The birth of the internet shook the foundation of many markets, in particular our own media industry. Over the past decade and a half, online advertising has blossomed into the potential genius alternative to the more traditional media outlets -- wooing media buyers with the promise of "real-time metrics" and the technology that allows better targeting capabilities. It's all these promises of marketing capabilities that has propelled online media into securing a larger slice of overall budgets -- stealing dollars from traditional media bucket. However, in order for online media to surpass traditional media budgets, the industry must further innovate in terms of scale and targeting. This will involve delivering value above and beyond what advertisers are currently capturing from TV.
One of the critical components that will enable the shift in budget allocation is in the one thing that makes online advertising so unique: the underlying power of ad-targeting data. As it currently stands, the digital industry is overrun with data: "Data, data everywhere, but no one has the means to truly leverage it." The digital ad industry is still a ways off from being able to leverage "data potential" from ad context and consumer behavior.
Reid Hoffman, the co-founder of LinkedIn, has often said that "Web 3.0" will center on data. However, if the next era of improving digital media involves "data," then there are going to be a few different phases within this era for us to test and refine. We've peeked into Pandora's Box and have just begun to scratch the surface of what intelligent data can deliver, but we have yet to truly understand the capabilities. Ad targeting has a lot of growing up to do in anticipation of the next phases of the web, and I have outlined a few potential life-stages of data below.
I have no doubt that we are rocketing through phase two, given the proliferation of cloud computing solutions. As a case-in-point, the number of ad firms built on top of Amazon's AWS platform is incredible. Simply put, the capability to access data at fast speeds is now becoming a commodity. However, if we look farther out to phase three, we see things fall off a cliff; the simple fact is that data currently lacks a means to measure quality.
As we know from Google and the retargeting industry respectively, speed and recency are key. However, given all of the different sources from which we gather ad-targeting data, who is keeping track of the quality? A huge opportunity exists in scoring the quality of data from the different sources that created the intent data. For instance, how are we answering questions like, "How many times has the data changed hands if I'm not receiving this data from the source?" I will concede that some of the quality of the underlying data will prove itself out in the performance of the associated ads, but there's tremendous waste in between. However, if we can't immediately solve with scalable data-scoring solutions, then the next best focus should be centered on consumer behavior.
We've solved contextual ads with "blunt scale" through products like Google's AdSense, but we're not solving for the incredible behavioral data that publishers hold. Another case-in-point, there's a reason why AdSense doesn't perform nearly as well as AdWords for marketers (besides the quality of publishers). The context aspect has yet to be figured out. Google may know the words that are most represented on a page, but it does not mean it understands the behavioral context. Take for instance the specific behaviors exhibited within each of the unique publishing categories below. Each of these verticals has behaviors entirely unique to its type of website. These actions are not exhibited, nor leveraged to their potential, anywhere else on the web.
If the ad industry can find a scalable way to feed these vertical behaviors back into the ad ecosystem, it would take us far beyond just contextual relevance. However, luckily, there is behavior that is consistent across publisher verticals...search! It is one of the core navigational tools we use across almost everything digital (phones, PCs, laptops, TVs, etc.). The opportunity for leveraging search data in targeting has only just begun, and I do believe it's one of the first steps toward extracting behavior to improve digital advertising. It doesn't mean that we'll solve for every "bad advertising example." In fact, those types of digital train wrecks may increase as we try to improve ad relevance.
To improve digital ad-targeting data, not only do we need to capture category-specific events and feed these back into the ad ecosystem, but we also need to provide more feedback loops from the consumer. The more behavioral data we leverage to make ads relevant for the consumer, the more we'll need to give them tools where they can control the inputs. Facebook has begun to build tools here, but has not yet leveraged the data set.
The first base layer example of this on Facebook is where a user has the ability to "like" or "X" a display or text ad. This is a simple form of a user feedback loop. However, even though this data may or may not be leveraged to improve ads within Facebook's media business, it's ultimately held within a walled garden on Facebook. No other ad technology has access to leverage this data for better ad targeting elsewhere on the web. For example, if you "liked" a Nike ad on Facebook, Nike would definitely want to target you as you surf the web, but instead it's outbid by the dancing mortgage lady because Nike doesn't know it's you.
For the sake of simplicity, let's just call this user-specific meta data the "consumer ad feedback info." If this tool were created for improving personalized ads across the web, this would be meta data that the users could carry with them to improve their experience across all publisher sites. For instance, a good example would be if a user denoted a brand is "relevant" (aka, "liked"), then that brand would be shown more often and white-listed across ad exchanges. Conversely, if a brand or advertiser were marked as "irrelevant" or "never show this again," then the brand would be blacklisted for that consumer. This may not solve the scenario for consumers that dislike all ads they see, but it would be a start for improving ad relevance across the web.
In short, the digital ad industry still falls woefully short of where it needs to be in order to garner more budgets from advertisers (the right context, leveraging behavioral data, or providing feedback tools). Luckily, consumers are helping nudge a lot of these budgets online, just by the sheer volume of their digital activity.
On Twitter? Follow iMedia Connection at @iMediaTweet.
"Dartboard on wood wall (miss darts)" image via Shutterstock.