“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 Eric Berry, co-founder and CEO at TripleLift.
Despite being one of the most revered techniques in digital advertising, optimization remains a highly misunderstood concept. While everyone responsible for running a digital ad campaign figures optimization will ultimately help boost performance, what they want or should optimize against is not always as clear as it was back in the early days of the digital ad game.
Before I pivoted to the native advertising world, I spent several years at AppNexus and had a front-row seat to how various optimization strategies evolved and matured in our ecosystem. Although it was exciting to see how marketers were beginning to embrace real-time bidding, I couldn’t shake the nagging feeling that the industry was taking a myopic view of the potential of RTB optimization.
RTB originated as a means to monetize remnant inventory and was used primarily by performance-oriented marketers, especially in the beginning, so they had a disproportionate impact on how these optimization systems evolved. Direct-response marketers make money when they can achieve their goals for less than they get paid, meaning arbitrage. Cheaper clicks and cheaper conversions are always better in direct-response campaigns than anything that costs more.
Thus, the modus operandi has been to focus on purchasing the cheapest impression possible that generates the desired response. RTB optimization has traditionally been executed by giving the engine a direct-response signal on which to optimize. That’s great for advertisers measuring success by conversions or actions, but what about brand marketers who were steadily increasing their spending on digital media? When building brand equity and awareness online, there isn’t the same signal. Clicks aren’t necessarily indicative of the success of the campaign, and conversions are rarely the goal of a branding campaign.
Since the RTB market was in its nascent stages, few people’s feet were held to the fire in terms of the tactics employed. Performance campaigns effectively “optimized down” to find the cheapest inventory for the advertiser that would perform adequately; the goals were simply to drive performance at the lowest price, regardless of anything else.
This would drive the optimization engines to find the cheapest clicks satisfying advertiser brand-safety constraints, but often would actually be below-the-fold placements not often seen by most users – and possibly optimizing to sites more frequently visited by bot traffic. These strategies don’t cut it for sophisticated brand marketers.
RTB does, however, offer an impressive opportunity to deliver meaningful brand impact – brands just need to think differently.
For instance, if a brand employs the practice of “optimizing up” to find the best inventory that drives the most impact, the value attained would easily trump that of low-cost arbitrage. By optimizing up, you try to engage with the target audience in the most effective way possible. This means trying to identify the sites that deliver the best impact for your brand, trying to understand how much it costs to buy media on those sites and then determining where on the site’s price curve you should bid based on the likelihood of high-quality branding impact.
As the site’s price curves change, as you show ads to the same user multiple times, and as the nature of the users changes, your bid will change accordingly. For a site that delivers an incredibly high-impact ad unit that is on-point for a brand’s audience, an advertiser should be willing to pay higher on the publisher’s price curve. If you start with a maximum price and a fixed technology margin, you will optimize towards high-quality branding impressions, as opposed to cheap, direct-response clicks.
While costs may be higher, you get the job done in a more meaningful manner. It is time that RTB be used for getting beyond just clicks and conversions and also for building relationships with consumers. The truth is that in the end, only a small fraction of users will ever click or convert on an ad, and indeed measuring this response barely scratches the surface for the efficacy of display advertising.
Impact can be measured in contexts such as visibility indexed against target demographics and target audience over indexing on brand lift – what does it take to buy your impact on each site, where are your demos going and how much can you buy them for.
By switching how you optimize and rethinking how you value display advertising, you can see incredibly different performance, which may result in more people actually seeing your ad, even though the clicks might not be as cheap. In the long term, your ads will be more effective, because they’re shown in better placements and seen by more people.
Like the adage says, you get what you pay for.