Criteo S.A. Q2 FY2020 Earnings Call

Good morning and welcome to the Criteo Second Quarter 2020 Earnings Call. All participants will be in listen-only mode. Please note this event is being recorded. I would now like to turn the conference over to Edouard Lassalle, VP of Head of Investor and Analyst Relations. Please go ahead.

Thank you, Kyle. Good morning everyone, and welcome to Criteo's second quarter 2020 earnings call. We hope you all are safe and healthy wherever you are. With us today are CEO, Megan Clarken; and CFO, Dave Anderson. Please note that because of the ongoing restrictions, we're all joining this call from different locations today and as a result may face unwanted technical challenges. In the course of our call, management will make certain forward-looking statements. These forward-looking statements reflect clear judgment and analysis only as of today. And actual results may differ materially from current expectations based on a number of factors affecting Criteo's business. Importantly, at this time, the global COVID-19 pandemic is still having a significant impact on the global economy, the business of our clients, as well as on Criteo's business, and may further impact Criteo's financial condition, results of operations, and cash flows in the future. There are significant uncertainties around the duration and extent of the pandemic. The dynamic nature of these circumstances may impact what we say on this call today, and those expectations could materially change at any time. For more information, please refer to the risk factors discussed in our earnings release, as well as the most recent Form 10-K and Form 10-Q filed with the SEC. We do not undertake any obligations to update any forward-looking statements discussed today, except as required by law. In addition, during this call, we will also discuss non-GAAP measures of our performance. Definitions of such metrics and the reconciliations to the most directly comparable GAAP financial measures are included in the earnings release published on our website earlier today. Finally, unless otherwise stated, all growth comparisons made during this call are against the same period in the prior year. With that, it is my pleasure to now hand it over to Megan.

