Earnings Call: H2 2019

Feb 6, 2020

Ladies and gentlemen, good day and welcome to the full year 2019 results presentation of Publicis Groupe. For your information, this conference is being recorded. At this time, I would like to turn the call over to Mr. Arthur Sadoun, Chairman and CEO of Publicis Groupe. Cecilia, bonjour, and welcome to the Publicis Groupe 2019 full year earning call. I am Arthur Sadoun, and I am here in Paris with our CFO, Jean-Michel Étienne. The two other members of the Directoire are here with us, our Secretary General, Anne-Gabrielle Heilbronner, and Steve King, CEO of Publicis Groupe. As usual, we will take your questions all together at the end of the presentation. Alessandra Girolami is also here and will be ready to take your question offline after this session. We will start this call with an overview of 2019 results and highlights. Jean-Michel will then give you a full detail on our number. I will conclude by sharing our 2020 business priorities and outlook before taking all of your question. Now, let's dive into the presentation. But before anything, please take the time to read the disclaimer, which is an important legal matter. Okay. 2019 has been, without a doubt, a challenging year, but a constructive one. We have definitely seen the two phases and faces, actually, of our transformation. Our organic growth was severely impacted by the headwinds on our traditional business and the cost of our transition. That said, thanks to our new model, we are shifting our revenue stream, delivering very robust financial results, and the best new business performance in the industry. Now that we have finalized our transformation in term of assets and structure, our focus is only on execution. Before Jean-Michel take you through the details of our number, let me share with you the four key highlights of last year. First, 2019 was a pivotal year in our transformation with the acquisition of Epsilon. With this decisive step, we now have the technology and the data capabilities to compete in a market that is incredibly and increasingly driven by personalization at scale. Epsilon is now positioned as a core expertise in building, enriching, and activating first-party data. We haven't lost one second to include this capability in our broader offering to clients. We have merged Epsilon advertising activities into Leo Burnett and Arc, and decided to place CJ Affiliate under strategic review to explore different ways to unlock value. The integration is largely complete, and Epsilon has had a significant impact already on our client relationship. It has also been instrumental in our major H2 new business wins, with Disney and Novartis being the most notable examples. We are shifting our revenue stream towards data and technology, which now represent nearly 30% of our group revenues. With Epsilon, the reported net revenue growth for the group reached 9.3% in the full year and 16.2% in H2, fully oriented toward innovative products and digital services that our clients increasingly need. Second, our organic growth for 2019 stood at -2.3%, and came in line with the expectation that we shared with you in October. Our overall performance was severely impacted by three well-identified headwinds. The first is the attrition on our traditional advertising business, which represents an impact of around 200 basis points globally over the year, which means over 300 basis points in the U.S. Second, we have seen the negative impact of a few media loss of 2018. Third, we felt the cost of our transition, mainly with the repositioning of Publicis Sapient in the U.S. The good performance of our game changer, growing by 80%, and the positive effect of our new business were not sufficient to offset these headwinds. This situation is reflected into our geographical performance. North America was at -3.5% for the year, Europe at -2%, and APAC slightly positive at +0.8%. The U.K. and France delivered a resilient performance despite tough comparable at the end of the year. Overall, Q4 came in line with expectations, but the headwinds we are facing continue to be strong. As we already shared in October, they will continue to have an impact, particularly in the first half of 2020. Third, thanks to our model, we continue to deliver strong financial results while reinvesting in our business. Our operating margin increased by 30 basis points to 17.3%, including significant additional investment in talent for an amount around EUR 100 million. Headline EPS is up 8.2%. We will be proposing a dividend of EUR 2.30, representing an increase of 8.5% versus last year, and a payout ratio close to 46%. We maintain free cash flow generation at a high level, reaching nearly EUR 1.3 billion. This will enable us to fully de-leverage in the four years following the acquisition of Epsilon, as planned and announced. This financial performance was driven largely by our activity mix and our continued efforts in simplifying our structure. Last but not least, for the second year in a row, we are leading the industry in new business performance, as you can see in JP Morgan and Goldman Sachs research. As I said earlier, Epsilon has made an important contribution to that momentum. When you look at our most significant wins, you can distinguish actually three different buckets of win, and we are equally proud of each of them. First, the new clients that have elected the group for their marketing transformation, like Disney or NIVEA. Second, existing clients with whom we have pitched a new part of their business, like Mondelez, where we won the creative assignment, or GSK, that awarded us the Pfizer business they just acquired only one year after we won the account. Finally, the existing clients that have organized a review to consolidate, like Novartis or AXA, where we have been able to regain their trust for the future and get more business. It is important to underline that at the same time, while we were doing this effort, we have not lost any major global accounts in 2019. The new business momentum is actually continuing in 2020. Just yesterday, Bank of America has decided to consolidate its creative business with WPP to accelerate their data-driven creative approach with Publicis. I will now let Jean-Michel go into the detail of our numbers, and I will then come back to you with our priority for 2020. Thank you, Arthur, and good morning, everybody. I will now present our Q4 review and 2019 financial results. As usual, let's start with the net revenue, which is up 9.3% on the reported basis at EUR 9.8 billion. Currencies have a 3.1% positive impact. Acquisitions impact is at +8.5%. All in all, the organic growth in 2019 is -2.3%, in line with our October update. For Q4 2019, net revenue is up by 15.2% on a reported basis. Currencies have a 2.1% positive impact. Acquisitions have a positive impact of 17.7% thanks to Epsilon. However, organic growth is down by 4.5% in Q4. Moving to page 9, let's present the net revenue by geography. In Europe, net revenue is down 7% on an organic basis. This is largely explained by the consequence of the few 2018 media losses, combined with some specific items. First, the high comps in Q4 last year in the three main countries of the region, with a double-digit growth in Germany in Q4 2018, for instance. Second, in the U.K., the uncertainty linked to the Brexit had an impact in Q4 because of our exposure to financial services. Finally, France is impacted by some projects that have ended as expected and some reduction in scope of work at the end of the year, which is typical in Q4, which is an adjustment quarter. North America at -4.2%. This is explained by the U.S., where attrition is continuing, but also with the 2018 media losses with high and high comparatives at Publicis Media, and finally, the repositioning of Publicis Sapient on business transformation. Asia Pacific is down 2.3% with some pressure due to macro environment, mostly in China. Latin America returned to positive at 0.9%, supported by easier comparatives for this quarter. Middle East and Africa is down by 1.2% in the context of high comps and some digital business transformation projects which have been completed. Thus, for the group, organic growth is down 4.5% in Q4. In page 10, you will find the performance for the entire year by geography. I will just give you the growth on an organic basis. Europe is down by 2%. North America is down by 3.5%. Asia Pacific is up by 0.8%. Latin America decreased by 4.9%. Middle East and Africa is growing double digits for the full year at +10%. For the group, organic growth is at -2.3% in 2019. On page 11, let's have a closer look at our performance by country. Above 10%, we have, for example, United Arab Emirates at +17% and India at +10.4%. Between 5% and 10%, you will find Canada at +8.3% and Malaysia at +5.3%. Between 0% and 5%, Israel is at +3.2% and Spain at +1.1%. Below 0%, we have the U.K. and France close to zero, very close to zero, with respectively -0.2% and -0.8%, China -1.8%, and the U.S. at -4.1%. Let's move now to our consolidated income statement on page 12. I want to