1. 1 Google confidential Cross Media Analysis Increasing ROI in the media mix LOreal case study

2. 2 Google confidential2 Google confidential Main content Summary (slides 3) Set-up and Objectives (slides 5-7) Study objectives Campaign overview Set-Up research Analysis (slides 9-13) Google Display Network (GDN) Optimizing ROI Curves Optimization of Pre-roll campaigns Frequency capping Recommendations (slide 15) Appendix (slides 17-19) Monthly efficiency indices ROS (untargeted) and targeted Pre-roll Explanation regression analysis 3. 3 Google confidential3 Google confidential Increase ROI by 32% by shifting 10% TV budget Current budget allocation New budget allocation 72% 17% 11% TV Print Online 65% 21% 14% TV Print Online 10% of TV budget towards Print/Online ROI indexed on 100 ROI = 132 4. 4 Google confidential4 Google confidential Set-up and objectives 5. 5 Google confidential5 Google confidential Study objectives Main Research questions Research questions for optimizing the media mix How do cost per GRP for TV, Print, Display, Video Ads and Rich Media compare? How to optimize reach and (effective) frequency across media/formats? How does targeting of media and ad formats compare to each other? What is the impact of the different media and ad formats compared to each other? How do cost/effect (ROI) scores compare? 1 Online and off-line have different languages: how do we integrate them towards one currency? 2 Online and off-line have different impact and cost: how do we compare and optimize the media mix? 6. 6 Google confidential6 Google confidential Research and campaign period (18 campaigns included) The research and analysis were done over a 5 month period, including 18 campaigns of 14 different beauty products. February March April May June July Week 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 Product 1 Product 2 Product 3 Product 4 Product 5 Product 6 Product 7 Product 8 Product 9 Product 10 Product 11 Product 12 Product 13 Product 14 7. 7 Google confidential7 Google confidential Method: Media Efficiency Panel Purchase and media behavior measured from the same individuals (single source) Print Folder Loyalty Price TV Radio Online Psycho graphic Demo- graphic Media Purchase Personal Promo 8. 8 Google confidential8 Google confidential How to reallocate 10% TV budget and increase ROI by 32%? Cost Targeting Impact Cost/Effect 9. 9 Google confidential9 Google confidential Low cost/GRP on GDN results in opportunity to increase ROI using a small part of the 10% TV budget The cost/GRP (or CPM) on Google Display Network on average are ~20 times lower than TV, targeting is 9% better and impact per served impression is about a fifth of TV. This results in ~4 times higher ROI at current spend level. A small part of the 10% shift in TV budget is allocated to GDN resulting in 4 times more GDN impressions. Cost 20,11 Targeting 1,09 Impact 0,19 Cost/Effect 4,20 556 23 - 100 200 300 400 500 600 TV Google Display Network Net Cost/GRP Source: price of GDN is based on the auction + Cost/GRP TV from Carat Media facts booklet (TV net cost based on 70% average market discount) 10. 10 Google confidential10 Google confidential Shift budget to media that are in steeper part of ROI curve Shift part of TV budget towards print and part of Rich media other budget towards YouTube 0.00% 0.05% 0.10% 0.15% 0.20% 0.25% 0 2 4 6 8 10 12 14 TV Print Rich Media YouTube Rich Media Other Average number of contacts Increaseinsales = Average number of contacts 11. 11 Google confidential11 Google confidential For many audiences targeted pre-roll is cheaper than TV On TV smaller audiences have a higher GRP price whilst each GRP delivers fewer impressions. YouTube Pre-roll has more stable rate by audience. Net Cost/GRP Indexed against TV for common TV audiences 1.0 0.1 1.0 0.2 1.0 0.6 1.0 0.3 1.0 1.1 0 1 2 3 4 5 6 7 8 9 10 M20-34 20-34 20-49 V20-49 13+ Net cost/GRP (euro) Population size Impressions per GRP Average Net CPM TV (euro)* Average Net CPM lOral 13+ 625 13,753,000 137530 5 5.5 20-34 795 2,996,000 29960 27 5.5 20-49 691 6,832,000 68320 10 5.5 Shoppers 20-49 633 3,732,000 37320 17 5.5 Shoppers + Kid 625 1,896,000 18960 33 5.5 Men 20-34 925 1,502,000 15020 62 5.5 Women 20-49 624 3,398,000 33980 18 5.5 Source: CPM targeted Pre-roll on YouTube in the auction was 5,5 euro + Cost/GRP TV from Carat Media facts booklet (TV net cost based on 70% average market discount) 12. 12 Google confidential12 Google confidential HIGH SEASON HIGH SEASON Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Targeted Pre-roll: 14,- CPM Untargeted Pre-roll (Run of Site): 10,- CPM Monthly TV costs are STER, SBS, RTL and and average. YouTube Pre-roll costs are based on the YouTube 2012 rate card Cost/GRP Pre-Roll is extra efficient in expensive TV months Large Pre-roll opportunity in expensive TV months due to seasonality. Pre-roll does not have this seasonality over months. (calculation Pre-roll versus TV ROI by month in back-up) 13. 13 Google confidential13 Google confidential Cap frequency for sites with very frequent visitors Sites/networks with high visiting frequency create too much contacts. For those sites, frequency capping increases ROI strongly. 18.6% 7.3% 9.2% 4.6% 11.6% 4.0% 8.2 4.2 3.0 8.2 3.6 5.1 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 0% 2% 4% 6% 8% 10% 12% 14% 16% 18% 20% Display Google Display Other Rich Media Google Rich Media Other Pre-roll Google Pre-roll Other Reach Frequency 14. Google Confidential and Proprietary14 Google confidential Recommendations 15. 15 Google confidential15 Google confidential Recommendations: shift 10% TV budget 1 Shift part of the TV budget towards Google Display Network. The combination of 20 times lower cost/GRP and one fifth of the impact compared to TV results in ~4 times higher ROI at current spend level. 2 Shift budget at the end of the TV ROI curve towards media that are still functioning in a steeper part of their ROI curve (Print and Rich Media). 3 Use variation in TV pricing over months and target audiences as an opportunity to find month versus audience combinations where Pre-roll is a cheaper solution than TV. (see back up slides for ROI calculation Pre-roll versus TV by month and audience) 4 Apply frequency capping particularly on sites with very frequent visitors (e.g. social sites). Otherwise campaigns build an extreme amount of contacts with a relatively small part of the target audience.