Thank you, Edward. Good morning everybody. It is a great pleasure to be with all of you for our third call together. Once again, I'm joining you from the safety of my house in upstate New York, and I hope everyone is staying safe as well. On our call today, I'm going to cover five key topics. Number one, the current business context and our learnings from it. Number two, better than expected results in Q2. Number three, a brief recap of our strategy, including executing on our four strategic pillars. Number four, a progress on online identification. Number five, our operational priorities going forward. Now, let's start with the current business context. In the unusual times that we're living in, the world is changing fast and some really interesting trends are shaping up. Global retailer sales are shifting to e-commerce. Based on eMarket data, e-commerce is predicted to grow from 16% of total retail sales today to 23% in 2023, heavily influenced by COVID consumption trends. In the United States, the growth is partly driven by an increase in click and collect, allowing consumers to make immediate contact with purchases. This strong growth benefits many non-large platform retailers and direct-to-consumer brands, particularly in the mid-market. For example, according to a survey, 81% of e-commerce purchasers who tried new retailers during COVID expect to continue to use them in the future. And marketers are making moves to redirect the sizable amount of their trade budget to media and advertising, injecting ad spend into the market and towards lower funnel campaigns, meaning targeting and retargeting. All of this clearly is very encouraging for our business on all fronts. Let's move to quarter two and our performance. Revenue ex-TAC of $180 million and adjusted EBITDA of $39 million were respectively 21% and $30 million above the high end of our initial guidance at constant currency. We achieved these solid results thanks to the contribution of all Criteo's who have not missed a beat during the lockdown and work-from-home situation. I'm really proud, incredibly proud of the team and our performance. In line with the Q1 trends, our mid-market business remains resilient and continues to grow, especially in retail, thanks to healthy online retailers and direct-to-consumer brands. Our number of large customers in travel, especially in brick-and-mortar retail, remain deeply affected by COVID. We continue to sign new accounts in both large and mid-market segments and expect growing contribution from new business in the second half. As the economic aftermath of the pandemic progressively unfolds, we believe direct response and targeted marketing remain key to help all clients weather the pandemic and accelerate their recovery. For our Retail Media business specifically, same client revenue ex-TAC growth was up close to 60% in Q2, a strong acceleration from prior orders. In Q2, we also signed a new agreement with one of the very top U.S. retailers. While higher e-commerce usage triggered by COVID was a clear positive for Retail Media in Q2, accelerating the move of brand dollars online, we believe this tailwind is going to continue albeit at a slower pace in Q3 and beyond as online consumption moves beyond the lockdown period peak. I want to say a few words about COVID. The net impact of COVID on our business was lower than expected in Q2 and proves to be positive in Retail Media and mid-market. Yet, the pandemic continues to be dynamic and business conditions remain incredibly fluid, particularly in the United States. Our current view is that the pace of recovery may be much slower than originally anticipated, particularly for the travel and classified verticals. As a result, we may not see full recovery before sometime in 2021 at the earliest. And Dave will talk more about this later. Now, moving to the third point of strategy, as we've discussed in the past, our mission is to power the world's marketers with trusted and impactful advertising. In order to leverage the key trends I spoke about, our strategy is to build and operate a world-class demand-side platform or DSP specifically for e-commerce. With this, we want to earn additional marketing budgets from existing clients and win new customers in the consumer brand and commerce space. To succeed with this strategy, we are drawing on our strong and differentiated asset: our deep relationships with commerce clients and retailers, our rich commerce data, and large ID graph, as well as our fast-growing Retail Media platforms for brands and retailers. Given the short and long-term growth of e-commerce and what we believe the long-lasting effects of COVID could be, our position in the space provides an exciting opportunity. Executing on this front, we focus on four strategic pillars. Firstly, strengthen the core. Secondly, expand the portfolio. Thirdly, explore strategic game changers. And last but not least, drive tech and operational excellence. With respect to strengthening the core, we continue to build our platform and release DSP features as they become available. For example, we now offer enhanced commerce audience targeting, allowing marketers to connect more product categories and brands into powerful additional audience segments. Our new reporting dashboards now provide marketers with additional transparency on the performance and measurement of their campaigns. Transparency is a critical requirement for marketers. Responding to our clients’ retargeting needs while pushing up the funnel to increase our capabilities beyond retargeting is our focus here. As a matter of fact, global research firm IDC recently named Criteo the number one independent ad tech company for 2019. With respect to our second pillar, expanding our portfolio, I'm pleased to report that our new solutions now represent 20% of our business on a revenue ex-TAC basis, growing 67%. In particular, the momentum around Retail Media accelerated to plus 84%, growing more than twice as fast as in Q1. Retail Media, as you know, helps brands reach critical target consumers on retailers’ websites and apps. In June, we launched our Unified Retail Media platform in the U.S. This platform combines various Retail Media offerings, including sponsored products and other rich ad formats, and provides self-service transparency and control that they need as they manage and monitor their marketing campaigns. We plan to launch the platform in Europe in Q4. Recently, we hired the prior CEO of Triad Retail Media, Sherry Smith, to join our team and lead our U.S. Retail Media business. Sherry is a terrific addition to our team. Though other new solutions continue to perform strongly. For example, our omnichannel business, helping clients target and engage their offline or in-store customers online, grew more than 120%. We expect this to continue to grow nicely even after we get out of the pandemic. And our app consideration solutions continue to grow very nicely as lockdown restrictions drive more user time on apps, driving more installs, more traffic, and more customer acquisition on apps for our clients. Regarding strategic game changers, we continue to grow