How Publishers Got Ass Whooped By RTB

I have written about the secular trend toward advertisers building out their own trading desks, after agencies built out their own trading desks, in an effort to capture margin from ad networks. What I haven’t written about is how publishers are absolutely suffering in this onslaught. All of this momentum on the advertiser side of the business is occurring because RTB generally, AppNexus (the exchange of exchanges) and OpenRTB specifically, have made it so easy for people to connect to exchanges.

As people connect to exchanges, they recognize that the value proposition of networks has lessened. Who needs a network to pre-aggregate inventory so you can reach an audience when the exchanges provide that service and DMPs are actively integrating into those exchanges to market their services. Suddenly, publishers hear a giant sucking sound and the 60/40 rev share (or even 70/30 back in the day) with networks starts to decline. “The advertisers are moving over to the exchange, you should get there and participate!”

Now publishers have opted for private exchanges to protect the quality of their brand, but that is really about making sure that shady advertisers don’t show up on their inventory. Once you are in the exchange, when the advertiser places a bid, he is bidding on your stuff right along with everybody else. Now, it is true that some advertisers want the good stuff, but some agencies want to burn budgets and show high CTRs/ROI/whatever. I know your inventory performs great, but the trick is inventory that is half the price only has to perform half as well.

Tweets on MarginThere are a couple of signs that point to me as leading indicators that networks and trading desks that rely heavily on RTB to acquire inventory are sticking it to publishers. Exhibit One: Rocket Fuel.


Rocket Fuel has always touted their investment in RTB and optimization. Even as their business has grown, their margins have skyrocketed to 60%. It is no wonder they have grown into a dominant ad network: They are making 20% more than everybody else! That gives you a lot of flexibility to fund growth. What is really amazing about this is that the margins continue to increase even as they grow. If you think about how a performance-based business like Rocket Fuel should work, to deliver incremental performance for advertisers should require additional investment.

So advertisers are increasing their spend with Rocket Fuel plus new advertisers are introducing competition to Rocket Fuels internal auction, yet prices are going down? Now, I am sure that what Rocket Fuel would like you to believe is that every 90 days their optimization algorithm is getting fantastically better. But I would suspect that it would take amazing algorithmic improvements just to hold margin flat while the business grows that fast. Think about it: If I want to buy 100 conversions in Q1 and I pay $100, if I want to buy 200 conversions in Q2, the odds that I can acquire them for $200 is very, very low. So it requires massive yield improvement just to stand still.

marginal cost

A far more likely explanation for these improvements is that the yield improvement comes from massively lower supply prices. There are any number of ways to generate these: Rocket Fuel could have

  1. integrated with new exchanges where inventory was far, far cheaper,
  2. publishers with strong performing inventory could have added more of it to the exchanges, or
  3. publishers with average performing inventory introduced tons more of it to the exchange, lowering average prices.

Now, Rocket Fuel was pretty big before they went public, so I think one could safely rule out the first theory. The odds that awesome inventory was just waiting to be scarfed up and Rocket Fuel spent years unaware of the existence of this inventory seems low. I am sure that some publishers with strong inventory probably grew, but the most likely scenario was probably the third: Average and poor publishers flooding the exchanges with increasing quantities of poor impressions. There are two effects of this: It draws down the average price as the average piece of inventory becomes cheaper and poorer performing. The second, and probably equally important effect, is that people with poor algorithms, regardless of whether those people are Rocket Fuel or not, fail to recognize that this inventory is the bad stuff. This draws competition away from the highest quality inventory, dragging down top-end prices. And because the optimization is happening cross exchange, the protection of a private exchange is probably limited.


If you are looking for further proof that this trend is real, look no further than PlaceIQ. PlaceIQ is a relatively young mobile ad network that is growing incredibly fast. Here, they indicate that they grew revenue tremendously and became profitable.

How does profitability in the face of rapid growth happen? One of two ways. One is that the money came so easy that they just blew through the top line without having to ramp the bottom line. There is almost certainly an element of that. They indicate they did 2x the quarterly goal. Accidental profitability sometimes happens when you do that. But the second way is awesome margins. If margins are skinny, becoming profitable requires a lot more growth. Accidental profitability is a lot easier at 60% margins than 40% margins.


I don’t have good data on other networks (that I can publicly disclose), but I suspect this is a trend. So all of these networks are experiencing explosive growth on the top line while margins are increasing. That means that all of this growth is making less and less money for publishers. This is a big transfer of wealth as the market moves money from inventory creators to arbitrageurs and value adders.

Publishers viewed RTB as a big opportunity: Theoretically, the real-time auction would allow them to extract maximum rent for each impression. In truth, I think what we have seen is that the clear visibility into the supply-side that RTB has offered the demand side has taught the demand side that there is tons of supply and none of it has any value.

More posts coming on Programmatic Guaranteed, War Stories of Raising Capital, etc. Make sure you sign up for my newsletter.

I am sure there are more examples of networks with explosive margin growth. Further, I would love to hear stories about how CPMs for publishers are being affected since I am not on that side. Let’s discuss in the comments.


Opt In Image
Get Access To My Best Start-up Ideas, Tools, and Growth Hacks
Learn whatever tricks I have from starting and selling 2 companies and running digital advertising products for more than a decade.

You will find helpful stuff on this blog, but I have built a free VIP section where I host the best downloads, more content, and tools you can use. For example, decks I used to get term sheets from VCs are posted there. Free parts of the book I am writing about lean startups, growth hacking, and digital advertising. You do want to be a VIP, right? Get in here!

Leave a Reply