Description

The last time Criteo (CRTO) was posted as a long idea on VIC, it was 2016, when the share price was double the current share price, while at the same time operating profits were half of what the company generates today. Based on Google's recent announcement regarding the changes it was planning to the Chrome browser, we think the uncertainty which has depressed investor sentiment recently should have lifted.

BUSINESS DESCRIPTION

Criteo S.A. (“Criteo”) is one of the world’s largest ad tech companies.

[1]

In the fourteen years since its launch in 2005, Criteo went from a small startup in Paris, France to a truly global player with 2,700 employees across 31 offices serving over a trillion online ads and analyzing over $800 bln of sales transactions per year for almost 20,000 clients. [2]

Here is a sample of Criteo’s customer base:

Between FY2012 and FY2017, Criteo’s revenues grew at a mindboggling CAGR of 45%, mostly organically. FY2018 became its first year of no growth solely because of Apple making it much harder for Criteo to operate within its vast ecosystem. Without this massive blow, Criteo’s revenues would have probably increased by ~20% last year (more on this later).

Criteo revenues, mln USD

Criteo’s margins in its early years are rather meaningless. Revenues grew by a factor of 11x in FY2009 and by a factor of 4x in FY2010. Spending didn’t keep up with this crazy pace and margins went through the roof. In the following couple of years, revenue growth slowed down and the continued aggressive growth spending put pressure on the margins. In addition, as a privately held company, there was no pressure to deliver short-term profits. After Criteo achieved scale and went public in FY2013, its EBIT margins stabilized in the 5.5-7.0% range.

Criteo EBIT margins, %

So, what exactly does this company do? Criteo is a retargeting company. Here is what this means.

Every day, millions of people browse all kinds of e-commerce websites such as Walmart and Expedia. Some of them know what they want, others don’t. Yet another group falls somewhere in-between. They kind of want a certain good or service but haven’t committed to any action yet. One of the worst things that can happen to an e-commerce business is to have one of these indecisive customers visit the store, put something in the cart and then disappear for good.

Criteo solves this problem. It identifies these high-intent customers as they browse the Internet and feeds them highly personalized ads. Criteo’s software is embedded into thousands and thousands of websites owned by all kinds of retailers, travel agencies and publishers. Every time you visit one of these properties, Criteo gathers a small amount of information about your interests. It then combines all these tiny pieces into a much bigger mosaic of your virtual persona. Criteo does not keep track of your sensitive information such as name and home address. However, they don’t need to. With access to over 1.5 bln users and over $800 bln in sales transactions, they know enough to predict when and where to show the right ad to the right user.

Advertisers such as Booking.com tell Criteo how much incremental revenue they would like to get from their retargeting ads. This goal determines how many clicks Criteo will attempt to generate for the advertiser and how much the latter is going to pay for every click. Then, Criteo will track advertisers’ potential customers across the web and feed them retargeting ads whenever it makes economic sense. Thanks to the breadth and depth of its accumulated knowledge about millions of online shoppers, Criteo is able to purchase ad inventory on a per-impression basis, but charge advertisers on a per-click basis. The company claims that every dollar their customers spend with them, on average, results in thirteen dollars of incremental sales. They also say that they win 90% of head-to-head tests against the competition. [2] [3]

Unfortunately, I have been unable to obtain independent verification of these claims. Apparently, Criteo’s customers are happy enough with the in-house tests they run to see whether Criteo’s offering works for them. After all, conducting simple controlled experiments is not rocket science and even startups can do it these days. That said, there seems to be plenty of indirect evidence that supports Criteo’s statements. First of all, the very number of heavy-weight brands that work with them is telling. So is the 90% customer retention rate and the fact that 73% of net revenues come from uncapped budgets, i.e. clients that don’t limit Criteo’s capacity to spend money on their behalf. Secondly, Criteo had been able to deliver impressive topline growth year in year out until Apple’s decision to pull the plug on third-party cross-site tracking in CY2017. These were not easy years. Google and Facebook were 800-pound gorillas of online advertising long time before CY2017. And then there were ad-blockers, consumer privacy regulations, header-bidding technology, and other stuff. It is inconceivable to grow so fast and so profitably for multiple years in a row in such a hostile environment without a major competitive edge.