remind you that Epsilon is consolidated from July 1st. In order to present an accurate comparative analysis, we have isolated the Epsilon transaction cost, as already disclosed in H1, for an amount of EUR 40 million before tax. To introduce our income statement, we have in reality three buckets. First bucket starts with the EBITDA, which stands at EUR 2,245 million, up by 9.6% year-on-year. Excluding Epsilon acquisition cost, EBITDA is up by 11.5% and operating income by 11.6%. Operating income stands at EUR 1,699 million, Epsilon contributing, as said already, on the last six months. Second bucket is our headline group net income, up by 9.8% at EUR 1,188 million, after taking into account net financial expenses of EUR 106 million and income tax of EUR 396 million that I will detail later in my presentation. Third bucket is further down in the P&L with the group net income, which is down by 8.5% at EUR 841 million and includes the Epsilon transaction cost, net of tax, at EUR 30 million. A capital gain of EUR 21 million, mostly coming from the disposal of PHS health services. It also includes some non-cash items, namely the amortization of intangible of EUR 153 million net of tax, the real estate consolidation charge due to vacant location, as well as an impairment charge for a total of EUR 163 million net of tax. The revaluation of earn-outs, representing a charge of -EUR 22 million. On page 13, let's now enter in our operating margin details. To start, personnel costs decreased by 2.1 percentage points, representing 60.8% of net revenue. The reduction of this ratio is mainly due to the consolidation of Epsilon, which is typically less people-intensive than the rest of the group. Restructuring costs at EUR 118 million, up by EUR 12 million versus last year, in line with our transformation plan. Other operating expenses amounts to EUR 1,586 million, representing 16.2% of net revenue, compared to 14.2% last year. Again, this is largely due to the Epsilon consolidation, which have higher other operating costs than the rest of the group, and EUR 40 million of Epsilon transaction cost before tax. This leads us to an operating margin of EUR 1.7 billion, excluding Epsilon acquisition costs, representing an operating margin rate of 17.3%, up by 30 basis points versus 2018, in line with indication given in October. If we move now to page 14, showing the change in operating margin rate. Foreign exchange, mostly due to the U.S. dollar, combined with M&A with the disposal of PHS and Proximedia, had a 50 basis point positive impact on operating margin rate. This does not include the impact of Epsilon, which is shown separately. Higher restructuring charges than 2018 had a 10 basis point impact. As part of our strategic plan, we continue to invest in talent to grow our game changers. We have also been active in hiring key talents to support the future of our creative and media operations, notably in the second half. All in all, this has an impact of 120 basis points. The implementation of our new organization, especially regarding the country model, implies some simplifications, which have brought 40 basis points to the margin. We also benefit from 30 basis point in lower occupancy costs coming from our real estate consolidation plan. Variable compensation has been mechanically affected by the miss of our targeted organic growth. On the other hand, the portion related to the operating margin and cash flow has been raised as we over-delivered on these targets. The net impact is 70 basis points. Other operating costs, including bad debt, had a negative impact of 20 basis point. Depreciation had also a negative impact of 20 basis point. Finally, Epsilon has a positive impact of 10 basis point on the margin, as expected. Overall, this has led us to deliver in 2019 an operating margin rate of 17.3%, excluding the Epsilon transaction cost. Regarding now the net financial expense on page 15. The amount of interest paid has doubled in 2019 due to the Epsilon acquisition, largely financed with a EUR 2.25 billion bond issue. This is partly offset by the increase in interest income coming from our cash generation over the last 12 months. All in all, interest on net financial debt amounted to EUR 25 million versus EUR 11 million last year. When we include the three elements, which are, first, the interest on lease liabilities following IFRS 16 implementation. Second, the net exchange gain or loss. Third, the other financial expenses. The headline net financial expense is a EUR 106 million charge versus a EUR 80 million charge in 2018. After changing the fair value of financial instruments, net financial expense are up by EUR 20 million at EUR 91 million. Regarding the income tax on page 16, the 2019 tax rate is at 25%, up 100 basis points versus last year. We should note that in 2019, we have additionally the impact of the BEAT tax, which is the U.S. tax on imported high added value services for roughly EUR 20 million, which compare to EUR 30 million estimated at the beginning of the year. For 2020, we expect an effective tax rate to be slightly lower than 25% in a context where some countries have decided to lower their corporate tax rate. On page 17, the headline earnings per share fully diluted is growing by 8.9% year-on-year to EUR 5.02. At constant currency and excluding the BEAT tax, headline EPS fully diluted is up by 8.2%. Moving on page 18, relating to the dividend. We propose a dividend of EUR 2.30, representing an increase of 8.5% compared to last year, reflecting the increase in the headline EPS. That would imply a payout ratio of 45.8%, up by nearly 100 basis points, slightly above our commitment of 45%. The dividend will be submitted to the AGM on May 27th. On page 19, free cash flow before change in working capital is EUR 1.253 billion, up by nearly EUR 100 million compared to 2018, improving by 8.2%. The main items are, first, the EBITDA at EUR 2.245 billion, up by EUR 196 million, mostly thanks to the contribution of Epsilon in H2. Second, the tax paid is EUR 349 million, up by EUR 21 million. An increase mainly due to the BEAT tax, as already mentioned. Third, the CapEx is EUR 225 million, up EUR 29 million with a contribution of Epsilon in H2. Going into the change in working capital, it is a positive EUR 394 million this year, supported by a few positive elements, such as our cash management policy and a better overdue recovery, as well as the impact of Epsilon. As I've been saying that during years, for years, the objective is to be neutral on the change in working capital year after year. Overall, group free cash flow amount to EUR 1.647 billion, growing by EUR 336 million. If we move now to page 20, showing the use of cash, you can see that the main impact on the increase of our net debt is, of course, the acquisition of Epsilon. Indeed, acquisition net of disposal amounted to nearly EUR 4 billion in 2019. The earn-out and buyout paid is EUR 125 million. The dividend is an outflow of EUR 297 million. We have also a non-cash impact on net debt of minus EUR 283 million. Last year, it was a positive impact of EUR 183 million due to foreign exchange impact on the debt. This year, foreign exchange impact is smaller, but there are two main other items, which are the change in the mark-to-market value on cross-currency swaps due to the drop in the US dollar's interest rate. The second reason is the impact of the new earn-out and buyout. That leads to an increase in net debt of EUR 3 billion. On page 21, we are presenting the balance sheet. The biggest impact is, of course, the acquisition of Epsilon and to a lower extent, currency impact. Goodwill and intangibles are increasing by EUR 3.7 billion to EUR 13.6 billion. On the liability side, you can obviously see the increase in net debt to finance the acquisition. We are also regrouping all our people in the same building in New York, allowing a better collaboration between our teams, and thus committed a lease extension in June. This translate into an increase of EUR 400 million in the net right of use and a rise in lease liabilities under IFRS 16. Let's move to page 22 about the net financial debt. Average net financial debt is EUR 2.375 billion, up by EUR 1 billion versus last year. I will remind you that in 2019, the debt linked to the acquisition of Epsilon has been accounted for only six months in the average. As it will be taken into account for 12 months in 2020, of course, average net debt will naturally increase. Net financial debt at the end of 2019 is at EUR 2.7 billion. Regarding our financial ratios on page 23, we show the evolution of our financial ratios, including the IFRS 16 impact, which means including the impact of lease liabilities. With six months of Epsilon, including our average net financial debt, our leverage ratio is increasing by 0.5 times to 2.1 times EBITDA. We also confirm our goal to deleverage in four years, reflecting our prudent approach in term of balance sheet management. To conclude my presentation now, a brief word on our strong liquidity position at the end of December, standing at EUR 6 billion. I hand