It seems to me that the idea of a DSP was not powerful enough, nor was it innovative enough to stand alone, and so DSPers had to link them together to make a “powerful,” “new” technology and service. This enabled them to proclaim the “death” of the ad exchange (and maybe the ad network as well) and allowed them to create a buzz that will occupy us all (the online advertising industry) for a long time to come.
DSP + RTB is far from being a game-changer in the online advertising industry. A DSP service can be achieved without having to use RTB, and RTB can be used by ad exchanges and networks with equal success for the right media. I will strengthen this theory through a broken down analysis of the basic offerings and characteristics of a DSP.
A DSP has four major uses:
It is able to buy media from various traffic sources (including ad networks, exchanges, publishers etc…), in a single interface for the advertiser, and with targeting and pricing accuracy.
It can manage a global budget.
It can manage a global frequency cap.
It can provide unified reporting for all media sources for optimization purposes or otherwise.
The first and most basic offering of a DSP is its ability to buy media from multiple sources. This ability is achieved, according to the DSP, by using RTB. I agree that this can be achieved by RTB, but it is far from being the only means with which one can buy well-targeted media at the correct price.
DSPs claim to be more transparent than ad exchanges, and thus are able to be more accurate both in pricing and in targeting, but if we ask ourselves where does the DSP get its data (as it is an advertiser platform solely) than the answers are twofold:
From the ad exchange that it is buying from (or publisher or ad network).
From analyzing previous media buys.
It’s clear the ad exchange (which is supposed to have less information about the impressions, and less transparency too) is the provider of a large part of the information that the DSP is using, and thus uses this same information in its targeting and pricing.
As for analyzing previous media buys, the ad exchange cannot do that over different traffic sources, but any ad network can do it for itself and just feed its conclusions to the buying process of the ad exchange. (I will touch upon how to do that that later.)
Managing a global budget and a global frequency cap, assumes an advertiser will run ALL of its traffic on one DSP alone. As we have seen in the current market setup, this does not happen in reality. Almost all advertisers run their campaign with several ad networks, and some of the campaign is even managed by the advertiser themselves and their team, so global budget and global frequency caps are not tallied. They are relevant to the media that is bought through the DSP.
The sophisticated ad networks that use a single ad server to traffic all their campaigns (even the ones bought on other ad exchanges and media sources) can achieve these goals rather easily along with global reporting, as all the traffic is flowing through one ad server.
DSPs represent an evolution of the online advertising industry in terms of its being a platform that integrates more data and more technologies in order to get a more accurate pricing and targeting model — this is its major advantage.
For ad exchanges and ad networks to be competitive and complimentary to DSPs, they would have to do three things that are currently almost impossible in all ad exchanges I know:
The first thing is to create an interface that allows the integration of new technologies. These technologies will mainly add additional information, layers on top of the information that the ad network provides.
The second would be to separate the targeting from the pricing.
Once the first two are accomplished, then they should be ready to receive pricing based on the data they accumulated in the first task (not necessarily in real-time).
If they manage to do that, than the DSP threat will be much smaller as the other three uses can be achieved by any ad network that works with several ad exchanges.