Don’t get me wrong, though. I am not discounting competition altogether. Google, Facebook and Amazon have humongous amounts of data and the technical talent to develop sophisticated algorithms – two key ingredients for a successful retargeting engine. That said, I believe Criteo has a moat in its small niche in the market. Let’s start with Google and Facebook. Both are omnipresent and collect information about users not only on their own properties but also on third-party websites with embedded Google and/or Facebook functionality such as ads, search tools and “Like” buttons. The problem is that quantity does not necessarily translate into quantity when you are trying to predict shoppers’ behavior. As I understand it, Criteo’s information is more granular – which SKUs are associated among each other, which kind of behavior when browsing an ecommerce web-site is associated with the highest probability of successful retargeting and so on. Facebook may know your favorite vacation destination, but Criteo may know what color is your favorite swimsuit. Now let’s turn to Amazon.

Unlike its two cousins, Amazon knows about the swimsuit too, at least if you bought it on Amazon. So, Criteo cannot rely on the granularity of its data alone in the fight against the Seattle-based behemoth. Conflicts of interest to the rescue! Part of Criteo’s competitive advantage stems from access to their clients’ customer relationship data. 73% of Criteo’s clients grant them access to at least some CRM data. [4] Obviously, this greatly enhances Criteo’s retargeting capabilities. Now, let’s imagine that I am an online retailer of state-of-the-art foam rollers, those cylinders people use to massage their muscles. Let’s also say that I am trying to decide who to do my retargeting business with: Amazon or Criteo? Both companies have lots of data and amazing algorithms and may look very similar to me. There is one teeny-tiny problem, though. As it turns out, Amazon is a foam roller company too!

Now, who do you think will get my business and access to my CRM data? And this problem applies to so many other verticals. Amazon is an infamous destroyer of brands. Who cares about Nike sneakers when Amazon makes it possible for people to find the same quality sneakers from a no-name company at half the price? Maybe you can even get a pair of Amazon’s private-label sneakers for a third of the price adding a few dollars to Jeff Bezos’ rainy day fund? How about online travel agencies? I bet Booking.com can’t wait to open their CRM books to Amazon!

Neither Google nor Facebook has been able to destroy Criteo with much lesser conflicts of interest and I don’t expect Amazon to be any more successful in this endeavor. Criteo’s success hinges on their ability to identify the most valuable ad impressions out of gazillions, bid enough to win the auction and keep the spread. This, in turn, depends on the strength of their algorithms, which depends on the quality of data that feeds them. It is this access to extremely valuable data and the absence of conflicts of interests that differentiates Criteo from the juggernauts of the Internet.

Before I move on to the crux of my investment thesis, I would like to talk a little bit about the special treatment that Criteo receives from thousands of publishers around the globe, which is viewed by some as the key driver of Criteo’s historical success and its key vulnerability at the moment. Here is the story. Because Criteo is really good at identifying the most valuable impressions, it is logical to assume that they win a good chunk of the auctions that they participate in. This means that they can offer major publishers a direct relationship leaving any intermediaries and their fees out of the equation – a win-win for Criteo and the publisher. At the moment, Criteo is directly connected to 3,500 publishers globally. [2] Here are some examples:

Now, some analysts appear to be concerned that Amazon will woo these publishers away from Criteo and destroy Criteo’s business model. I believe that the cause and the effect are mixed up in this line of reasoning. Publishers let Criteo take the first look at the inventory for a reason. They acknowledge that if Criteo decides to bid they will likely be the highest bidder anyway. So, why share anything with the intermediaries? This is not the reason why Criteo is competitive in the first place, but rather a nice side effect. As long as, Criteo has the best retargeting engine in the industry, Amazon is not to be feared because publishers will always be eager to maintain some kind of direct relationship with them.    

INVESTMENT OPPORTUNITY

As you can see in the chart below, Criteo has always been a very volatile stock, to say the least.

Criteo stock price, $

Here is a brief overview of the five most recent selloffs in Criteo’s history:

1. Apple ad-blocking: In Jun-2015, Apple announced that iOS 9 would support ad blocking on Safari. Ad tech stocks took a battering. Nothing bad happened to Criteo’s business, though.

