Let a Thousand Ad Networks Bloom
There's no end to predictions of the demise of ad networks. The basic story is the same: Ad networks used to differentiate themselves based on inventory, data and relationships -- but now everyone has the same access to inventory and data through exchanges, and relationships come and go. While there's some truth to this narrative, the part no one talks about is the blooming of whole new categories of ad networks based on differentiated inventory, proprietary data, and industry-specific relationships. Let me explain.
Over the past month I've had conversations with three entrepreneurs leading successful startups outside of advertising, all of whom were contemplating an entry into the display ad space. The businesses were different, but the rationale was the same: "We've got great relationships with media companies in our space, we've got better data on customers in our market than anyone else, and we already service the advertisers looking to reach these customers." Voila! An ad network is born!
Let's start with data. Call it behavioral if that makes you comfortable, but it's really just cookie-based data that says the user is interested in something very specific. For example, you run the definitive mailing list and event planning business for certified plumbers (it's possible!). You've got some kick-ass data about plumbers. Now what are you going to do with it?
- You can sell ads on your site, which is great, but you probably don't have a lot of volume.
- You can sell the data to other marketers, but how liquid a market are you going to find for buyers looking for plumber data?
- You could use the data yourself to target display ads to plumbers in the wild. Congratulations, you've just created a plumber ad network!
But isn't selling and buying ads a hassle? Why can't I just sell the data through a data management platform ("DMP")? Good question. DMPs are great, and an important part of the ecosystem. But they're largely designed to sell broadly applicable, high-value data, not niche data. To get the most out of niche data you need the relationships with sellers and buyers, which is hard to find, except ... that's exactly what you have!
Consider that there are different appropriate data strategies based on the value and breadth of applicability of your data. If you've got data with broad applicability then it makes great sense to extract value by selling or renting it. And if you've got data that's generic, like demographics, then it makes sense to use it yourself as a way of improving your yield or adding value to advertisers. But, if you're got narrow data that is valuable, it's really up to you to extract the value -- and jumping into the marketplace as a fully participant is the best way to do that.
The next challenge is accessing the inventory to create a successful ad network for plumbers at scale. In the first phase of Internet advertising, that meant paying up to a portal for access. In the second phase, that meant using keywords on search to buy users based on their intent. Now, you can go into the open real-time bidding markets and grab that inventory on a per-impression basis with little overhead or technology investment. This is display advertising that scales from the smallest to the largest opportunity.
As for finding buyers, if you already are the definitive source for the plumber community my guess is that you have better sales relationships with companies who want to reach plumbers than anyone else. And while those buyers may not know a DSP from a PVC, part of your value-add is packaging and simplifying the value proposition so that even a grout salesman can understand it.
Plumbing is a slightly absurd example, but it's just as real as many other opportunities that can be used to bootstrap a next generation ad network. Next time someone tells you that ad networks are dead, you remind them of the old adage, "there's never a plumber when you need one."