back to Arthur Sadoun, who will give you an update to our strategic outlook and, of course, I remain available for question during the Q&A session. Thank you. Thank you, Jean-Michel. Before we take your question, let me give you an overview of where we stand strategically and our priority for 2020. In 2019, we completed our transformation in terms of assets, but also in terms of organization. During this period, we have been obsessed with what our client wants. More than ever, they have to differentiate and add value to their product through creative excellence and navigate the increasingly complex media landscape. But they also need new capabilities to take back control of their customer relationships and adapt their business model to the digital ecosystem. Thanks to the investments we have made and all the changes we have implemented, we are now uniquely positioned to deliver what they need to continue to grow profitably. This through the four pillars of our offer. First of all, with our iconic creative brands, we can deliver groundbreaking dynamic creativity. Thanks to our unique media networks, tool, and partnerships, we have unparalleled clout and scale with a strong leadership position in the U.S. With Epsilon, we have the technology and data expertise to build, enrich, and activate clients' first-party data. Finally, with Publicis Sapient, we have the engineering and the consulting capabilities to redesign their operation with a customer-centric view. We not only have all of this expertise at scale now with Epsilon, but we are also able to deliver them seamlessly, which is exactly what they need. Thanks to our country model, we are able to do that in every country, and by the way, to do it globally through our unified client P&L and the Power of One. Today, our model is clearly adapted to our client needs, as you can see on this chart. It has been the case, and it has made the difference when we have been pitching those new business track record that you have seen. Now, for 2020, our priority is to execute on the four steps of our organic growth recovery plan. First step, stabilize and progressively return to growth in our advertising and media activities. In 2019, we completed the necessary changes to our organization to set the foundation for this. This is done and won't need any further improvement. In 2020, we will focus on leveraging our new structure to foster cross-fertilization in order to improve our performance, thanks to three structural actions. We have finalized our country model over 10 operation, each with a single P&L and a single leadership team. In the U.S., which represents more than 55% of our activities, we have appointed a single governing body, the U.S. COMEX, that I am personally chairing, composed of our key leaders in the market. This year, for the first time, business plans have been following this country led structure. The incentives model has shifted from global expertise to country or regional performance and is designed to reward over delivery on growth targets. This should accelerate our ability to cross-sell and upsell our expertise in creative and media, of course, but also in other areas like production, commerce, and CRM. We are leveraging our client-centric organization with our top client to fight attrition and grow with them. Growth plan have been designed for each of them across our 10 markets and around six different capabilities. Finally, we need to improve our win rates in regional and national accounts and pitches. We will apply our new business model that has enabled us to win most of the largest pitches in the last two years to midsize opportunity. Second step, we will bring back Publicis Sapient to growth in the U.S. by making the necessary changes to turn around the situation there. As we have announced at the beginning of last year, we are focusing Publicis Sapient U.S. on business transformation through industry verticals, actually reflecting the structure that supports growth in Publicis Sapient internationally. These involve, again, three decisive moves. First, we integrated Publicis Sapient U.S. digital marketing activities with our creative agencies under the brand of Razorfish to strengthen our relationship with shared clients and drive cross utilization. Second, as we did this, we are also disengaging from small one-off projects and digital marketing services to focus on resources that are less labor-intensive with additional growth potential. Finally, positioning Publicis Sapient U.S. solely around digital business transformation means we are also shifting from project-based work to multi-year client assignments. These are necessary measures, and it is the right thing to do if we want to drive sustainable future performance in the U.S. for Publicis Sapient, and do exactly the same as what we have done internationally. These changes produce already some positive results, but it will take time to fully materialize in our numbers. Third step, we will leverage Epsilon expertise for all of our assets. In only six months, we have put Epsilon at the heart of the group, and the integration is largely complete. We now have one unified team gathering all the Publicis Groupe data and capabilities under Epsilon leadership. We have built a unique data platform, Epsilon PeopleCloud, that spans the whole data value chain in onboarding, enriching, and activating our client data. Epsilon has been connected to our shared services. We have also started to expand Epsilon outside of the U.S., thanks to our footprint, and we are preparing the rollout of Epsilon PeopleCloud in key European countries. There have been many questions during the last months on whether a group like Publicis should own data or not. We actually believe that the right question is another one. Can our clients afford to fully outsource the management of their own data to a walled garden or an e-commerce platform? Or should they also start to build their own data capabilities? Some will tell you that an agency role is only to help clients navigate around a complex digital ecosystem of competing platform. Let's be clear. Of course, our clients need the scale of the platform to be heard. But at Publicis, we also believe that they are actually jeopardizing their future if they don't start building a direct relationship with their own customers through first-party data. It is not a coincidence that the companies that are the most advanced in this domain are also the most successful ones. This is why we offer both. On the one hand, with Publicis Media, we have developed over the last decade the right capabilities and partnership with the Google, Facebook, and Amazon of the world. The billing we are managing and the new business performance in 2019 clearly demonstrate that we are a leader in this market. But as we believe clients need to control their data, we are also bringing with Epsilon something unique that none of our competitor have, data and technology at scale to help our client build their first-party data and start to engage directly with their customer in a personalized way at scale. This is actually becoming even more important in an environment where privacy requirements are increasing, third-party cookies are likely to disappear, and some platform are changing the rules constantly, as we have seen recently with Google Chrome. Combining Epsilon capabilities with our traditional business give us a unique point of difference on the market. As I said earlier, today, our data and tech capability now represent nearly 30% of our revenues. In the future, they should be the biggest contributor to our growth, seamlessly connected with our traditional activities, in creativity and media. We will now focus on execution with three clear areas for investment, talent, training, and CapEx. This is the fourth step of our organic growth recovery plan. We are a people business. In 2019, while we were implementing our model, we continued to promote a new generation of leaders in strategic positions in our most iconic brands, our biggest countries, and our top clients. Only last year, we have invested an additional EUR 100 million in talent. We promoted 100 leaders internally and hire 150 top executives. This is one of the reasons why we are actually performing so well in new business. We made the decision to transition our talent profile and actually proceed to a massive talent uplift. This means we have also had to let go some people, as reflected in our restructuring cost of EUR 116 million last year. We will continue exactly the same strategy in 2020 by attracting the best people, but also continuing to transform and change the kind of profile we need for the future. To do so, we have put in place a new talent organization, and we will also align our incentive plan to encourage greater collaboration across teams. Embarking everyone in our transformation journey is exciting and demanding. This means we will need to over-invest in term of trainings. Our L&D program, learning and development program, should allow everyone to play an active role in our transformation. We will