PubMatic CEO: “Media Arbitrage Models Aren’t Profitable”

Rajeev Goel CEO PubMaticAs publishers respond to requests for private marketplace and automated guaranteed deals, logistical challenges mount. Publishers must bring in an advertiser’s data and package inventory, and make it discoverable to buyers.

Which is why PubMatic on Monday rolled out a buyer-side portal where an advertiser can shop for automated guaranteed inventory.

PubMatic founder and CEO Rajeev Goel also plans to strengthen PubMatic’s analytics products, allowing direct sales staff to use the product to monitor deals and gather intelligence across all their sales activities.

Goel spoke with AdExchanger.

AdExchanger: Why do private marketplaces have so much more momentum in the market than automated guaranteed?

RAJEEV GOEL: The shift to programmatic started in the indirect category, and PMPs are a very natural extension of open RTB programmatic. It’s using the same underlying tech to transact, but making it more transparent and putting controls around it.

In automated guaranteed, it’s a different set of tech capabilities. It’s about workflow automation. Then adoption becomes a challenge. You need a catalyst to drive that adoption, and that’s why I’m excited about our approach. It will take all the inventory in private marketplaces, and make it available through automated guaranteed. That will solve that technology adoption problem.

Why have public ad tech companies not done well as a whole?

When I talk to investors, there’s confusion about the different business models and which models will generate profit in the future. The media arbitrage model or the ad network model generated revenue quickly, and they were the first public companies in the space. The other model is the software platform model. They take longer to grow, but they are more profitable models. I don’t think any of the media arbitrage models are profitable.

I think it’s similar to what we see in the 1999-2000 dot-com era, where Wall Street was fooled on the revenue growth front, and willing to give credit to companies generating profit on non-profit-based metrics, like eyeballs, traffic or number of transactions. Then the cycle changed, and it came back to profit.

Now what you see companies taking more time to achieve a high rate of profitability before they approach the public markets. I think all of that will be positive. It’s forcing entrepreneurs like myself to focus on building a long-term profitable, viable business, and that’s a healthy thing for the industry.

What metrics do you look at to evaluate your business?

We made a decision to focus on revenue growth and profitability vs. looking at revenue growth at all costs. I can’t share too much about financials, but we hired 225 people in 2014 and last raised money two and a half years ago, so we’re building a self-sustaining business. We’ll continue that balanced approach.

Will you grow through acquisition?

Mocean was our third acquisition in three years, so we’re fairly acquisitive. We look for three types of acquisitions: something to help us speed up entry in a market, like a local player with a team on the ground. Second would be a product and engineering team. If there’s a particular area of product development focus, we might acquire a team. The third would be acquiring a product.

What’s your take on the imbalance between the sophistication on the buy side vs. the sell side?

There’s certainly a provider imbalance, but the sophistication on the sell side is now rivaling the sophistication on the buy side.

The buy side is transactional in nature. On a campaign basis, people are thinking about efficiency and optimizations. There’s a lot of room for people with better mousetraps to take some of the spend.

The sell side is much more of a platform model. The publisher side is constantly challenged with simplification through unification, it’s much harder to run their entire business. That’s the reason there’s a provider imbalance.

Where are you in native programmatic?

We are in the early stages of bringing that to market. Programmatic native will be deconstructing an ad unit into its components, and assembling an ad on the fly based on contextual relevance and the user. We have unique access to the page or app environment, and the attributes of users, which you don’t get on the buy side

The Bidtellect [integration with its native ad exchange] is one of many to make in that space. We’re testing that out now in the first half of the year so we can scale up with more publishers. One of the great assets we have, from acquiring Mocean, is we support native on a direct basis. Publishers can do direct, programmatic or a mix in our platform.

Yet fragmentation often helps publishers. When they work with multiple supply sources, they get higher bids and more revenue.

There are temporary dislocations. If we integrate a DSP for a specific type of ad format before someone else, there could be a pocket of time where demand is unique. Or there could be technical issues: When you get into mobile, for example, there are different ways to pass location data, and that can affect things. All of these drive the temporary dislocations in monetization from one platform to another. Depending on [the] publisher, that can be a meaningful difference for three to six months.

But where the industry is headed, it’s becoming less about the eCPM an SSP can create [for publishers], and more about how partners help them understand their business. It’s workflow capabilities with direct and indirect, real-time analytics. The value prop is shifting away from a CPM of X and shifting into “Do I have a better understanding of my business as a publisher?”

Enjoying this content?

Sign up to be an AdExchanger Member today and get unlimited access to articles like this, plus proprietary data and research, conference discounts, on-demand access to event content, and more!

Join Today!


  1. Albert Lin

    I completely disagree that the ad network model isn’t profitable. From first-hand experience, I’ve been at networks (desktop and mobile) that sold conversions (purchases, app installs) and bought impressions and were extraordinarily profitable – 40-70% gross margins. Not sure what companies Goel has in mind but they must not be making good use of data.

    • Jack Kaiser

      “I don’t think any of the media arbitrage models are profitable.”

      He’s saying that the arbitrage model isn’t profitable. No shots fired at networks, from what I can see.