Is Hyper-Targeting Overhyped?

The promise of behavioral targeting largely hinges on achieving a granular understanding of customers at scale. Some advertisers can accomplish this relatively well with first-party data, which tends to be more accurate because it is gleaned directly from customers and audiences. But to inform hyper-targeting, most advertisers resort to third-party data, which is often aggregated, probabilistic, and sold by brokers.

The problem with depending on third-party data and even second-party data (information offered by other organizations with first-party audience relationships) is that it is often inaccurate. For example, a joint research study from the MIT Sloan School of Management and Melbourne Business School found that the accuracy of gender-based targeting with third-party data was worse than 50%. That means you would have a better chance of serving an ad to the desired gender if you did no targeting at all.

On top of that, behavioral targeting is only getting harder due to the increased fragmentation of user attention across devices, platforms, and channels; data hoarding by walled gardens; and anti-tracking changes like the sunsetting of the cookie and Apple’s deprecation of its mobile identifier. In short, hyper-targeting will get both harder to execute and more expensive. While some brands do see success with personalized hyper-targeting tactics, the realities of that success can be a bit more complicated, especially given the use of flawed measuring tactics. It’s a bit of a chicken-or-egg dynamic: a brand may be seeing high conversion rates based on hyper-targeting, but the advertisements in some cases may just be getting served up to the individuals who are already most likely to convert and would buy without the ad.

So, if contextual targeting is more effective at similar cost, many brands urgently need to reassess the balance of their targeting tactics.

Many in digital advertising are familiar with the argument that behavioral targeting is risky because it often relies on low-quality data. But, proponents of hyper-targeting might argue, what about first-party data? Surely, if advertisers can use granular, privacy-safe, accurate information about audiences to target them at scale, that will produce high conversion rates and superior business value, right?

Not exactly. Analytic Partners analyzed the ROI of five different targeting tactics by an automobile business. Those targeting strategies included hyper-targeting based on first-party data, contextual, broad reach, programmatic, and third-party. Of these five, the brand was surprised to learn that hyper-targeting was the least efficient, especially due to its high costs. Contextual topped the list.

The brand recalibrated its targeting strategy to deprioritize third-party and hyper-targeting while increasing the share of its ad spend allocated to contextual advertising. Focusing ad buys on media partners who would draw audiences interested in the category and/or related areas and that were aligned with the brand’s messaging on safety allowed the advertiser to quickly increase its ROI by 30%. This strategy was more successful because it helped the brand reach relevant audiences that it might have otherwise missed.

Concerns about an impending recession are incentivizing advertisers to cut unnecessary marketing spend and build programs with the highest possible ROI. For many, the first step in that process should be determining which targeting methods are both high cost and relatively ineffective, redirecting ad spend to more cost-effective targeting methods.

Hyper-targeting and third-party data-driven targeting do not just lure brands into potentially privacy-unsafe territory, but they also tend to be expensive. This gives advertisers another reason to interrogate their targeting investments with an eye toward the possible efficacy of contextual advertising.

Brands should also remember that conversion tracking and click/view through metrics fuel the perception that certain executions work, but these methods often overstate marketing’s role. This is particularly true for investments geared toward clicks from loyal or likely customers. Ensuring you have a comprehensive measurement in place to measure incrementality vs likely to convert is key to combat these overstatements.

Some questions advertisers should ask to begin auditing their targeting programs include:

• What is the cost of buying against each of your targets? What is the cost differential between your broad reach and more narrow targets?
• How confident are you that these targets are accurate, in regards to both who you’re looking to reach and whether they are the right audience?
• If narrow targeting is charged at a premium, are the additional working and non-working costs offset by gains in effectiveness particularly when third-party data is used?
• What first-party data are you collecting, and is your database large enough to guarantee cheaper costs?

Trends like a recession, fresh regulations, and questions about the efficacy of targeting can lead brands astray, encouraging them to pull back on ad spending and cede ground to competitors. But advertisers do not need to cut back on advertising altogether to get more for their dollars, and they do not need to feel overwhelmed by market challenges.

With the right questions, capacity and approach to measure which tactics and executions are working and driving incremental business, marketers can run a leaner operation while continuing to make data-driven decisions that drive success and grow market share.