INside Performance Marketing
Top 50 Industry Players from 2017
Q&A: Criteo's pink pony

Q&A: Criteo's pink pony


What’s your view of the display market right now?

Let me talk about the market first. Just the general market. As you know, I was almost eight years at Yahoo and my job was running the worldwide sales there. So I had a very good view of the digital landscape, the digital inventory, the CPMs and the branded world. I left Yahoo in 2008. In the four years since there’s been lots of change that you’ve been seeing, that I’ve been seeing. The biggest one is there’s so much more inventory on the market. I can’t count that high, there’s so much more. It’s just huge!

Marketers are looking for great return on investment. The brand marketers are still going to be there. The brand marketers are still looking for the best contextual placement of their advertising, but that’s starting to shrink a bit as the inventory continues to grow. One of the things I’m seeing is there’s real pressure across the internet on CPMs and on the publishers that have premium inventory. When you have supply increasing and demand staying the same - they taught me in business school that prices tend to go down.

If you have the chance to take a look at the big portals now and look at the home page, the home pages are filled with direct response advertising now. There’s still a bunch of premium, but that was the best price that will hold on. So demand for that to do other things has increased.

I have friends that are in the branded world of creating content and selling advertising. There’s more pressure on them to deliver better yields and better CPMs when the market is pushing against them. So as a general rule in the display space, that’s kind of a high level macro view I’d say.

What we’re seeing in the performance space is that clients are starting to understand that display advertising can actually be utilised as a performance vehicle. This was not the case when I was at Yahoo. People would spend a lot of money on untargeted advertising, what we call tier 2, and hoping to get a return on investment that they could measure, that they could say the return was positive.

We’re now getting a lot more marketers that are starting to see some products in the marketplace that’s creating an even greater appetite. They’re saying if I give you some money and I want to run a display ad, I want to have a great positive return on my spend. Of course I want to get some branding in that as well. We’re seeing this.

The reason I joined Criteo a year and a half ago was because I was watching these macro trends. The thing that I found unbelievable when I first met Criteo was that they had created a product that allowed display ads to perform like search. I thought it was an exaggeration. I didn’t think it was possible.

Before I joined, I had the good fortune of being very close friends with two of the board members in Europe. I was on the board of advisors of Criteo back then. They showed me the numbers. They showed me renewal rates that were 96%. They showed me that the marketers were running 70% uncapped. I made personal threats to their lives if they were lying to me, but I found something that I didn’t know existed.

Today what I’m seeing now is that there’s really a race around the science behind making display ads perform. At Criteo we have almost 300 engineers. To make a display ad perform you can’t take it lightly. You have ad networks that try to do some of that, but they have three people in technology, in R&D; and 150 people in sales marketing.

Is that what sets Criteo apart, the volume of its engineers?

GC: The numbers are important, but the mission that they have, the leadership that they have. It’s the quality of people and the quantity that we need to do what we need to do. So we’re building a vision for the many years to come. As a company, as a private company, we don’t share our internal figures, but the amount of money that we reinvest in our engine would shock people. It’s exactly what we should be doing.

The technology allows us to grow everywhere. I call our technology a pony and our engineers like it. I call it our pink pony that we have built. I get the chance to take a cord from that pony in the United States and plug it into all these advertisers and they blow up and become huge success stories.

We’re doing it in Japan now. We’re going to do it in China soon. We’re doing it in 32 different countries. Having the technology engine in a central location in Paris allows us to do magic around the world. The scalability of what those 300 engineers do is why I came to this company. It’s the only reason I came because they were committed to investing in the greater products that without question deliver what a marketer wants: a positive return on ad spend that they can measure and that they can increase their spending. Here at Criteo you can see [our slogan], Expand Your Search. We think search is great, but we think search plus performance display is the best. We think that we can augment.

What are some of your engineers' more notable achievements?

GC: A lot of what we do is always going on, always iterative. There are some things that I just can’t talk about, but we have a number of different releases that go on through the year. Usually two big releases a year.

Without doing anything different, a client will come to us and say, “Whoah! My performance just increased. What’s going on?” It’s the constant work the engineers are doing in a number of different areas that are allowing us to do this magic. So it’s part of our secret source. It’s part of what we don’t talk about. Even if we did talk about it, it would be impossible or very difficult for people to copy.

Criteo started as a recommendation engine and one of our real competitive advantages is that a marketer - so if you’re Priceminister, Macies in the United States, Travelocity, Expedia. As they’re looking to grow, they also have to look at how many places Criteo is buying from.