invest more than EUR 50 million in 2020 to continue to develop the initiatives taken in 2019 to raise knowledge on our expertise, products, and services. And on CapEx, our approach will reflect the fact that we are now both a service and a product company. This is why our CapEx should reach around EUR 300 million in 2020, including the integration of Epsilon. As a percentage of net revenue, this is above a traditional agency profile and perfectly demonstrates the uniqueness of our business model with investment that will make a difference to future growth initiative and new capabilities. So, in 2019, despite the challenge faced on organic growth, we are shifting the revenue profile of the group. We are demonstrating the attractiveness of our offer in transformative pitches and consistently delivering robust financial ratios. Now that our transformation is completed in term of asset and structure, our organic growth recovery plan will be our main focus. We have to be realistic. This recovery won't happen overnight. We have designed our roadmap to drive profitable organic growth with a revenue mix that fits with our client needs in the future. But in the short term, we will still experience headwinds, and the cost of our transition will continue to impact our growth rates. When it comes to 2020 outlook, we are confirming what we said in October. Our organic growth should be between -2% and +1%. We, therefore, anticipate that the first half of the year should be negative, particularly in Q1. Most of the improvements will come in H2. We will update you on outlook at the end of H1. We will then have more visibility on the four well-identified elements that will shape our performance for the year. The attrition we are experiencing on our traditional expertise, the new business ramp-up, mainly in our media activities, the result of the shifts we are undertaking with Publicis Sapient in the U.S., and the contribution of Epsilon in H2. As for margin, our model makes us confident that we will sustain our margin around 17%, and it will continue to be supported by the simplification of our structure on one side and our new revenue stream. 2019 was a tough fight, but it made us stronger. We are more ready than ever to work as a united team. Now, we will put 100% of our energy in executing our roadmap. Voilà. To conclude our presentation, I would like to thank our clients for their support in this transformation journey, and of course, everyone at Publicis Groupe for their efforts. I would also like to express our solidarity with our teams and clients in China. We are constantly monitoring the situation and are here for them at any time. Thank you very much for listening, and now we will take your question. Thank you. As a reminder, to ask a question, please press *1 on your telephone keypad. We will now take our first question from Lisa Yang from Goldman Sachs. Your line is open. Please go ahead. Good morning. I have three questions, please. Firstly, to follow up on your comments around Google Chrome dropping the third-party cookies. Investors seem to have been very worried about the impact on Epsilon, and it looks like you implied this could be an opportunity. So could you elaborate on that more specifically? The second question is on Sapient. Could you maybe give us some color in terms of how much of the business has been already transitioned away from the long tail/digital marketing services? How much business is there left to do? Is that mainly in Q1? Could there still be a few left to do in Q2, Q3? And have you seen any positive impact from that transition on your DBT business? If you can comment on the pipeline of new contracts you might have got so far. And thirdly, it's about the level of attrition. I'm just wondering how much of that is related to current specific issues. Is there anything you guys could do in 2020 to reduce that? Related to that issue specifically, P&G has been talking about reinvesting in marketing spending lately and has also outperformed on organic growth. Just wondering if that could ease the level of attrition for 2020. Thank you. Thank you very much. I'm afraid the three questions are for me. I was hoping that someone here could take one of those, but I will start. I will take it the other way around, or maybe I'll start with attrition. First of all, we don't comment on any particular client, but the good news is the client you just mentioned comment on us. My feeling is that you should ask him a bit our position there. He will tell you more things than I can in my position. Are we fighting attrition? Of course. Actually, the exercise we did this year with the business plan, with our top clients to make sure that we find every opportunity for cross-fertilization is a big operation. I don't know if you remember last year, but I announced that someone called Ros King, nothing to do with Steve King, by the way, that was coming on board. She was a CMO of Lloyds. She was coming on board to help us manage a business plan for every of our clients. This is in place. We did all the business plan. Of course, we have a growth plan for every of our clients. We are doing everything we can in this area. As you have seen, this is a very well-identified point that we have to manage. Now, to give you a bit of flavor on that, although we are seeing encouraging signs that actually attrition could slow down, and we are again doing everything we can, we are not seeing yet this in our numbers. So far, as you have seen in our press release, we are taking a conservative view on attrition, hence the minus 2% low end of our organic growth outlook. Again, this is things that we'll be able to update you at the end of Q2, but I want to make sure that we do the things properly. It's way too early, honestly. Sapient, I'm not going to go into every detail, but I think you raised a fantastic question about marketing services. Sapient, what we did is pretty simple. We have a model internationally that is based on industry practices or industry verticals, we call it, where we sell business transformation that has been delivering growth every year for the last five or six, or at least for the time I know Sapient. It's brilliant because it's exactly what our client needs, and it's very complementary to our marketing offer. We are really bringing together marketing transformation with media and creative and business transformation with Sapient. Now we can put data at the core. This is very powerful and the reason why we won what we have won this year. By the way, I will invite you to look at the press release of some of the clients that we win, and you will understand that they are not seeing us as a communication partner, but as a transformation partner. Again, it's not me saying this, it's our clients. But coming back in the U.S., we had a very different position because it was organized by geographies. It was a mix of marketing services, a bit of consulting here and then different things. We took the hard decision to rationalize all this. Nigel Vaz, that hasn't slept a lot for the last year, has done an outstanding job since he took his job in March to actually come with industry verticals, with new kind of leaders. We brought people from the Accenture, BCG, and McKinsey of the world to help us bring this culture in the U.S., and we are starting to see the beginning of traction. But we had all those people working in digital marketing services. That's actually brilliant. It's great talent, and it's people that we have here. This is exactly what our creative brands need to actually come with a fully-fledged offer that goes end to end in term of marketing transformation. Again, we didn't think in term of operation. We thought in term of talent. We brought, I'm not going to give you all the name, but we brought at least, I'll give you one, the CEO of our East operation on creative, Jem Ripley, that came from Capgemini, but that was a former Sapient manager, so knew the company, to actually take on our creative agency there and bring with him all the marketing services expertise, and more importantly, talent that we were having in Sapient. It made two things. First, it enabled Nigel Vaz and his team to focus on business transformation and on experience. Because don't get me wrong, the difference at Sapient compared to an Accenture is we know the brand, we understand creativity, and we are about to deliver experience. They brought that to their client and focused on that while all of these people, and you will forgive me if I can't give you the detail of the number, moved into our creative agencies. Again, I have to be realistic. We will see progress on our goals, but it's going to come step by step. But what we are seeing there is for sure incredibly encouraging when you look at the dynamic of the talent, very encouraging when you look at our new business, and hopefully we start to pay in our organic growth. Now, the Google question is a key one. Again, I'm sorry, my friend. I promise the next question is for you, whatever it is. If it's on the creativity, we'll give it to Jean-Michel Étienne. The decision