our partner network. Partnering with Yahoo! Japan since 2012, we extended our agreement to now include our consideration solutions, including web traffic and app install, enabling marketers to display ads for additional marketing objectives on Yahoo Japan's inventory. We also just signed a partnership with customer data platform Zaius to strengthen our custom audience capabilities for online and in-store targeting. Additionally, we broadened our partnership with Lengow, enabling retailers to target offline customers with more relevant product recommendations based on what's available in local stores. We expect these two partnerships will further enhance our audience targeting success and support our focus towards e-commerce. Lastly, with respect to driving operational excellence, as discussed earlier for enhancing our platform with new DSP features, we're also working on strengthening our contextual targeting capabilities, building on our partnership with Oracle Data Cloud and enriching the Criteo engine to enable bidding on contextual signals. We continue to invest in our identity solutions. I'll speak more about this in a second. I'm also thrilled by the new talent we're bringing into Criteo, strengthening the caliber of our leadership team. Most recently, we announced Dave Anderson's arrival. Those who know Dave know that he brings experience, transparency, and integrity. He's a hard-working trusted partner, and I'm thrilled to have him on board. David Fox officially extended his role to Chief Commercial Officer, working tirelessly to bring sales operations excellence and accountability to our commercial teams. I'm also delighted to announce that we've just hired a world-class Chief Product Officer named Todd Parsons, joining us from OpenX. The Chief Product Officer role is critical for us, and Todd brings a proven track record as a product leader and deep expertise in advertising technology. I'm energized by the team we've built in a short time—fired up and ready to go. Regarding our cost containment program, we exceeded our plan in Q2 and remain well on track with the targets that we shared with you. As we said in our Q1 call, we intend to implement further cost control and organizational measures to right-size the business while maintaining investment in growth areas. It gets me to my fourth topic, our focus on identity. We see changes to online identity presenting a step forward for consumer privacy. However, we stress the importance of industry collaboration in thinking through the implications of some of the proposed changes on the wider ecosystem and the consumer experience. Remember, privacy challenges, including changes to Chrome, affect the entire ecosystem, whether that be ad targeting, supply monetization, measurement, or attribution. For our part, we see this moment as an opportunity to apply our unique assets and work for the open Internet, the advertisers, and ultimately, the consumers. We believe we have assets that no other tech company has, and we’re stepping up to use them. As our CTO recently put it, online identity may go dark at some point, but we have a powerful flashlight. Our powerful flashlight is built at scale on three capabilities. First, our ability to do contextual and cohort-based advertising based on user profiles built on context and affinity. Secondly, one-to-one advertising based on consumer identification data coming from first-party integrations with over 20,000 advertisers and 4,700 direct publishers. And thirdly, adding to our first two capabilities, our massive ID Graph with 2.5 billion unique users globally, of which 98% have persistent identifiers beyond cookies—a critical asset for online identification that nobody else has. Enabling marketers to deliver the right message to the right person at the right time and on the right device, whether in-app or CTV, is predicated on our ability to best understand anonymized consumer identification and their shopping interests. Our three capabilities combined provide a massive headlight and will continue to set us apart as we move forward. With our expertise and confidence, we created our response to Google’s privacy sandbox on behalf of the industry. We call this SPARROW, or Secure Private Advertising Remotely Run on Webserver. Criteo is the only ad tech company, alongside Facebook, to have progressed its proposal to Google Chrome from a W3C group to a more formal group called Web Platform Incubator Community. We propose methods related to consumer experience, control, and continuity of the ecosystem while improving performance for advertisers. This work is ongoing, and we recently published an extension to our original proposal to add near-real-time reporting capabilities, which received positive industry feedback. In fact, the CTO of a large well-respected supply-side platform recently praised SPARROW and called out Criteo for creating good harmony across the working group. Most recently, Adweek called Criteo the number one retargeting company within the 15 most innovative and relevant companies in ad tech and martech. Overall, our work around identity is deep and wide, building both inside and outside the sandbox and extending beyond Criteo to the industry and consumer experience. As you probably recall, Apple announced that with iOS14 expected later in Q3, all apps will need to obtain user permission before leveraging its anonymized identifier for advertising. We believe this is a move in the right direction, based on respect for user choice and control. Yet, we also believe this can be a significant industry issue. This unilateral decision, with little time to adapt, will create troubles with deep implications for the advertising ecosystem, especially on the supply side. Both advertisers and publishers are looking to us to help address this change. Here again, we believe we have a powerful flashlight. As I said on our last earnings call, we’re building a revocable identification system, allowing users to build their own privacy profile and easily adapt it as their preferences evolve. This personal profile would be available across all web and app environments and would not be owned by any commercial entity nor limited to a software, operating system, or device. We’re working with the industry to make widespread use of this ID solution. This is very promising, and of course, we’ll update you down the road. That leads me to my last key topic for today: We remain focused on the five priorities for the short and mid-term that we indicated on our Q1 earnings. First, to ensure employee safety and business continuity. Second, support our clients with direct response marketing through their business recovery. Third, further align our organization and cost structure to support our strategic plan while investing for growth. Fourth, develop enduring, industry-leading solutions for online identification. And fifth, execute against our full-stack DSP strategy, both organically and via partnerships, while staying nimble as the landscape unfolds and new trends emerge. With that, I’ll now hand it over to Dave to go over our Q2 performance and our financial outlook. Thanks, all.