2. SteelHouse lawsuit and header bidding: In Jun-2016, Criteo filed a lawsuit against its much smaller competitor SteelHouse alleging that the latter ran a “counterfeit click fraud scheme”. SteelHouse responded with a lawsuit of its own. Companies settled out of court in Nov-2016 and Criteo’s stock price rebounded. During the same period, there were concerns about the impact from the new header bidding technology, which made it possible for publishers to offer ad space to multiple exchanges simultaneously. While header bidding was, in fact, bad news for supply-side platforms, it had no discernable impact on Criteo.

3. Apple Intelligent Tracking Prevention (ITP): In Jun-2017, Apple announced its new Intelligent Tracking Prevention policy, which made it effectively impossible for Criteo and other ad tech companies to track Safari users across different websites. Initially, there were hopes that Criteo would be able to circumvent these restrictions, but over time it became clear that there was no easy workaround. Notice that during this period Criteo stock price declined by more than 50% and the company’s enterprise value fell by 65% even though less than 20% of revenues were at risk.

4. More ITP and reorganization: By Jul-2018, the full impact from ITP became evident and Criteo announced a reorganization plan. Jean-Baptiste Rudelle, the original founder of the company, returned as the CEO. A number of leadership changes followed and investments into both core and new products were ramped up.

5. Google ITP: Finally, on Mar-22-2019, Adweek exterminated the remaining few Criteo bulls with an article speculating that Google would implement its own version of ITP in Chrome.

My thesis, in short, is that the last three selloffs have been massive overreactions on the part of the market. Let me explain why.

Apple ITP

Apple’s new policy was undoubtedly a bad thing for Criteo. Roughly a sixth of their revenues were impacted and, in my opinion, their enterprise value should have decreased by the same amount or maybe a little more than that. What actually happened? Criteo lost a whopping two-thirds of its enterprise value following the announcement. The eventual impact on the bottom-line was nowhere as scary, though. Criteo posted no revenue growth in FY2018 and its EBIT margin even expanded somewhat despite the blow. To give investors partial credit, Criteo stock was on its way back to $30s when the Adweek article came out in March.

Google ITP fears

The Adweek article shaved a third off Criteo’s still recovering enterprise value in a matter of days. As a result, Criteo is now trading at all-time low valuation levels.

Obviously, investors took the Adweek article seriously. Chrome is responsible for a much larger portion of Criteo’s revenues – roughly 50% of total – and losing this much of your business is a scary scenario indeed. However, we now know that Google is not going to implement a copycat ITP solution.

During Alphabet’s 1Q19 earnings call, Sundar Pichai noted that even though “Chrome is super committed to making sure it's best in class in privacy and security … we need to be mindful of the content ecosystem and the publisher ecosystem”. [5] Shortly after this, Google announced that a new set of controls will allow Chrome users to see all of the cookies currently stored by the browser and give them the option of blocking any trackers they don’t like. “Our experience shows that people prefer ads that are personalized to their needs and interests—but only if those ads offer transparency, choice and control,” Google SVP of ads and commerce explained in his blog post. [6] [7] While specific user interface has not been presented to the public yet, we can be reasonably confident that the “nuclear” Apple-like option is off the table. In a letter to its shareholders, Criteo noted that “these changes should have an impact on our business ranging from neutral to potentially low single-digit negative”. [8] I completely agree with this assessment as the vast majority of users tend to keep default settings unchanged, which is why Google’s decision was criticized as inadequate by some. [9]

Why did they do this? The answer is simple. Unlike Apple, Google cannot join the anti-tracking club without hurting its top line. Both its own properties and the exchange business that they run for third parties greatly benefit from detailed user profiles. For instance, knowing the age and approximate income of a person searching for BMW cars helps advertisers decide which ad to show to the user: a dealership, a video game or a toy shop. Without this level of granularity online ads would be less valuable, which means less revenues for Google. At the same time, banning third-party tracking, which Criteo depends on, without making any changes to Google-owned businesses would still send the message that individualized ads are bad and most likely lead to antitrust action. After all, Google paid multibillion fines for much smaller offenses.

Why does the opportunity exist?