announced by Google will have a significant impact on the entire marketing ecosystem. Agencies, publisher, a lot publisher actually, clients, and customers. The truth is we still don't know at this stage what will replace cookies at Google Chrome. I have to be cautious in what I'm telling you. There are some certainties that I can share with you at this stage. First, when it comes to Publicis, this announcement implied limited risk and a far bigger opportunity, thanks to the acquisition of Epsilon. If you give me a minute, let me remind you what is Epsilon, why it will not be affected, and how it will be an opportunity for us. First, I have read during the quiet period in some reports that Epsilon is heavily reliant on cookies. This is completely inaccurate. Let me explain why. Because the data gathered by Epsilon is actually very diverse. It's online and offline, also offline, it's behavior and social demographic. It's intent and transaction. Let's go into facts. Epsilon can see more than 65% of the non-cash transaction in the U.S. You are what you buy. We see one bought by out of two on a credit card in the U.S. Okay? They have gathered, and this is critical, they have gathered 200 million individual IDs, which is the base to create customer relationship. Out of those 200 million individual IDs, they are going up to 7,000 attributes to look at how people behave. Clearly, this is why any decrease in third-party cookies will not materially impact Epsilon rich data sets. That's the first point. There is a second one. We believe that what has been announced will be an opportunity for us as our clients will need to find alternative ways to deliver personalization at scales with transparency in measurements. It's as important to deliver personalization at scale than to deliver it with transparency in measurements. If you don't have the transparency, you won't build a long-term relationship with your client. We can do both. Where it gets very interesting, and maybe Steve can elaborate a bit on that, when you combine Epsilon expertise with the leadership position of Publicis Media in the U.S., we are building a unique point of difference in the market that is evolving constantly. We have seen that in the recent past through our new business track record, and you already have. I won't comment on any client that we won, but if you take some client that we won and you compare with the kind of results they have posted on things that we have done together, you will understand that our unique capabilities to understand customer, to build, enrich, and activate first-party data, and bring the scale of media, makes the difference. Everyone talks about personalization at scale. We know how to personalize. You should never forget that we also have the scale with media. You need both. If not, for our big clients, and you named a few, they won't be able to fight against the direct-to-consumer brand. Steve? Thank you, Arthur. I hope you can understand my accent. I think as Arthur says, what Epsilon very simply brings us is technology and data and scale and helps our clients manage their own data and other first-party data much more effectively. I think you saw that during 2019, we finished number one in new business again. Certainly for the major clients that we converted in the second half of the year, after we closed the Epsilon transaction, Epsilon was a core fundamental component of those successful wins. I suspect that without Epsilon, I do not think we would have had the same differentiated offer for those clients. You could see the scale of those clients. Arthur referred to some of those earlier. In the same way that we have been using Sapient behind all of our successful pitches before 2019, the second half of the year, we really saw how clients were reaching into how we could use technology and data in a really compelling way. I think everyone needs to assume that the wall of the walled garden will get higher, and we want to make sure that we can help our client build an alternative when they want to take back control on their customer. Again, we should not oppose both. As I said in the presentation, we need to make sure that we leverage the scale of those platforms, and Publicis Media is doing an outstanding job on that. But at the same time, they need to build a direct relationship with their customer. Thank you, Lisa. Sorry for the long answer, but I guess it was important point, so hopefully I was not too long. No, thank you. That was really helpful. As a reminder, to ask a question, please press star one on your telephone keypad. We will now take our next question from Conor O'Shea from Kepler Cheuvreux. Please go ahead. Yes, good morning. Thanks for taking my questions. Three quick questions as well. First question, just following up from the previous question on the repositioning of Sapient U.S. Just to clarify, did that already start in the fourth quarter and in your guidance for 2020, -2% to +1%, what kind of range of headwind are you assuming for that repositioning? Second question on Epsilon. I think in the previous Q3, you gave the number for the organic growth for Epsilon. I wonder, could you share it for the fourth quarter as well? Then the final question, maybe for Jean-Michel. Could you give us some guidance on the financial costs and restructuring costs in 2020? Thank you. Thank you. If you don't mind, I'm going to take question one on Sapient, then I'm going to give Jean-Michel, and I will end up with Epsilon for the last quarter. Sapient U.S. Yes, we did start it. Again, this had a big impact in 2019. It is one of the well-identified elements that will make the difference next year. Let me tell you quickly what we did. We changed the management team in February or March and put Nigel Vaz, that is the one who made the success of Publicis Sapient internationally, to build it in the U.S. He know how to do that. He has a fantastic group with clients. He has been with Sapient for 20 years, but he's a very modern manager. I hope he's not listening to the call. He's doing very tough because we're talking about a big shift, but great job at the moment. As I said, we are starting to see the results. But it's way too early to see it in our numbers. To make a long story short, we have not waited to start it. Actually, the structure is in place. The organization is in place. Most of the talent are there. For example, we brought a fantastic creative called John Maeda, because again, one of our differentiator is that we take an experience view and a customer view on business transformation. This is actually starting to get traction. We have taken the tough decision to cut the long tail, to move from project-based to long-term contracts. We have decided to focus our best people on the projects of the future versus short-term organic growth. Now, as you rightly pointed out, in the well-identified elements that will make the difference between the -2 and the +1, Sapient in the U.S. has definitely an important role to play. No pressure, Nigel. I think Jean-Michel is going to take the next one, and then I answer on Epsilon. Conor will answer regarding the restructuring charge for 2020. We expect resourcing charges to be at a similar level as in 2019, with some postponements from 2019 to 2020. We should expect an amount between EUR 110 million and EUR 120 million. Okay? Okay. What about the financial costs of with a full year of debt from Epsilon? Regarding the financial cost, we could expect +EUR 60 million regarding the acquisition of Epsilon. This is what we have in our forecast. Okay. All right. On Epsilon results. First of all, most of them are in the U.S., so I guess not a lot of them are listening to this call. But the least we can say is that the Epsilon team had a very busy year. When you take it back, we had the sales process during all H1. We had the integration as soon as H2, and honestly, it is almost completed, and more importantly so far, it is really a success. We didn't wait actually one day for them to focus on new business with us. This is what you see, what you have seen on the Novartis and even more importantly, on the Disney. If it was not enough, we have done everything that we said in term of efficiencies, which had obviously a very important impact on the top line. So very busy year. Now, as anticipated, as we said from day one, Epsilon total organic growth for the year is at -2% on the full year. This is, and it's a very important point, fully explained by CJ Affiliate and the agencies. The truth is, we did not wait one second to address the difficulties of these both businesses that represent more than 100% of the decline. CJ Affiliate, I told you quickly, we placed them under our strategic review to explore different ways to extract value while they are starting to collaborate with Publicis Media. The agency took a week. We merge it with Leo Burnett and Arc, which by the way, is giving great talent to those networks. I was talking about our win on Bank of America with Leo Burnett. You can see how energized the brand is. We created immediately, thanks to those