Thank you, Megan, and thank you very much for those very kind words. Good morning everyone. Let me echo what Megan said about everyone’s safety in this time and also appreciation for joining our call today. I'd like to start with our quarter two performance, and then I'm going to close with a discussion of third quarter guidance and provide some perspective about how we're thinking about the outlook for the remainder of 2020. To start, Q2 performance was obviously a lot better than we originally expected, with about two-thirds of our improvement compared to our April guide attributable to lower COVID impact and one-third to better underlying performance, including several positive one-time items. Revenue for the quarter was $438 million; ex-TAC, the revenue declined 18% at constant currency to $180 million. We estimate that the net negative impact of COVID on the revenue ex-TAC during the quarter was about $41 million, or about 19 points of growth. So if we exclude this estimated net impact, revenue ex-TAC growth was slightly positive, which was obviously encouraging. Another COVID headwind more than entirely impacted large clients using marketing solutions, particularly in travel, classifieds, and, of course, brick-and-mortar retail where some clients temporarily paused or even reduced their campaigns during the quarter. Travel, which remained deeply affected, contributed to over 50% of the net COVID headwind, with the remaining 45% to 50% of the impact evenly spread between classifieds and other verticals, while the impact on retail was close to zero overall. However, it was very contrasted between large clients and the mid-market; spending by the mid-market and direct-to-consumer brands remained resilient and was actually supported by COVID. Retail Media also benefited from a COVID tailwind. We estimate that about $4 million of this tailwind for Retail Media benefited Q2 only due to COVID. Excluding this positive one-time tailwind for Retail Media and the $3 million favorable technical items related to client rebates, the traffic acquisition costs in the quarter reviewed the normalized level of revenue ex-TAC, which was about $173 million in Q2. Now, changes in foreign currency in the quarter provided a $3 million headwind compared to Q2 last year, but a $2 million tailwind compared to our guidance assumptions. The revenue ex-TAC margin in the quarter was 41%, in line with the prior quarter and our expectations. Let's look at some of the operating highlights quickly. Our new solutions doubled their contribution to revenue ex-TAC from Q2 last year to 20% of our total business. Obviously, the solid growth of Retail Media was a significant driver here. Excluding COVID, retargeting only declined 7% at constant currency, compared to 27% on an adjusted basis. Net client additions were flat in the quarter, and we ended March with close to 20,400 clients, which is a 3% growth year-over-year, despite client retention declining to close to 80% for all solutions. COVID had a significant impact on same-client revenue ex-TAC, which declined 14% at constant currency, of which 21 points were attributable to the pandemic, and 7 points to positive contribution from client stickiness across Marketing Solutions and Retail Media. On the supply side, more than 4,700 direct publishers are now connected to one of our Criteo Direct Bidders on web and app, showing real nice growth on a year-over-year basis. Turning now to regional performance. Revenue ex-TAC in the Americas grew about 2% at constant currency if you exclude an estimated COVID impact of $14 million, mostly on large customers and the broader classifieds vertical. Retail Media’s strong momentum continued to accelerate in the U.S. For EMEA, Revenue ex-TAC grew about 1% at constant currency, excluding an estimated $16 million COVID impact, largely driven by travel weakness. Our new business was healthy across EMEA, with success across both direct clients and ad agencies. We’re also seeing a few large clients adopt more stringent banners for obtaining user consent in some markets. In APAC, Revenue ex-TAC declined about 3% at constant currency, excluding an estimated $11 million COVID impact on travel and classifieds verticals, while retail saw a continued shift from offline to online, especially in mid-market. Shifting to expenses, we reduced total expenses by $34 million on a GAAP basis and $27 million on a non-GAAP basis, which is $3 million better than our Q2 target before $4 million of unplanned bad debt accrual, most of which was related to U.S. customers. The cost favorability was driven by a number of initiatives that Megan and Benoit talked about last quarter. Operating expenses declined 22% on a GAAP basis and 20% on a non-GAAP basis. Headcount-related expenses represented 72% of GAAP OpEx, in line with last year as we reduced expenses across all areas of the business. And as we indicated on our first quarter call, we're being very disciplined with all of our expenses and our headcount. If you look at the non-GAAP expenses by function, we also had favorable year-over-year expenses across R&D sales, operations, and G&A. Adjusted EBITDA for the quarter reached $39 million, significantly $37 million at guidance rates, and $30 million above our original guidance for the second quarter. This drove the adjusted EBITDA margin to 22% of revenue ex-TAC, highlighting our increased focus on cost control in light of the current economic challenges. D&A expenses decreased 5% as a result of fully amortizing some technology intangible assets and our office right-sizing policy that we implemented in late 2019. The effective tax rate for the quarter was 