually made a purchase would not have done so without the voucher, while those waiting to check out had already decided to buy. In both cases, Banana Republic paid for the 10% discount and commissions. Move this online and that equates to the media spend costs and fees paid to the targeting provider.

So much of the people-based marketing industry just chases consumers who have already signaled that they’ll make a purchase eventually. A person who puts products in a cart and then abandons it is giving a very strong signal. And while solving cart abandonment is at the heart of many retailers’ ad strategies, it may not be as urgent a priority as it often feels. Ad platforms jump on these consumers to steer them back to complete a purchase, and doing so may bring incredibly high return-on-ad-spend (ROAS) figures.

Many tests show that recency and engagement are negatively correlated with incrementality. The ads look successful, but it’s debatable if they’re actually influencing consumer decisions.

Some make a purchase from the same retailer every week. When ads are targeted toward this cohort, they will return great results. But that’s because this segment was likely to buy anyway. It’s hard to comprehend the value and influence of that ad if buying propensity has nothing to do with it. Contrast this with a customer who was a regular purchaser over an extended period of time but recently stopped buying. Reaching this audience with an ad might make them more likely to buy, and that’s incredibly valuable.

One way marketers can learn is to measure incrementality, or how much more revenue the brand gets as a direct result of advertising. This is well within the retail marketer’s reach using Google Analytics. Marketers can run A/B tests that split their site visitors evenly into two groups. The test group will see ads, while the control group will not. They can then follow these groups over time to see if showing ads to the test group increases conversions and revenue in comparison to the control group.

Over time, the brand can test smaller, more specific audience segments, such as cart abandoners versus people who have already bought. However, marketers may find that some of the metrics they prided themselves on, such as ROAS, are now much lower than before. That’s because this is finally an accurate read on the ads that do the heavy lifting in driving purchases, not the ads reaching existing customers.

Getting away from people-based marketing’s flaws will force marketers to wean themselves off problematic metrics and identify better KPIs for measuring success. The better the numbers look, the less incremental they may be. Those who approach their campaign results with healthy skepticism are more likely to find efficiency from their ad spend, along with a better sense of how to invest their money and when to pursue consumers.

Follow Andreas Reiffen (@AndreasReiffen), Crealytics (@crealytics) and AdExchanger (@adexchanger) on Twitter.

This post was syndicated from Ad Exchanger.