capabilities, center of excellence in shopper and CRM in Chicago within Burnett and Arc. But we did that, I would say, in the first month, which by the way, had an impact again. This is why you have this number. We cut the long tail. We cut the good people. We did the work that needs to be done when you start it. Now, let's focus on Epsilon 2.0, which is roughly 80% of the business, which is our expertise in building, enriching, and activating first-party data. The growth of 80% of Epsilon, which is what we bought, what is making the difference in new business and will help us and help our client to navigate in the change we just talked. The growth over the year is at +1% with a flat H2, which is exactly in line with what we are expecting. So we are on track. Now, it's really about using our Epsilon PeopleCloud platform to actually deliver personalization for our clients, continue to grow our business through cross-organization and new business, and don't miss the culture. I made the point about learning and development. This is also why we are increasing our pool there, is we have a fantastic engine that represents a massive opportunity to help our client in a new way. Again, the Google Chrome announcement helps that. But we need to make sure that it's spread everywhere. Hopefully, we made a simple and comprehensive presentation about our four pillars. Epsilon is a new one in the family, and we need to make sure that people that are not part of Epsilon, but we need them in media, in creative, in Sapient, can understand how it works. Great. Many thanks. Thank you very much. We will now take our next question from Julien Roch from Barclays. Your line is open. Please go ahead. Yes. Good morning, Arthur, Jean-Michel, Steve, Alessandra, and Brice. I have four questions, but they are all numbers driven, and you can answer in less than one minute. I will ask them one by one, if that is okay with you. You are moving to one Publicis, but you still have verticals like communication media. How much did media grow in 2019? If you do not want to give us an exact number, and I would prefer an exact number, can you tell us whether it was negative, zero to 2.5, 2.5 to five, or five to 7.5? As you know, we do not give the number per expertise. The reason why we are doing it, honestly, is because it is becoming more difficult also to differentiate the value that each of our operation brings. I am going to give you a very simple example. When you win Disney, it is a media pitch. Okay? When we go on stage, we said, "We are not going to give you a media proposal. We are going to explain you how we can help you to transform your marketing." True for Beiersdorf, true for Novartis, true for Daimler before that, is we come on stage and we explain that. The win is going to be in Publicis Media. The different component of the pitch will have the help and then follow up in the same way. This is why we are moving to a country model. I guess it's difficult for you to understand how much of the shift it is. When you start to say in the U.K., not to talk all the time about the U.S., that you're going to have a single CEO with a single team under a single P&L, and that what our client needs is more integration. It's to connect creativity with media, with data and technology. For the sake of this transformation, we need to start thinking in term of the Power of One and bring an end-to-end solution. This change the culture, definitely, but also change the way we report our number. I can't give you a detail on that. I can tell you that if you look at the new business type table, you will see that Publicis Media won most of the pitches, and at least the two biggest in 2018, which were GSK and FCA. This year again, not most, but all the significant pitches with a significant difference, which is we haven't lost any major account in 2019. That was the first one. Sorry, it was not a one-second answer. Yeah, that's okay. Attrition cost 200 basis point in 2019. But how much was attrition in terms of revenue, either in EUR or percentage, in 2018, the starting point, or in 2019, the ending point? Roughly the same. That's fast, no? How much- It was roughly the same. This is why we are staying very realistic for next year. Again, we are feeling that attrition is slowing down, mainly because we are cross-fertilizing better. For the moment, there is no impact on our number. By the way, we continue to see some clients that are calling you. You just have to look at the earnings and compare it with the clients that are saying, "We are in a significant massive cost-saving plan. As a partner, you will have to take a part of it." As you know, we are very exposed in traditional advertising, which is a good news for our margin, but not always a good news for our revenue because this is what they cut first. We have to assume at this stage that if we want to be realistic, there will be attrition, for sure. Will it be the same level of 2018 and 2019? Too early to say. We will tell you more at the end of H1. Sorry, Arthur, I was unclear. I am not asking what is the impact of attrition, which was 200 basis points in 2019. I am asking what is the percentage of revenue that you consider to be attrition. Is attrition 30% of revenue? Is attrition 50% of revenue? If I understand well your question, you are asking what part of our activities is impacted by attrition? I am asking what percentage of revenue. When you say attrition impacted revenue by 200 basis points, which is the annual impact, what is the starting point or the ending point? When you say- We are talking about the creative business that is roughly 30% of our revenue, so roughly EUR 3 billion. Okay. Thank you. Third question is impact of net new business in 2019 overall. You said negative, but how much? And whether you can give us any indication on 2020 already. No, the net new business is positive. You just have to look at the different report, but I can't give you a number. I saw you say in the statement that net new business in 2019, if you include the losses in 2018, was negative. No. Or I missed something. Okay. But maybe Jean-Michel, you want to add something on that? The contribution is positive. There are other factors which explain, of course, our negative organic growth of the year. No. If you want to decompose organic growth, it's pretty simple. There is two good news and two structural headwinds. The good news is our game changer are continuing to grow, of course, less than the past year because the scale is bigger by 18%, and our net new business is definitely positive overall. But then when you add to that the attrition, we're having our traditional business on one side and the cost of our transition, you understand the net. What I was saying earlier on my presentation is that although our game changer are progressing in a good way, and although new business is helping our growth, not enough to compensate the headwinds we are seeing on the both items I just mentioned. Okay. Last one, CJ Affiliate, which you're selling, what were the revenues and operating profit in 2019? So we can have the base and they consolidate CJ Affiliates. Not a lot of change compared to what you know already regarding CJ Affiliate. This is roughly $150 million revenue with a high margin, as you know. But again, we put it under strategic review, and we are exploring the different option at this stage. The thing is, hopefully, you would agree that we did not lose a second to integrate Epsilon. The reason why, honestly, integration is going so well is that it's not one or two people that bought Epsilon, it's an entire management committee that decided that this is what we needed. Once we did that, integration has been way faster. We have spent most of our time on what is the future of Epsilon and the future of Publicis, which are their ability to build, enrich, and activate first-party data, which is 80% of the business. We have been fast in integrating the creative business because we knew that for revenue and cost reason, it was making sense. By the way, the expertise they were having was precious for Leo Burnett. On CJ, we are taking our time. We are going to do what is best for everyone, starting with our shareholders. We are taking our time. Okay. Thank you very much. We will now take our next question from William Packer from Exane BNP Paribas. Hi there. Thanks for taking my questions. Three, please. Firstly, you didn't really want to comment on the performance by vertical, so could we have a performance by client type? FMCG was around 25% of revenue in 2019, I think. How rapidly did it decline in 2019, and can it improve in 2020? Secondly, could you just remind us on Publicis Sapient, what percentage of group revenue is still exposed to those difficult dynamics that you mentioned on the call? Then finally, if we rewind to this time last year, we had an outlook which was focused on an improvement in the second half of the year. We don't have the same by now, but certainly it's going to be second half weighted. Could you just talk us through what's different this time? What gives you the added