October 4, 2019

Taboola & Outbrain Joining Forces

Taboola & Outbrain, the largest content recommendation companies, are finally joining forces. The two firms had been circling each other for years, but are finally pulling the trigger.

It shouldn’t come as a surprise that we’re in a period of consolidation and it’s not just on the publisher side. Taboola & Outbrain have been trying to outcompete and out pay each other to accumulate marketshare. But ultimately, they viewed the real threat as Facebook, Google and Amazon—not each other.

Both companies are reportedly profitable. Outbrain shareholders are getting $250 million in cash plus 30% of the combined entity. The name will remain Taboola and Taboola’s CEO will be leading the operation.

Unfortunately for publishers, this merger is the first step in a slow reduction in the amount of revenue that these sources will provide. That’s where I want to spend time today.

Outbrain and Taboola launched over a decade ago a year apart from each other in Israel. The original goal, I can only imagine, was to give user’s the ability to find new, interesting content while also paying publishers for the exposure.

Unfortunately, the product devolved into what many in the industry now call the “Chum Box.” Because they operate on a cost-per-click basis, the headlines became more audacious with images to go along with them.

For example, here’s an ad I found on The Washington Post, which uses Outbrain:


The problem is this comes at the end of an article with the headline: “Trump wanted Ukraine’s president to launch investigations before face-to-face meeting, State Dept. texts show.”

A user starts their scroll by reading very important information about politics in the United States and then ends with that ad. You would think a reputable news organization wouldn’t want that on their site.

And yet, these ads are everywhere. Bloomberg, The Washington Post, CNBC, you name it, a publisher likely has it. And it’s not just Taboola and Outbrain. There’s Revcontent, ZergNet, Dianomi, media.net, and the list goes on. They’re all hawking this chum and publishers are eating it up.

It makes sense why, though.

With Taboola and Outbrain, in particular, going head-to-head for the best publishers, they started forking over large guarantees in exchange for exclusivity. I’ve heard stories of publishers getting deals worth over seven figures per month for running the ad units. In an era where publishers were desperate to get revenue in the door, I can see why they ran these units.

We used to run units like this too. And although our guarantees were lower, the money was really good. Everyone hated the units, but it’s hard to turn down that kind of money when you’re trying to build diversified streams of revenue.

The talking points out of Taboola/Outbrain are that this merger will provide more scale for advertisers to convince them to bring more budget over which in turn means that publishers will make more money. It’s possible that more advertisers will come over because of the combined scale making it easy to buy, but that doesn’t inherently mean that publishers are going to make more money.


It used to be that if Taboola was threatening to cut your guarantee, you could just call up Outbrain and they’d give you something similar. And if they didn’t, even the threat of Outbrain doing something would be enough to keep Taboola engaged.

With Taboola and Outbrain now Taboolabrain, publishers can no longer pit them against each other. I imagine this is was a reason why they merged: they were tired of the constant back and forth.

That’s not to say publishers weren’t already beginning to question whether guarantees made sense. In a Digiday story in 2018:

Publishers that spoke on background for this story because they were in the midst of negotiating said either that the content recommendation companies themselves weren’t offering guaranteed payments anymore or that the publishers themselves were opting for a revenue-share model where they’d be paid a share of the revenue when visitors click on a paid ad in the widget. One exec at a digital lifestyle publisher used to get a $1 million-plus guarantee a year and gave it up for a rev share with Taboola.

“When we looked at the stipulations, we decided we wanted flexibility on placement, and a guarantee doesn’t let you do that,” said the exec.

For those publishers that do rely on guarantees, I expect them to get smaller. For those that don’t rely on guarantees, but opted for a revenue share, it’s still likely that revenue will drop for them. Now that the two former-foes have combined, maybe a 70/30 split becomes a 65/35 split. And after a few years, publishers might be looking at a 50/50 split.


When we think about a service like Taboolabrain, we have to understand the complete supply chain to understand where the money is coming from. So, let’s make a theoretical one up.