30%, thanks to the positive effect of the French patent box regime. Net income was $6 million, down 51%. Adjusted diluted earnings per share were $0.27, only down 43% due to the positive impact of the share buyback program on the share count over the period. Cash flow from operations declined 37% to $33 million, almost entirely driven by negative changes in working capital due to a four-day higher DSO, reflecting continued impact of COVID on client payment terms. Despite some clients’ payment mentality, including in the U.S., our teams stayed focused and did a tremendous job of collecting payables in a timely fashion. CapEx declined 43% as a result of cost control, including through optimized server usage. Free cash flow came out solid at $15 million, thanks to reduced CapEx and resilient collections during the period, with free cash flow conversion for the first six months coming in at a strong 61% of adjusted EBITDA. Cash and cash equivalents on the balance sheet stood at $578 million as of June 30, after spending $15 million on share repurchases in Q2, and it also includes a preemptive draw of $157 million on our EUR350 million revolver. We purchased about 1.3 million shares under the buyback program at an average price of $11.70, financed through our free cash flow generation in the quarter. Let me now shift from the quarter and provide a little bit of guidance and perspective on the third quarter of 2020. The following obviously reflect our expectations as of today, July 29. As Megan explained earlier, our current view is that the pace of recovery from the pandemic may be much slower than most everybody anticipated, particularly for travel and classifieds clients. The lack of international travel remains extremely complicated for OTAs and all clients in the travel space. This will negatively impact the third quarter, traditionally our heaviest quarter for travel. With continued uncertainty around COVID, we expect continued impact on brick-and-mortar retail clients. The global situation on the job front and real estate sector will continue to be detrimental to clients in classifieds, particularly in Japan. As a result of the still highly uncertain economic situation, we don't perceive a full recovery happening before sometime in 2021, and obviously that's still very much a judgment. Going into Q3, we've seen our July performance decline a bit below 20% year-over-year on a global revenue ex-TAC basis. As of the date of this call, we expect August and September to show slightly more positive momentum, despite the weak expected contribution of travel and a more muted back-to-school season this year as a result of current restrictions. We estimate that net COVID impact on our Q3 revenue ex-TAC to be just below $40 million. Additionally, we’ve accounted for some headwind for the early adoption of iOS14 and also advertisers’ and publishers’ increased adoption of more stringent consent banners in some European markets. On the expense side, we expect Q3 non-GAAP expenses to be approximately $17 million, or 9% at constant currency, below Q3 last year, in line with our prior indication. Taking all of this into account, and as of July 29, we expect revenue ex-TAC for Q3 to be between $171 million and $173 million, translating into a year-over-year decline of 20% to 21% at constant currency, with foreign exchange expected to be a headwind of about 150 basis points of growth, or more than $3 million. It is important to say a few words about the Q3 sequential growth, i.e., Q2 to Q3. As we discussed earlier, our normalized revenue ex-TAC was $173 million in Q2. The midpoint of our Q3 guidance, which assumes a $3 million headwind from iOS14 and stricter consent banners in Europe, would be $175 million on a Q2-comparable basis. In other words, we see slight sequential growth on a normalized basis in Q3, before highlighting the expected weakness in travel, in what is traditionally our largest quarter of the year for travel. For reference, revenue ex-TAC of about $25 million was done with travel clients in Q3 last year. On the profitability side, we expect the still meaningful net COVID impact on our topline to translate into an Adjusted EBITDA for Q3 in the range of $31 million to $33 million. As usual, the foreign exchange assumptions supporting our guidance can be found in our earnings release. Now, importantly with respect to the business outlook for the rest of the year, if you recall, we withdrew our guidance for fiscal year 2020 on April 1st. Given how uncertain the situation still remains, we believe we’re still not in a position to reliably quantify the COVID-related impact on our financial results beyond the third quarter. Therefore, we continue to not provide guidance for revenue ex-TAC and Adjusted EBITDA for the fiscal year 2020. Let me share some of our current thinking for the fourth quarter. As of today, we’re currently modeling little economic recovery from the aftermath of COVID. Based on this macro backdrop, coupled with assumptions around the early impact of iOS14 from mid-September and the increased adoption of explicit consent in Europe, we believe the year-over-year decline of our revenue ex-TAC in Q4 could remain in a high-teens to low-20s percentage range. On the expense side, we remain focused on managing the expense base and delivering favorability in 2020 compared to 2019. This includes more moderate assumptions for the second half regarding the bad debt reserve. We continue to closely monitor our cost base and look into further opportunities to ensure that our cash flow generation is strong in 2020 and beyond. So with those highlights, Megan, let me turn it back over to you.