confidence that the second half of the year should improve? Thanks very much. Maybe I'll start with the outlook. Thank you. Hopefully, what you will feel here is we have decided in 2019 to do the hard work. I mean, the country model, the repositioning of Sapient, the acquisition of Epsilon, bringing more than 150 new talents, doing more in promotion, investing more than EUR 100 million in talent at a moment where we have a negative growth. We did a lot of things to make sure that we can come back to organic growth. We have been feeling good about it because whether you like it or not, winning new business shows that you have the right model for the future, and we have been able to sustain a very strong financial, which, by the way, shows the fantastic work that has been done to simplify our structure. Which is not only for cost reason, it is first because this is how you should manage a company in a leaner and more agile way. But it is also showing that despite the negative growth, we are improving our margin, which means that something should happen in what we are winning versus what we are losing. All of this is good. Now we know that organic growth is our top priority and focus. We have a very clear roadmap, as I described, to create cross-fertilization, will it be at country level or at client level. We know how we can turn around Sapient because it is the case outside of the U.S., so we need to bring it there. We have an Epsilon that is starting very good momentum and traction that has already won new business with us. That will, of course, bring new things next year. And we have very solid financial ratio, which means that we can continue to invest. So you take all of this, you look at the dynamic we are having at the moment, again, Bank of America yesterday. And yes, we feel confident that you will see sequential improvement over the time, and although I mean, H1 should be negative, particularly in Q1. We think that we will see some improvement in H2, and time will tell, but we are focusing only on that. We gave a large guidance from -2% to +1% for the same reason, which is we know that if we want to have a good H2, attrition has to go down, new business ramp-up has to accelerate, Sapient in the U.S. has to win project, and the contribution of Epsilon should be better. But that things that are on the way, we are doing this 20 hours a day. And the truth is, it will take time. We have to be realistic. We need to make sure that we manage that through time, and we do that properly, but we feel extremely confident in the model we are building. We went through tough time, we took tough decision. It is a moment where we are working like crazy. But there are signs of the fact that, yes, contribution in H2 should be better. That was for this. Then, Publicis Sapient. Well, just by memory, Publicis Sapient is roughly 20% of our revenue, and the U.S. is roughly two-third. And out of those two-third, you can extract the marketing services Part that which is more compared to the rest still, and you have an idea of the level. Now, I am not going to comment on every industry vertical and how they are growing, but I am happy to take the FMCG question. For those that are European, for me, FMCG is a bit like GAFA. We are mixing things that are completely different. Within the FMCG, you have roughly those that have been investing in technology, investing in their brand, investing in innovation, and that are having a fantastic track record, and again, during this year earning result, you will see some. And another that did not enough. And the truth is, the good news is, the ones that are winning, we have some. The good news is, the ones that are not winning yet, we are doing everything we can to be on their side, to help them grow in the future. But it's true that for some of them, we have to take the pain for them, with them, and sometimes for them. And we will make sure that in the future, we leverage our loyalty and our commitment. But honestly, it becomes more difficult every day to talk about FMCG in general. Thank you. Just in terms of an improvement in trends with those customers, are you expecting an improvement? Improvement with what, sorry? With the FMCG client spends with you? Do you expect that to improve then? We are seeing some early sign of slowdown in attrition. But when I say we are seeing, it is more in the commercial relationship, in the time we spend with clients, in the way they want to invest in their business in general, but it is not yet materialized in our number. This is why, again, we presented the -2 to +1. Thank you. We will now take our next question from Adrien De Saint Hilaire from Bank of America. Please go ahead. Hello, good morning, everyone, and thanks for taking the questions. I have a few of them, please. First of all, we saw this week that Accenture decided to stop their media auditing practice. Do you expect Accenture to be more aggressive around media buying and planning for 2020 as they drop this activity? Secondly, can we have an idea of the different growth rates between Sapient International and Sapient U.S., if the international bit is the blueprint for the U.S.? Then thirdly, Arthur, I am very sorry to press you around short-term trends, but should we deduct from your comments that you expect Q1 to be in line with the Q4 trends? Thank you very much. Thank you, Adrien. First of all, as I said, H1 should be negative, mainly in Q1, but we expect sequential improvements. We are not expecting Q1 to be at the level of Q4, but Q1 will be negative. H1 should be negative. You understand that, again, if we are negative on six months, the impact on our growth in Q1 will still be significant. Okay? The second question was on the growth rate of Sapient internationally and in the U.S. Let me do a short answer that is in line with what I just said before. We are growing outside of the U.S., we are not growing in the U.S. That is it. Then the question of Accenture is a very important one, actually. Let us start by where we are. First of all, and they said it properly well, by the way. In shutting down their media audit department, Accenture is aligning itself with the kind of new non-conflict rule that are in our business and that are very important. They had to do so. You can take the glass half full or half empty. If you take it half empty, you say, okay, you have a company that is very powerful, that is coming into our field. Because the truth is this, they are coming into our field. If you take the glass half full, our opportunity for us, this is actually, and this is very important because this is something that we have been discussing for a while. We had the discussion with Adi Troga five, and more importantly, this is something that we are claiming since Maurice Lévy did the acquisition of Sapient, which was honestly maybe the most visionary acquisition that we did so far. What is happening with Accenture today is definitely an additional sign that the convergence between marketing transformation and business transformation is happening. If you believe that Accenture is going to spend more time on our field, the fact that we have with Sapient a 15,000 strong engineer, data analyst, consultant that can bring the same firepower in term of technology and data to a pitch or to a client, put us in a unique situation. I will not mention any pitch, but you will find it in the press. It will not be the first time that we have been opposed with them. On every big pitch where we have been confronted to Accenture in our marketing field, where they were bringing their expertise, we were able to win for two reason. The first is we know our field for decades, and the second is we have an arm with Sapient that can fight back pretty easily. It does not mean that we are going to win everything. We are in a trillion dollar market, so it is not like the market of transformation. The truth is there is room for everyone, but this is an additional sign of the convergence between marketing transformation and business transformation. Thanks very much. We will take our next question from Matthew Walker from Credit Suisse. Your line is open. Please go ahead. Thanks a lot, and thanks for some of the explanations around Epsilon. If you could just clarify. So within Epsilon, you have a lot of client first-party data, that is usually matched with third-party cookies to see how people are moving around the web and what they are doing. So could you say how much revenue in Epsilon does come from that area where you are matching first-party data with third-party cookies? The second thing is, could you maybe explain the costs of expanding Epsilon around the world? You mentioned that you are expanding it into markets outside the U.S. So if you could tell us how much is that going to cost and how long it is going to take. Okay. As I said, on the first question, we are definitely not depending on third-party cookies. We cannot give you the detail. Maybe Steve can tell you what, by the way, on how it moves also with media, because this is more. On Epsilon, it is an opportunity, way more than a