Question from a buyer: It seems lately I’m finding it harder to justify buying inventory on premium content sites. So many options exist to buy audiences based on data; I wonder if the content around an ad really matters any more. And the premium sites tell me that they can use data so that is like an endorsement. Is reaching audiences through data the only way to go now?
Amy says: We’ve come a long way from the early days of behavioral targeting to today’s data-driven audience buying. There are so many providers of both single-sourced and aggregated data sets that it is hard to keep up. Holding company trading desks add another layer of complexity, but mostly for sellers. There are just too many companies in the data supply chain, which makes it hard for digital media buyers to decide the best way to reach their target audience online.
The clearest benefit of using data to find your audience is price. In most cases, the CPM with a data partner will be much less than if you buy on a premium content site directly. Online ad networks have evolved quite easily into using data to sell their inventory, so you can work with the same companies you have worked with in the past to gain efficiencies in your plan. But I wouldn’t say that content is completely irrelevant. We can all cite incidences where advertisers’ ads have shown up in places where they shouldn’t. When you buy audiences based on data, I would say some kind of brand safety technology is a must.
Determining which data sets are the best for your client is a matter of testing and learning. Some data providers are companies that have been in business for decades and are now bringing their offerings to the online space. Other companies track online behavior directly (aka behavioral targeting) and use those signals to create data sets. Depending on the advertiser, you may not really find data helpful. I think it is more important for high-priced, longer-purchase-process items like travel, finance, and auto. But if there is a particular psychographic you are trying to connect with, data profiles can help you increase frequency at a lower price than a premium content site.
As viewable impression standards emerge, content will once again become important but data will continue to be an important tactics in online media buying. And just like the old days of behavioral targeting, sometimes it will work and sometimes it won’t. Hopefully as the business matures, we will have more public information to inform the data vs. content decision making.
Jason, is data-driven audience buying going to take over?
Jason says: Let my answer be informed by our recent presidential debates: Yes and no. Yes, as for audience buying, data is the only way to go -- but audience buying will absolutely not take over. What we use the data for is the important question that is still unanswered.
As with most things involving digital media, remember that we’re still early in our lifecycle. Even if you consider the full extent of digital marketing’s existence, let’s say we’ve been at this for 20 years. If we took a snapshot of other advertising platforms at their 20th anniversary, television advertising would have been held to practices set in 1961. Print advertising would have remained in its current form in the mid-1700s (I think I worked with a few of those people before).
Don’t take my position on data as being anti-technology. The opposite is true. Television has also long used data to decide on which programs to buy. I embrace the technology as long as it is used judiciously. The problem with our current state of targeting is that the data is used for the most banal of purposes: driving click-throughs. This shortsightedness from marketers, catered to by agencies and performance vendors, is what leads to the draconian thinking of our buyer/inquisitor.
I suggest all sellers and buyers use data to prove the hard questions we want to tackle. How does serving an advertisement to this person, right now, make this individual think my shampoo will be better for him than my competitors’? Work to prove that an ad can convince a woman that your client’s cereal is healthier than that of the competition. Those are the real challenges our industry faces, not deciding whether we can get a monkey to click on a mortgage ad.
Let’s not question whether using data is going to rule the day. It is. Let’s make sure we are using that data to tackle our more difficult tasks more efficiently. I am Jason Krebs and I approve this message.