Thanks, Dave. In closing, we delivered solid performance in Q2 during highly unusual times. While the pandemic remains dynamic and the business context very volatile, retail continues to hold up well and to enjoy rising e-commerce sales. The mid-market remains strong, and Retail Media proves counter-cyclical in helping brands accelerate the shift of their trade marketing dollars onto e-commerce sites. We have unique assets for direct response marketing and identity resolution to help marketers get the performance they need. Our financial position is strong, and we’ll stay very disciplined in our capital allocation. The leadership team is on board and fully focused on building our full-stack DSP strategy for the future of e-commerce to capture large opportunities across the consumer brand and commerce space. And with that, I’d now like to open up the floor now to questions.

And we will now begin the question and answer session. And our first question today will come from Rocco Strauss with Arete. Please go ahead. Rocco Strauss, your line is open.

Sorry, I was on mute. Hey, altogether. Megan, I think you touched on IDFAs, the changes coming with iOS 14. Is there a way to size the share of inventory to data transactions on IDFAs in the quarter retargeting business as well as in the app install business within new solutions? And how quickly are advertisers actually able to shift targeting towards other IDs? And secondly, with the hiring of Todd from OpenX, the ID initiatives, and growing number of publisher integrations, it seems you are aiming more and more to cut out the SSPs and the exchanges fully from the equation, becoming your own SSP. I guess the question here is, is there some thinking of opening Criteo’s direct bidder to other third-party DSPs to enable campaigns on those third-party DSPs to be enhanced by Criteo’s ID graph? Thank you.

Thanks, Rocco. That was a hard one, good for you. Good to hear you. IDFAs, it is difficult to size just because it is not necessarily something that we're seeing the effects of yet. It is a little new in terms of how fast the pipeline has come up to speed with getting consents and everything that goes along with it. We've sized it enough to be able to make some assumptions around the hit it causes in terms of our revenue forecast for the rest of the year, where it impacts us. And Dave went through that before. I'm happy to have him share those numbers again if he has them handy. But again, what I was talking about earlier was our plan to work with the industry to actually get in front of this through some of the technology that we're building out in order to put in an ID system that will allow consumers to be in control, and directly use that to pass data across to us that we can use to continue to run our business. So, we do see we do see an impact for this year. Sizing it is complicated. We've focused on getting a solution in place to get profitable as soon as possible and something that we can utilize across the industry and help the ecosystem out. On Todd Parsons's role coming into our business, I'm thrilled with that. I don't know if you know Todd, but he's a terrific guy—extremely talented. We have a strategy in front of us that goes full-stack DSP, and there are exciting nuances now that sit in the center of that because of the strong performance of retail and the growth of e-commerce. So I haven't fully focused on that; clearly, it will become other opportunities to expand that capacity into other areas. But in terms of SSPs and becoming SSPs and going beyond what we have in terms of strategy on paper today, we don't have anything until he looks at it and he can tell us which is the right direction to go. And we have a lot on our plate. And the number one priority for me is that the team remains extremely focused.

Thanks, Megan.

All right. Good morning everyone. Thanks for the comment on IDFA. And trying to size it, as you noted, is complicated, Megan. I think we'll probably not try to dig any more on sizing for it. But some of the color on the Q4 impact was helpful. I guess, what I’d like to dig instead, you elaborated off the top on the Retail Media business, which is understandable why retargeting has gotten most of the focus. Retail Media is really coming on strong. What I’d like to ask a little bit more about is, I think the role of the use of an outsourced ad tech backend is always made sense for small and midsized players who don't have the resources to build themselves. But I’d love to hear you just talk a little bit more about how the conversation works with larger retailers who may have some of their own resources, but also tap into your offering and how you maintain strong relationships there. Thanks.