risk. On media, it will mean some changes that we cannot anticipate too much at the moment, but it will mean some changes. Maybe, Steve, you can tell us a word on that. Yeah. Hi, Matthew. Yeah. As we said earlier, this decision, which was just announced by Google, but we had anticipated, as you will have seen and heard earlier, it is definitely going to have a significant impact on the entire ecosystem that we operate in the whole marketing sphere, obviously on agencies, on publishers, clients, and of course, customers. The impact probably is going to be on clients, as you ask, is probably going to be threefold. Enriching their own data is going to become more difficult, definitely. It is going to narrow the number of potential partners that are able to do the activation, and it is going to limit their ability to measure effectiveness in a really transparent way. This is likely to translate into less relevant advertising for consumers and more marketing waste. I think as you have seen earlier, when we first met with Epsilon during the due diligence, one of the most impressive things is they have been planning for a cookieless world back from 2012. They are really prepared for this. This, we think, is a huge opportunity for us and as our clients need to find other ways, alternative ways, to deliver personalization at scale with transparency and measurement. Obviously with Epsilon, we can do two things. One, we can find some alternative ways of reaching these consumers, and two, we can use the technology to help them build their own first-party data solutions, the datasets that they are going to build to enrich it, and of course, a platform in order to activate it. The fact that we have Epsilon at the core of our business, as you can see, is something that we actually see as a huge opportunity. Although the impact of this by Google Chrome is difficult to predict on the whole environment. Certainly, I think our clients are going to see this as a positive momentum from us. Thank you, Steve. On international, I know that time is flying, but I will have to take a second. Epsilon does three things. The first is they have the technology to build first-party data. This is a technology that we can export in our main countries, and we have a plan for that. It's not a huge source of investment. We need to make sure that we have the people in the countries that can do it. The second point is the opposite. Epsilon in the U.S. has these massive datasets that is a fantastic opportunity at the moment where the wall of the walled garden are getting higher. That, of course, is not exportable. The question is, what kind of data do we have in the different countries? In some countries, we are pretty advanced, like in France or in Germany. In other, we need to make sure that we build the right dataset, and this is where there might be some investment to make. It's too early to say because we can start with what we got. To answer your question, this could be an area in investment, which is data that could be activated at the local level. The platform is going to be only one platform. It's the Epsilon PeopleCloud, and you will see the interface, the way we can work with that, the way we can boost it. From, by the way, our Publicis PeopleCloud. For those that remember the Investor Day, it's not that we started eight months ago thinking that data were at the core of what we do and that first-party data will be the future. We started two years ago, that we put a big investment on the table to make sure that we can accelerate and get the scale. The Epsilon PeopleCloud is something that we're going to export fast, that is going to help us, by the way, with Epsilon, but also with our media operation. Thank you very much. Thanks a lot. We will now take our next question from Omar Sheikh from Morgan Stanley. Your line is open. Please go ahead. Morning, everyone. Sorry to you with the time. I will ask just a couple of quick ones, if that is okay. First of all, Arthur, on Sapient, could you just quantify what you expect the shift from project-based work to have on organic growth in 2020? So, what drag do you expect during this year? That is the first question. Secondly, on attrition, you kind of tantalized us with the comments you made about seeing encouraging signs that it could slow down. Could you maybe just quantify that? You highlighted 200 basis points of drag in 2019. Do you expect that to be better in 2020? Thank you very much. Okay. I guess my English is not good enough. I am going to have to make sure that we are very clear on that. We see some early sign of our clients starting to spend and understand that there is a future in spending more in data, in technology, and in creativity, which is a good thing. The problem we are having there is that we do not see that in our numbers. We need to be realistic for the moment. I can promise you that at the second we see this improvement, we will come back to you and starting with you, Omar. But for the moment, we need to be very realistic on that. On Sapient, I cannot give you the detail of the project base, because again, it is going from one year to another, and some things that we have stopped this year, we started last year, we still have an impact this year. But again, as I said earlier, definitely, Sapient is one of the well-identified elements that will make the difference between the minus 2 and the plus 1. I am sorry I cannot give you more detail at this stage, but hopefully you will understand at least that we are sharing all of this with you to put you in our kitchen, to understand the kind of decision that we are putting on the table to make sure that we come back to organic growth and to come back to one of our previous question, we sequentially improve our organic growth while maintaining our financial ratios and bringing to our client what they need for the future. Okay. Thank you very much, Arthur. Thank you very much. I see we have time for one last question. We will take our next question from Laurie Davison from Deutsche Bank. Please go ahead. Okay, just one question for me, in the interest of time. The CapEx to EUR 300 million. When we are thinking about the new level of CapEx intensity for this business going forwards, can you give us some idea of where you are thinking about CapEx for sales now going forward? Thanks. Jean-Michel. I am taking you through this CapEx, where we say EUR 300 million of CapEx for 2020. This has increased due to Epsilon, of course, which is a techno company, so this means that we have more CapEx on the part of the group for sure. We should also take into consideration that in 2019 and 2020, we have still some CapEx which are incurred to support the real estate consolidation program, which is something that will be finished at the end of 2020. So we could have a little bit less CapEx in 2021, but we have also to invest in the platform that we have, the various platforms. So we consider that the EUR 300 million of CapEx that we have in 2020 is something sustainable for the future. It comes back to the previous question about the investment on Epsilon. Part of this CapEx is the development of Epsilon internationally in terms of technology and assets. So it is all in our plan, and again, the good news about Epsilon is that this is what people want now, our client wants now. This is why we are accelerating on the international side, is that our client have seen the kind of traction we are starting to have in the U.S., and they want to make sure that we can expand that. Look, just to wrap up, 2019 has definitely been a year of transformation. A tough year in a way, and a very exciting year on others. We have adapted our structure with a country model. We are, and we have transitioned to be Sapient in the U.S. to make sure that we can bring business transformation as we are doing outside of the U.S. And finally, we have acquired, and I would say sincerely, successfully integrated Epsilon. As you have seen in the presentation, we definitely have the profiles, the assets to deliver what our client needs, and we are obsessed by that at the moment. What has happened with Google Chrome, what has happened with Accenture, the way our client are moving now means that what will make the difference tomorrow, if we want to sustain a very valuable business and grow, is to bring what they need, which will always be creativity, which will always be media because they need the scale, but they will need the data to take back control, and they will need the technology to transform their business. It's very difficult to bring all of these people together. It's very difficult to make them understand that what matters is an end-to-end model. But when we do it well, we win, as we showed in our new business. When we are doing well, we are improving our financial ratios. To be clear, our main focus in 2020 is the execution of our plan to foster organic growth, and we will be focusing on that every day. I will just thank you very much for this attention and this long call. Normally, we do shorter. I hope to see you soon. Merci beaucoup. Thank you.