FormerStarClickbait123.com is my fake website. And I am putting up slideshows about how on former stars have had their careers come crash to the ground. All around the slide show, I put up as many ads as I can fit. Every fourth slide, I inject another ad. The goal is to get a user to click to the next slide so that all the ads refresh and the cash register can go off in my mind.

Here’s some hypothetical math… Across all the ads on a page, I figure out how to get a $10 RPM.

Now that my site is ready, I go to Taboolabrain and buy my ads. Knowing that my page-level RPM is $10, if I were a single-page publisher, I could only bid $0.01 per click. But that’s okay. There are 50 slides. So, I bid $0.20 a click. So long as I can get a user to visit more than 20 slides, I am profitable.

I’m not the only one doing this, though. So competition pushes up the price and maybe I have to bid $0.45 per click. I am only making $0.05 per user, but when you’re talking about millions of users, the money really starts to add up.

This is basic arbitrage… And it’s made it impossible for publishers of legitimate content to use these services to promote their own content. I’ve looked into it as a way to promote our very legit content. It just doesn’t work unless you really pack on the ads.

The thing is, this arbitrage is all predicated on the amount of money that I can make at the page level. In the above scenario, I’m making a $10 RPM. So, break even is one user hitting one additional page for each additional penny that I bid.

The reality, though, is that FormerStarClickbait123.com has no direct deals. The only money I am getting comes from cookie-based programmatic advertising. And this is what I feel is the biggest risk to Taboolabrain and, in turn, publishers that depend on this revenue.

As cookie-based advertising slowly dies, the amount of revenue that I can earn as the owner of FormerStarClickbait123.com drops. Publishers say they saw a 25% drop in revenue with Safari and Firefox blocking cookies, so let’s say that happens to my fake site. Now I am only earning a $7.50 RPM. If my bid stays constant, I have to add more slides to my content to recoup the investment. What’s more likely to occur, though, is that the amount I can bid drops.

That’s the supply chain in a nut shell. If the buyer starts to get margins crunched, the seller is going to have the same problem.

[Update: If you want a much more in depth read on how the arbitrage of this works, check out this piece over The Margins. So much more in depth and insane to read.

Revenue will drop

Both companies will likely disagree with this, but it’s hard to see it going any other way. Competition among ad companies is good for publishers. It means that the publisher earns more money because they get to play the middle. As these ad companies merge, margins start to improve for them while publishers see less revenue.

And even if that didn’t happen, the business models are flawed. If the way a premium publisher makes money is if someone clicks over to FormerStarClickbait123.com and the only way that site makes money is with crappy cookie-based programmatic advertising, what’s the outcome when cookies continue to die and revenue dries up for my crappy site?

Taboolabrain will continue to make amazing money, will likely IPO in the next couple of years, and could make the investors a lot of money. But for publishers, I expect to see revenue drop.

My one ask…

The one thing I do hope is that the combined companies keep Outbrain’s Sphere product. I’ve been investigating this for some time now and I think it’s a great opportunity for publishers that are looking to do smart paid audience acquisition.

In the above scenario where I am bidding $0.45 per click, there’s no way a legitimate single-page publisher can compete. That would require a RPM of $45.

But the Sphere product is unique in that its only for premium publishers and the cost really isn’t that bad. I pulled this image from Outbrain’s site:


Let’s use two real Sphere partners: CNN & Skift.

A user sees a Skift ad on CNN and clicks it. Skift pays $0.015 to CNN. All of that money is CNN’s. If the user then clicks over to a second page on Skift, Outbrain charges $0.015 to Skift. In total, Skift has paid $0.03 for a two pageview user. If the user goes to four pageviews, Skift still only pays $0.03. Skift would also have a unit on their site, so while they’re paying, they are also potentially earning.

I really like this model because it means that publishers are not competing with the fake sites and direct response marketers that can bid much higher than most publishers can.

I am hopeful that Outbrain keeps this product because I actually really see it as a viable model for publishers. As publishers move more aggressively into registrations and subscriptions, the LTV of a user should improve which should, in turn, make this a much stronger paid audience development channel.

We’ll have to wait and see what they decide to do with the product, though. I’m hopeful.