As I noted earlier in this report, $CRTO was on its way to $30 per share when the Adweek article came out. Sell-side updates that came out after this news confirm that investors were concerned about changes to Chrome settings more than anything else. Why didn’t the stock price rebound after Google announced its decision? I think that the reason is twofold. Firstly, investors who sold the stock following the Adweek article may not feel comfortable making a U-turn six weeks after the decision to sell. Analysts may get uncomfortable questions from portfolio managers and portfolio managers may get uncomfortable questions from clients. Sell-side has been surprisingly quiet about the Google news too. For some, it may be important to be consistent than rational. Secondly, Criteo provided what may be perceived as ugly 2Q19 and full-year guidance during its earnings call that took place just a few days before the Google I/O conference. 2Q19 revenues ex-TAC are expected to be down 1% YoY in constant currency terms and full-year revenues ex-TAC are expected to grow by 1.0% YoY instead of 4.5% YoY.  This is attributable to Criteo’s customers being somewhat slow to adopt in-app retargeting at a time when consumers are spending increasing amounts of time away from desktop computers. The in-app business still grew nicely (up 32% YoY), but did not keep up with consumers and management expectations (up 63% YoY). In addition, sales function optimization and the new self-serve retargeting offering for small and medium enterprises are taking more time than originally expected. Even though these headwinds are temporary in nature, they precipitated multiple downgrades and caused the consensus target price to fall from $30 to $25. As a result, the stock may look uglier to quant investors than it really is (especially after the I/O conference).

VALUATION

At the moment, Criteo is trading at ~10x forward earnings ex-cash, which is very cheap for a high-ROIC growth business with a large net cash position and significant insider ownership. Using any kind of historical multiples would result in tremendous amount of upside. In addition, historical multiples are not representative of Criteo’s future prospects anyway. On the one hand, Criteo was experiencing multiple perceived existential threats throughout its short history as a public company, which means historical multiples could have been higher. On the other hand, Criteo’s growth prospects are currently nowhere close to what they used to be in the past, which means today’s multiples should be lower. To keep it simple, I went with a DCF model. Here are the assumptions:

• No revenue growth and slightly below-average margins in 2019.
  
• 2020-24 revenue CAGR of 8% assuming mid-single-digit growth in the core retargeting service, i.e. slightly worse than the broader industry – digital banner revenues were up 22% in 2018 [10], but are slowing down to low-teens – with the rest coming from new initiatives, including the recently acquired HookLogic business and the new customer acquisition service. The latter is based on Criteo’s existing retargeting engine but will be used to identify high-value users who have never been to the customer’s web-site before.
  
• Long-term EBIT margin of 6.0%, which is slightly below the 5-year average of 6.2%.
  
• Terminal ex-cash P/E multiple of 16x in-line with historical market averages, which is not too demanding for a company with above-average ROICs, growth rates, balance sheet and insider ownership.
• WACC of 10%.
  
• The resulting target price is $34, which is 70% above today’s price.

REFERENCES

[1] IDC Worldwide Digital Advertising Software Market Shares, 2017

[2] $CRTO (Feb-2019) Criteo 101 presentation

[3] $CRTO: How to boost online sales with Criteo Dynamic Retargeting

https://www.youtube.com/watch?v=ANR6po62_sA

[4] $CRTO FY2018 10-K

[5] $GOOGL 1Q19 transcript

[6] https://www.theverge.com/2019/5/7/18535532/google-chrome-ad-tracking-cookie-blocking-io-2019

[7] https://blog.google/products/ads/transparency-choice-and-control-digital-advertising/

[8] $CRTO (May-8-2019) 8-K

[9] https://www.vox.com/recode/2019/5/9/18537250/google-privacy-tracking-cookies

[10] IAB internet advertising revenue report 2018 full year results

https://www.iab.com/wp-content/uploads/2019/05/Full-Year-2018-IAB-Internet-Advertising-Revenue-Report.pdf

Catalyst

• Criteo should be back to healthy growth once more of its customers adopt in-app retargeting, which is just a matter of time.

• Their new automated onboarding technology should bring in small and medium-sized accounts, which historically had not been covered by the sales team.

• At some point in time sell-side analysts will start updating their views to reflect Google decision not to follow Apple’s lead with respect to tracking prevention.