Beating “Hate for Profit” is Profitable for Advertisers. Here’s How.
The Stop Hate for Profit website is direct in its condemnation of Facebook as the facilitator of hate and disinformation: “Hundreds of businesses are standing together to support our deeply held American values of freedom, equality and justice and as a result, are pausing advertising on Facebook’s services.”
While this boycott may help to clean up Facebook in the near term, the reality is that hate, disinformation and low-quality content permeate the entire internet, not just Facebook. And in a game of whack-a-mole, if you whack Facebook, inevitably a new platform or site will pop up to deploy cheap & addictive content to steal consumer attention to win ad dollars. As long as advertisers use faulty metrics to guide their ad investment, there will always be organizations happy to take advantage and hurt the wider eco-system.
Two things have collided to get us to a place where advertisers are now inadvertently funding “hate for profit” on Facebook, and elsewhere: audience targeted advertising plus faulty (but more easily deployed) ad measurement techniques.
Buy Side
The promise of audience targeting is to place ads in front of the exact “right” person regardless of the content they’re consuming and its largely fueled by reinforcement learning algorithms. Tell the algorithm that success is achieved when action X is performed by the consumer at price Z or lower and get out of the way. Unfortunately, by getting out of the way and measuring success via CTR, last touch/click, or even multi-touch attribution, audience targeting algorithms are being trained to find consumers who are already bottom funnel and serving them low quality, inexpensive ads that are likely unviewable or unimpactful. Doing these two things represents the most efficient path to gaining the desired action at the lowest out-of-pocket cost for the algorithm…precisely what we told it to do. Each time it “achieves” the outcome, its learning is reinforced, and it will lean further into extracting value from an advertiser, rather than creating it. As advertisers we end up rewarding a correlation, in place of a causation.
Sell Side
Given this new method of buying media, platforms, networks and all manner of advertising sellers are no longer incented to create high quality content, they’re now incented to attract and retain audience attention at the lowest possible cost. And so…we end up with clickbait, addictive “newsfeeds”, shock & awe disinformation and general low-quality content online. These kinds of tactics play to our baser human psychology and attract eyeballs for a low cost and keep them glued so the seller can load up as many ads as possible (similar to a slot machine). And because advertisers are largely not measuring if these ads caused an action to occur, the quality of the environment, the content/ad experience, and the impact of the ads on consumer behavior doesn’t matter.
Marketplace Effects
Over the past decade the media eco-system has seen an explosion in the number of available ad impressions, and this new dynamic of buying & selling digital media is a large reason why. Creating low quality content and ad experiences is a cheap endeavor. As a result of increased supply, CPM’s have naturally gone down, and higher quality publishers have had to resort to pay-walls and most dangerous of all for advertisers: entirely ad-free experiences, in favor of consumer direct subscriptions. High quality audiences & content are going behind ad-free paywalls. Correcting the approach to how we measure ad success is the best thing the buy side can do to reverse this affect.
Proper Measurement
For advertising to be truly effective it should have a causation effect on a desired business outcome. Which means that if the ad had not been served, the consumer would not have taken the desired action.
In the medical community the gold standard of measuring any new medication is a randomized control trial where you take two equal populations with the same conditions and give one a test pill and the other a placebo. From there you can accurately measure the effectiveness of the pill administered. That’s the same testing setup and rigor that we can apply to media buying today in order to understand if an ad buying tactic, publisher or channel is truly impactful on driving sales.
Three Ways to Set This Up
- Leverage publisher direct measurement capabilities.
While not the preferred option, it’s still better than nothing. Many publishers, platforms, DSP’s or ad networks will provide their own system for measuring incrementality from the ad buy. Not exactly an impartial party measuring your results, but you will almost always see your “performance” metrics dip when you apply this new form of measurement. And that’s exactly what you want to see as you start to remove the fictitious performance & waste. - Pay an external vendor specializing in incrementality measurement.
A couple of interesting startups have emerged recently with proprietary methods of measuring for incremental sales impact across channels and we’d expect to see more of these companies crop up in the years to come. The advantage here: more impartiality, can be deployed across a number of channels and likely some very good data science/methodology behind the products. The downside…is outsourcing what will ultimately be a critical function for effective paid media buys moving forward. - Last, and most ideal of all, is to utilize an advanced analytics/data science team from your media agency of record or an internal team to build a customized solution.
While startup costs and legwork are downsides to this option, the long-term ability to separate true signal from noise and learn and optimize directly from the intelligence is a major competitive advantage. Let your competitors believe all is well because their ad-server report says so…. your top-line growth compared to theirs, will say differently. Companies that can measure & optimize paid media closer to home and based on true incrementality, will absolutely win an outsize share of their customer and/or market.
When we implement our own form of incrementality testing, a clear dichotomy emerges between the kind of environment that is truly impactful on sales and the kind of environment that only appears to be on the ad-server report. Advertising run alongside higher-quality journalism & publishers, in cleaner, well-lit environments, simply sells more product for our clients and delivers greater branding impact. By applying incrementality testing and optimization we reward good journalism, starve the bad, help promote a cleaner internet experience for consumers and, oh yeah…we sell more product.
Longer term, if advertisers right-size their investments toward quality, we should see a welcome reduction in inventory levels, an increase in CPM’s for the quality that remains and a lowering of paywalls once again to quality content & audiences.
The ability for advertisers to target specific audiences has moved directly in line with an increase in disinformation, but that doesn’t mean audience targeted advertising is bad, in fact it’s an incredibly powerful tool to make ad budgets more efficient & effective and improve consumer experience. But it needs to be measured properly or most stakeholders in the process, including advertisers, get hurt.