Thanks, Dan. Thanks for the question. It is a really good one. As you know, we do have relationships with some of the larger retailers; some of them use ad tech to offer some of the solutions inside of their own tech stack. For example, they pull up white-labeling, for lack of a better term, some of what we do. As we go into the next round about strategy and open our offerings up more, I suspect that there's going to be a lot more that you will hear from us in terms of what we're doing with the larger retailers. One thing that excites me about what we're doing here is if you think about mid-market being such a superstar during this time, and retail e-commerce being so strong is a step that there's a miss-looking at it now in terms of people searching to buy new products. The areas of retailer websites and other marketplaces have grown from 7% to 28% during COVID. Then you double that with a step we called out before in terms of 81% of consumers who switched products will now stay with those products post-COVID. There’s a whole marketplace here that’s appearing for mid-market retailers, which is exciting. What I think we have, what I know we have here in terms of Retail Media, plus the whole suite of the marketing solutions product as an offering that taps into both large retailers, but this growing mid-market range and mid-market retailers that are looking for an underlying platform like ours to be able to move advertising across—from products to people to sites, including retail sites—using privacy-driven identity. Everything we have today. Putting that all together, what I see is a massive opportunity for us. Just to circle back to your question, the large retailers that already have a presence are going to utilize some of that as well.

That's great. If I could have one follow-up. You spoke of course about the high-level strategy around identity and you come from Nielsen, where the whole idea of the single source of truth has always been the foundation of the business, and Dave obviously spent some time there as well. My question to you is, do you think that there can be a single source of truth in digital? And if not, what do you think the most important differences are? And how can an ecosystem work with multiple consensus identities versus the singular one? Thanks.

You guys are killing it this morning. All right. So I will put my Nielsen head on in the moment and respectfully answer the question. I think a single source of truth really means that somebody is a provider that has it. If everybody rallied around one provider that owned identity, then you might get to a place where you have a single source of truth, in the same way that Nielsen measurement is the single source of truth for television because everybody rallies around it and says, okay, this is what we’re going to align on. To your point, we're in a world where everybody has identity; everybody says they have the best identity graph. I truly think, from my experience, when I look at Criteo’s identity graph, it is the best I've ever seen. Ultimately, it is about how you use it; first, that you get it by having relationships with the sell-side and the buy-side, and ultimately with the consumer. When you have relationships, you get trusted access to information to do a better job of knowing who the consumer is and targeting them. If you do that well, then you may call it the single source of truth. The content doesn’t really matter. It is about you providing the service and getting to the right people at the right time, on the right device, in the right place—without annoying them because it is a contextual ad or contextual content that you've served up to. It is more about how you use it and the quality of the data you have, than this notion of a single source of truth for identity. I agree with you, and I happen to think that what we're sitting on here is gold.

That's great. Thank you for elaborating on it and congrats on hiring Todd.

And our next question comes from Nick Jones with Citi. Please go ahead.

Thank you for taking my question. Just one more on the iOS14 update. It sounds like you want to create or Criteo wants to create a solution that gives users control over tracking, or I think that's what iOS 14 is trying to do. But the idea is that I think a lot of people are going to opt out. So I guess could you maybe paint a picture of how you think the user engagement should be that I guess ostensibly would result in more opting in, and what that solution looks like in comparison to what Apple is doing?

Hi, Nick. We come from a place that we believe consumers should be in control. The operating systems are trying to, and admirably, trying to put privacy restrictions in place that protect themselves from sitting on information or the notion that they might be using consumers’ information in the wrong way. Consumers don't mind their information being used if it is used for the right reasons and if it is used by a trusted partner. Our notion is to give that right back to consumers, so they can set up their own way in which they want to be treated, how their data is to be used, what they like and what they don't like. Everything we're doing around the ID system, the revocable ID system, is to set up a structure in which consumers can have that control. We believe that if you give consumers back that control and they feel comfortable about how their data is being used, that they can choose to opt-in or out, they’ll be more likely to opt-in on a system like that, rather than just having one choice to opt-in or opt-out and trust that their experience is going to be a good one, which it won't. So, in short, we want to give control back to the consumer, and anything we can do to build technology to achieve that, we will.

And our next question will come from Sarah Simon with Berenberg. Please go ahead.

Yes, hi. I have a couple of questions. First one on the client retention: Dave, you mentioned it went down. Can you talk about the causes of that? I mean, do you think that it is basically COVID and classified and travel guys not spending anymore? Do you think there’s more materially we need to be thinking about? The second one is a really small point. Can you just remind us again what the $4 million one-off revenue was? I didn't catch the explanation. As we think about the 4,700 publishers you’ve directly integrated with, obviously that gives you first-party data, it addresses a lot of the issues people have around Chrome and cookies. So what’s holding back that number? Presumably, most publishers would want to integrate with you. Any thoughts on how you speed that process up or why you wouldn't want to would be helpful. Thanks.

