At first pass these demand-side networks seem like a good idea; they allow for the agency to retain more control over the inventory their clients are purchasing, thus stewarding their brands in an environment that can be risky at the very least. This is a strategic reason, one that I believe in wholeheartedly. Being something of an agency purist, I believe that strategy -- not technology -- is what should drive the relationship between client and agency, not technology.
Unfortunately, many agencies are developing these solutions for revenue reasons. Many media buying shops have identified the volume of dollars they spend with the ad networks and have theorized that if they create the same solution, using existing ad exchange technology, that they can avoid passing these dollars along.
In doing so they are gambling with their clients' dollars. Most of the ad networks, at least the reputable ones, have spent millions of dollars refining their algorithms and developing technology that makes their solutions work. In many cases they will happily open the kimono, so to speak, and share the breadth of this technology with you to prove the point that this is simply not easy.
The first dirty little secret of most of the demand-side networks is that they're staffed by two to three employees working some spreadsheets and balancing loads to optimize inventory for their clients. This methodology simply can't rival that of an investment in the scaled solutions offered in the marketplace. Trust me; I tried it and it simply doesn't work that well.
The second little dirty secret is that many of the agencies building these solutions are including them on all their clients' media buys and getting paid on both sides. When an agency buys media, it typically charges a media commission of from 5% to 15%. Many agencies are applying that commission to the media they buy on their own demand-side network.
This would be OK, if it weren't for the fact that agencies are also making money on the arbitrage buying on the other end, which is the way these networks make their money in the first place. As an example, that means the agency can buy the inventory at 50 cents, sell it to its client at $1, and keep the 50-cent difference. This would also be OK if agencies weren't charging commission on the media buy, thereby double-dipping.
If I were a media-buying client, I would demand in my contract that my agency not charge commission on any media run through the agency's demand-side solution. If the agency can achieve my objectives and make money through the arbitrage, that would be acceptable -- but the client's needs come first and should be the driving decision for using that platform, not the agency's desire for more revenue.
And of course, what goes without saying is these advertising buys on the demand side networks should be held to the same -- if not higher -- goals than the rest of the media buys. It's arguable that they should be held to a higher metric since the agency does indeed have a stake and does have stronger control of the inventory.
So if your agency is creating and spending your money on a demand-side network, don't be afraid to ask the hard questions and understand what their goals are with the platform. Ask to meet the team working on that solution and understand the technology behind the solution. But understand, also, that that solution is no different than the rest of the marketplace. There are other solutions that can work for you.
There is no silver bullet in online advertising. Beyond search, there are few scalable solutions that can rival ad networks and their targeting capabilities, so don't overlook them. Just be sure to ask the right questions before you spend your dollars with any solution that presents itself.