Ad Targeting 2.0: Timing is Everything
As marketers, we have all heard this mantra before: reach the right person, with the right message, at the right place, at the right time. But do we really have the tools to deliver on all 4 of these phases?
Over the years, I’ve held data roles at TubeMogul (now Adobe), Samba TV, and Microsoft, and I’ve evaluated many companies that seem to be leading the charge. I want to show who’s getting it “right” and expose new advances to reach customers at the right moment in time, like never before.
1. The Right Person
With programmatic advertising, “the machines” can now help us target any type of audience, with any type of granularity, and now, in real-time. Today for example, a marketer can can build an audience of females 25–39 years old in the greater Des Moines area with median household income of $70–99,000 that purchase a name-brand fabric softener every month. However, these audiences are 1) becoming less relevant as they are being affected by the decline of cookie accuracy and 2) becoming less important as real-time targeting becomes more effective at increasing sales (more on this below) and 3) are more exposed to data fraud. Fraud is growing even faster in mobile ad ecosystems, as its estimated that 15% of all mobile clicks are fraudulent, according to a recent study by Tune.
The solution is to work with partners that can connect audiences to more persistent identifiers, such as The Advertising ID Consortium’s Universal ID, led by Appnexus, LiveRamp, and Index Exchange, and recently The Trade Desk. An identity management partner, those that can connect users together to match identities across different online and offline sources like Drawbridge and Tapad is also helpful. And finally, look to independent entities in the fight against ad fraud. For mobile fraud, Forensiq is doing great work.
2. The Right Message
Now let’s talk about the ad itself. Dynamic Creative Optimization (DCO), or personalizing elements of the ad creative at the moment the ad is served, entered the scene 3–5 years ago. Companies such as Celtra and Criteo began offering marketers tools to optimize on a range of characteristics such as the audience data mentioned above, but also external factors such as weather, time and location. And the results? I’ve seen average ad performance increases from 40 — 60%, which are in line with the claims from DCO proponents. Recent advances in AI have accelerated DCO even further. Companies like Persado personalize the ad copy itself in real-time. Even Out-of-Home (OOH) advertising, with programmatic delivery, gives outdoor billboards the capacity to become dynamic. For example, in order to get more store visits, Burger King served ads on digital signage near their stores with dynamic messaging by time of day and proximity to local events.
3. The Right Place
Location-based targeting — the use of phone GPS info, WIFI signal, cell tower data, to determine user behavior — is having its moment right now. But like many other marketing technologies, it is facing its own share of accuracy and fraud issues, as summarized in Adexchanger’s latest Location Marketing Guide. Generally, vendors that use location-based data struggle with pinpointing users within a radius smaller than 10 meters in the real world. Plus, I have personally heard from many ad verification agents that location-spoofing (collecting location data from the bidstream) is fairly simple and prevalent in the marketplace. As you can see, the landscape for location-based targeting is rich with opportunity for improvements. I believe that adding time-sensitive triggers, based on activity and motion data (further described below) to location data is the winning combination for Marketers in 2018.
4. The Right Time
Compared to the others, “finding someone at the right time” is about to evolve into a new level of precision. Currently, most Demand Side Platforms (DSPs) allow for time-based targeting: dayparting, targeting weekdays as compared to weekends, pairing geography and time-zone, etc. However, peppering users with mobile ads throughout a specific time period, no matter how segmented, implies that the highest value user reached today will be available at the same time tomorrow. The technique is imprecise at best.
There’s an emerging technology at the intersection of IoT, AI, and ad-tech that I believe solves this problem: motion sensor processing. Everyone knows that phone sensors can track steps and sleep, but not until recently have innovators begun fully processing the rich signals coming from these sensors (which happen to be omnipresent in smartphones these days.) The applications appear to be endless, and as you can see, real-time motion targeting is an obvious segment ready for prime-time.
In 2017, I co-founded Velocity, a company that uses machine learning to unearth the nuanced information today’s motion sensors are able to detect. In other words, Velocity identifies a user’s actual behavior, using the motion sensors embedded in almost all smartphones & tablets.
Since Velocity can detect when someone is actually paying attention to their phone, entering the mobile advertising vertical with real-time targeting was a natural next step. The technology helps marketers understand when a person is standing in line, lying on a couch, riding as a passenger in a car, or commuting on a train, etc. to determine moments of optimal attention. Conversely, its clear when a device is resting on a table, or is in a purse or pocket, a person is not paying attention. Motion data interpretation can map a person’s behaviors to degrees of attention and receptivity. To make this data actionable, Velocity’s mobile DSP allows customers to target ads specifically during real-time receptive moments, increasing ad lift and conversion rates on average 300–400% compared to inattentive moments.
For marketers who want to control when to interact with their customers and reach them at moments of optimal attention, there’s no time like the present.
-Ishan Bhaumik