We now have over 5,000 publishers in our network. We spend a lot of money in real time bidding with all the exchanges, but again what sets us apart is our ability to buy directly from the publishers in addition to the exchanges. 5,000 publishers worldwide is a competitive advantage. Going back to the engine, if you’re a marketer and you’re selling red shoes as an example. When we see they’re shopping for red shoes, we’ll serve a red shoe as a retargeted ad. When they click on that red shoe ad we generally serve three, four or five other ads: maybe a hat, maybe a scarf, maybe a coat. Based on what our engine knows. What else they’re looking for. Most of the clicks and most of the merchandise people buy are not the red shoes; they’re the other items we serve up.

Having a recommendation engine at the heart of our company has given us a huge competitive advantage. While they’re still buying red shoes, they’re buying mostly other products. If you’re Macies, you don’t care. You’re looking to sell merchandise, not just red shoes.

In the coming years, where is display’s innovation going to come from?

GC: I believe the market will be looking for the display world, and we’re taking a leadership role here, to help the top of the funnel. Can there be better ways to track, in a more performance-like fashion, a new customer to Macies, a new customer to Marks & Spencer? What retargeting does really well is it brings in those incremental guys that are people in the market right now, but most of the advertisers budgets are trying to get new people to come and shop with them. Not so easy.

At Criteo, a big part of our product game plan is to be able to go to an advertiser and tell them we do a great job with retargeting at the bottom of the funnel, but we now have a pure CPC-based product that will deliver new customers to you. Our vision a year from now or two years from now, it’s to be able to say to a customer, “give us your entire display budget.” We can satisfy the bottom and the top in a way that you can clearly measure the efficacy of the campaign.

What I will tell you about the top of the funnel now. It’s pretty interesting because here’s how we look at it. We’re in the laboratory. So it’s all like we have white lab coats on. We’re experimenting. We’re getting some of the biggest and most important advertisers to come into our lab, to say here’s our vision. We can’t do it on a whiteboard. We need real live clients. The level of interest in cracking the code for upper funnel and performance is huge. We’ve created a great environment where the expectations are low with our marketing partners, but they all want to come into the lab to see if they can help us crack that.

From an innovation standpoint, having display that’s always been perceived as top of the funnel for brand awareness, that’s still going to be important. That’s not going to go away, but if you can do it with a defined performance element. That’s the Holy Grail and it’s not easy.

How do you think consumer perceptions of display have evolved?

GC: Let me answer it also from the macro standpoint. So today, click through rate is proxy for engagement. The average banner ad according to the IAB, clicks at 0.06% across the internet. 0.06% clicked on an ad. At Criteo our ads click at 0.7% so ten times greater. What it says if you can show the right ad at the right time to the right person you’re going to see interest.

Most of my friends say they don’t even notice. There’s ad blindness. I can’t disagree with them, but they do see the ads they just don’t remember it. I think you’re going to see a marketplace evolve where more and more consumers will want to opt in for certain things that are of interest to them and opt out for things that not interesting to them. If you’re looking five years from now, that’s probably where the market will go.

Today there have been lots of innovations around opt-in. Still, the scale is tiny to get people to opt in. It’s still pretty difficult to do that. I think with all the initiatives that are going on today, you’ll see much more about that.

So if I knew I was interested in buying a new trench coat there should be a way for me to say I’m in the market without just me going to the website and shopping. There have to be different ways to demonstrate intent where I want to see the offers and new products out there, but until then, the cookie-based targeting and cookie-based tracking will be very important.

I think it can be done with current cookie system. The consumer could also become, and is becoming, far more educated about what they can and can’t do. We’re finding that very interesting to watch. The consumer will lead the trend. They’re going to demand. I’m looking for a trench coat. I want to see this.

Right now, from a desirability point of view, people are clicking on ads with us if they’re in the market for the trench coat and you’re serving up different trench coat ads. It’s a good thing because they’re in the market, but if we were serving up trench coats after they’d bought a trench coat it would be a pain in the neck. I think it can be designed around what’s happening today.

Continue the conversation

Got a question or comment – tweet Simon simonnholland or comment on Twitter, Facebook or LinkedIN.

Simon Holland

Simon Holland

Simon is the news and research reporter at Existem. Previously a technology journalist, he now spends his time investigating both future and developing trends in performance marketing whilst producing editorial content for

Read more from Simon

You may also like…