Well maybe I could go quickly with the point about, I mentioned the 80% client retention. Sarah, we think that it is significantly related to COVID and it goes back to clients that were particularly brick-and-mortar retail clients that were severely impacted by COVID. This number is something that I think we'll have more clarity on as we progress through the third quarter, but I think that's a significant issue there. Regarding the $4 million, what I explained was the normalized revenue for the second quarter recorded at 180. And I said, if you look at that on a normalized basis, we had some technical items strictly related to the revenue accounting that were better than expected performance in the second quarter, which was $3 million. The $4 million is basically our best judgment into what was a tailwind for the Retail Media business that occurred in the second quarter as a flip side of some of the headwind we experienced on the marketing solution side. We don't think that will repeat in Q3. So that phenomenon, along with the $3 million, represents our best judgment, taking 180 recorded, and creating a more realistic view for actual performance and for comparative purposes looking at Q2 and Q3. And Megan, do you want to take the third element?

Regarding the publishers, so you're referring to 4,700 publishers and us getting first-party data from them. Let me flip to the buy side, where we have over 20,000 advertisers and we obtain first-party data from them. On the publisher side, the trick, as you know, is certainly we want as much as we can get. The trick is, how do they get it? The way to get first-party data is by creating relationships that be a sign-on subscription-based model, providing promotional opportunities, or anything else whereby you can get something back from the consumer. It is difficult for publishers to work out what their strategies are to gather such data. Some have done a fantastic job, while others are still trying to figure out what they can offer to gain something back. We’ll help as much as we possibly can, and I think we'll see that number increase over time. Regardless, that is, notwithstanding the publisher side being slower than the sell side, that's an enormous amount of first-party data that we have access to, which puts us in a very good place.

Okay. So basically, is it fair to say that you try and do these integrations more with those publishers that have first-party data than not? Or would you be trying to drive those integrations regardless? Because obviously, there are economic bidding considerations as well.

Yes, we try to drive them regardless. And data just gets better as they evolve. That's correct.

Our next question comes from Andrew Boone with JMP Securities. Please go ahead.

Good morning. Thanks for taking the question. Megan, some other people in the industry believe that Google eventually will pull cookies. Can you talk about your personal opinion there and how you see that playing out just given the impact on publishers, the open ecosystem, and COVID? Secondly, you guys talked about contextual targeting in the deck as well as in the prepared remarks. Can you help us better understand your capabilities there? Is there any degradation in performance compared to cookies? Thank you so much.

Number one, Google. Look, I hope they're right. We have a good relationship with Google. They have been a partner for some time. We respect what they do for the ecosystem. What Google is trying to do is clearly respond to the notion that they hold a lot of private information, particularly as it relates to CCPA. They want to make sure that they can do the right thing in terms of CCPA. So, they’re clearly focused on that. None of us can blame them. It is not about what they're doing; it is about how they're doing it. I suspect that for some time, COVID might slow them down. I still suspect that COVID might slow them down—not because they don’t move fast, but because the ecosystem needs time to catch up, and all of us have had a bit of a jolt. What I’ll say is that it doesn’t matter to us because this day is coming. The end of third-party cookies is inevitable. For us, it’s prudent to not worry about that timeframe but to think about how we get access to the information we need moving forward. Personal information will not go back; it will be about your relationship with the consumer and how you can access it in a safe and private environment so consumers feel comfortable and are happy with how their data is being used. So, I’m less focused on Google's timeframe than I might have been before. I’m more focused on the traction that we’ve made to take ourselves up five pegs ahead of where we might have been with Google. Regarding contextual targeting, that has been around for us for a long time. We’ve been doing it for quite a long time. It is never as good as third-party or first-party actual ID-based one-to-one type targeting, but it serves a purpose in terms of understanding the consumer you’re trying to engage with. I’m explaining it in very layman’s terms. It is a great fallback for us because it is such an old concept. It is not our first priority; it simply exists when and if we need it. We do contextual targeting today; it is nothing new for us, and we are quite good at it.

And this will conclude the question and answer session. I'd like to turn the conference back over to management for closing remarks.

Has the call dropped?

The conference is now concluded. Thank you for attending today's presentation. You may now disconnect.