Rubicon Project CEO Does The Math On Its Low-Fee Future

Rubicon’s CEO Michael Barrett will speak at AdExchanger’s upcoming Industry Preview conference on January 17 – 18, 2018 at the Grand Hyatt New York. 

Rubicon Project is in the midst of a risky move – one that CEO Michael Barrett thought he could avoid when he joined nine months ago.

Because of customer outcry, the company stopped charging nontransparent buy-side fees last quarter, which supplied half its take rate.

But this move decreased its take rate into the 10-12% range, well below what it needs to keep its business in the black.

Now, Rubicon must use its cash on hand to fund the lowered take rates until, hopefully, it can make up the difference by acquiring more clients. But that cash won’t last forever.

Rubicon once had hundreds of millions in cash, but during its Q3 earnings it had $99 million. That’s still more than its current market cap of $88 million.

Originally, Barrett thought he could lower take rates while having buyers and sellers pay equal amounts. But that proved unfeasible.

“I’ve been on record saying buyers should pay half, and sellers pay half. Buyers feel differently. And that’s fair,” Barrett said. “When you talk to folks behind the DSPs and agency holding companies, the concept of paying for access to media is so foreign that the idea of trying to go through an education process didn’t seem worth the effort.”

Rubicon is betting its lower, transparent fees will help it gain market share.

Since it halved its take rate, it must double the media spend on its platform just to end up with the same amount flowing in.

“If you were a competitor to Rubicon, all you had to do is slide our S-1 or 10-Q across the table and say, ‘Look at their total take rate,’” Barrett said. “We hope the same thing occurs now, because we are as low as you can get. Pricing is going to be part of our competitive advantage.”

Barrett talked to AdExchanger about his first nine months at Rubicon.

AdExchanger: Originally, you said Rubicon’s take rate would go down to the 15% to 20% range. But on the last earnings call, you revised that down to 10% to 12%. How did you decide on that range?

MICHAEL BARRETT: That was the way the math worked out in our case. If you eliminated buy-side fees, [10-12%] was where you got to.

If you’re running an organization [with buy-side fees], you tend to discount your fees on the sell side knowing in the end you can make it back up. That number can drift up slightly over time. The idea now is to run as pure sell-side only, and give buyers and sellers complete transparency.

The Trade Desk doesn’t have this reporting problem, because it’s not publicly disclosing its take rate every quarter. If you could go back in time, would you tell the CFO, “Don’t report this”?

Maybe they learned from us. The tricky thing is, I’m sure somewhere along the line the CFO at the time thought that we needed to combat an argument out there [among investors] that “The take rate is going to zero,” and he probably thought shining a light on it would placate investors and make them feel like it’s a good margin business.

The problem is once you do it, unless you want to restate [earnings], you’re committed to that number. It’s been a negative bull’s-eye for a long time.

Google AdX has a reputation for being a stickler on its fee, often quoted around 20%. So they don’t have that problem either. 

Rubicon had a really nice business for several years there. We had a nice spot in the waterfall – anything that Google didn’t monetize came to us. But it was always capped. You had someone playing with the inventory before you had a chance to see it.

Now we have equal access to inventory, and that calls for a different economic structure. The total addressable market is so much bigger than it was in the waterfall, pre-header day. It’s a challenging trade-off short-term. But long-term, it’s an enormously attractive business model.

Rubicon may now have a low take rate, but companies like Amazon, or Google, are in the position to offer tech as a loss leader, because they make money on media. How do you compete with free?

The one thing that gives us any semblance of comfort is the idea that Google, the most powerful advertising company on the face of the planet, has not been able to aggregate enough of their own demand to satisfy a publisher’s complete monetization strategy. If that were the case, every publisher would just be using AdX.

Amazon certainly can’t harness enough demand. It’s very unique demand and valued by publishers, but it’s a slice of what they need to totally monetize their site. We think a 10% fee is right.

Google’s been around quite some time and hasn’t budged on that 20% number, and the idea that they’d go from 20% to zero is a bit far-fetched. Even if they were, it’s not enough demand to fill a publisher’s complete need.

Having been through the transparency and fee wringer on the SSP side, what advice do you have for DSPs that might also have hidden fees?

The game is over. When you have the world’s largest marketers doing complete audits from the minute the dollar enters the programmatic universe and cross-checking with the publishers, it’s inconceivable that kickbacks and undisclosed fees could be part of the business model going forward.

Marketers and agencies are going to know their total cost, and it’s not going to be a matter of years, it’s going to be a matter of quarters.

AppNexus CEO Brian O’Kelley has declared a price war. Do you feel you’re under attack?

No. If anyone is waging a price war, it’s us. I know Brian has been quoting some low prices, but if you were to call him up as a regular publisher, you’d be hard-pressed to get that 8.5% rate.

If you are a large enterprise client, like Microsoft, you could get that price. Brian is an engineer at heart, but he has a heavy dose of marketing in there.

Transparent, low pricing was the right thing for us. I don’t think you have to be 100 basis points below your competitor. It’s not like pizza stores in Manhattan.

You could have the lowest cost from fees, but run first-price instead of second-price auctions and make a killing. We want the lowest total cost, which is a combination of total take rate and clear auction dynamics.

That seems like a summary of what AppNexus is doing: having a lower fee but using first-price auctions to make up for that lower fee.

That sounds too nefarious. My example is not that first-pricing is bad, it’s that first-pricing in a second-price auction is bad. I think AppNexus is very clear. There are folks out there that say they run a second-price auction that are first pricing, and that’s a way to really make a lot of money.

Do you think that these fees are sustainable for Rubicon?

For the short term, we chose to invest our capital in the business, which is a fancy way of saying we will be running at losses for a while. Our investors by and large have leaned into the strategy that [subsidizing lower fees] is a good use of our capital.

Our models predicate that this can be sustainable, but it’s a high-volume business. You don’t have to be an MBA to realize that if you halve the fee you charge, and don’t double your inventory, it’s a tough path.

This interview has been edited and condensed.

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 Comment

  1. Michael Barett said, “I know Brian has been quoting some low prices, but if you were to call him up as a regular publisher, you’d be hard-pressed to get that 8.5% rate.”

    He is 100% correct, most “regular” publishers are in the 10-15% range with Appexus.
    However, if you are a regular publisher who calls up Rubicon or have been a long term client, you are not getting lower than 15-20%.

    Mean while, Sorn and Index Exchange are laughing all the way to the bank with their 30% take rates..