9. Selecting the Right Segments is the New Challenge Trial and error method is a slow and cost prohibitive process across dozens of data providers delivering 10,000s of segments
10. Large Data Repository for Multifaceted Audience Analytics Syndicated Segmentation CPG: Loyalty Card Data CPG: Transaction Data Allergies Analgesics Baby Products Carbonated Beverage Diet Carbonated Beverage Regular Cereal Children's Cereal Children's Food & Products Coffee Cookie Cosmetics Cough & Cold Dieters Energy Drink Frozen & Packaged Food HH Cleaners Natural & Organic Petcare - Cat Petcare Buyers - Dog Premium Brand Salty Snack Sports Drink Value Haircare Vitamin Buyers Automotive Baby Products Beauty & Fragrance Big & Tall Apparel Cat Products Children's Apparel Children's Products Consumer Electronics Corporate Attire Corporate Men Corporate Women Cyclists Dog Products Fine Jewelry Fishing Fitness Flowers Furniture & Accessories Gardening Geriatric Supplies Gift and Flowers Gifts & Cards Golf and Tennis Health & Wellness High Fashion/Luxury Apparel Hiking & Camping PRIZM P$YCLE ConneXions DataLogixLifeStyles Home & Garden Home Entertaining Home Improvement Home Organization Home Renovation Hunting Jewelry Mens Jeans Men's Fashion & Apparel Natural Wellness Outdoor Sports Pet Supply Plus Size Apparel Runners Senior Fashions Senior's Products Small & Home Office Tools Toys Weight Loss and Supplements Winter Sports Women's Accessories Women's Apparel Women's Shoes Demographics Age Gender Family Composition Occupation College Grad Married Children Grandchildren Dwelling Type Length of Residence Financials Household Income Net Worth Credit Worthiness Credit Card Type Financial Services Investments Insurance Auto Car Brand Affinity Motorcycle Brand Affinity Truck Brand Affinity Vehicle Age Vehicle Budget Politics Geographic Voter Party Affiliation DMA
11. Large Data Repository for Multifaceted Audience Analytics 3rd Party Data Syndicated Segmentation CPG: Loyalty Card Data CPG: Transaction Data Allergies Analgesics Baby Products Carbonated Beverage Diet Carbonated Beverage Regular Cereal Children's Cereal Children's Food & Products Coffee Cookie Cosmetics Cough & Cold Dieters Energy Drink Frozen & Packaged Food HH Cleaners Natural & Organic Petcare - Cat Petcare Buyers - Dog Premium Brand Salty Snack Sports Drink Value Haircare Vitamin Buyers Automotive Baby Products Beauty & Fragrance Big & Tall Apparel Cat Products Children's Apparel Children's Products Consumer Electronics Corporate Attire Corporate Men Corporate Women Cyclists Dog Products Fine Jewelry Fishing Fitness Flowers Furniture & Accessories Gardening Geriatric Supplies Gift and Flowers Gifts & Cards Golf and Tennis Health & Wellness High Fashion/Luxury Apparel Hiking & Camping PRIZM P$YCLE ConneXions DataLogixLifeStyles Home & Garden Home Entertaining Home Improvement Home Organization Home Renovation Hunting Jewelry Mens Jeans Men's Fashion & Apparel Natural Wellness Outdoor Sports Pet Supply Plus Size Apparel Runners Senior Fashions Senior's Products Small & Home Office Tools Toys Weight Loss and Supplements Winter Sports Women's Accessories Women's Apparel Women's Shoes Demographics Age Gender Family Composition Occupation College Grad Married Children Grandchildren Dwelling Type Length of Residence Financials Household Income Net Worth Credit Worthiness Credit Card Type Financial Services Investments Insurance Auto Car Brand Affinity Motorcycle Brand Affinity Truck Brand Affinity Vehicle Age Vehicle Budget Politics Geographic Voter Party Affiliation DMA
14. ... and clicks a banner… A user visits an Auto-Magazine website ... ... to land on the Volkswagen website Surf behavior clicks Proprietary data base, agreements with publishers, search etc Look alike modelling to extrapolate fact based data points across whole internet audience Client site browsing behaviour Googlekeywords cliced surf behavior on client‘s site
15. 2nd Party Data 1st Party Data ... and clicks a banner… A user visits an Auto-Magazine website ... ... to land on the Volkswagen website Surf behavior clicks Proprietary data base, agreements with publishers, search etc Look alike modelling to extrapolate fact based data points across whole internet audience Client site browsing behaviour Googlekeywords cliced surf behavior on client‘s site
16. 16 Second party and look alike modelling, plus Third Party Awareness First party, client specific data Engagement Any data that drives and enhances performance Action Confidential | Xaxis Capabilities | June 2011
17. The stupid EU cookie Directive in 2 ½ minutes http://www.youtube.com/watch?v=arWJA0jVPAc Presentation Title | Date 17
18. Presentation Title | Date 18 Step 1: OBA ad appears on a website, including Icon (bottom left in this case) Step 2: User clicks on icon, receives a dropdown box of basic information and options Step 3: User clicks on ‘More Information and Opt-Out Choices’ and is taken to full www.youronlinechoices.com site, with extensive information about OBA and data use Step 4: User clicks on ‘Your Ad Choices’ and s provided with ability to opt-out of OBA targeted ads
22. Who We Are A GroupM company helping agencies and their clients use data and technology to reach and engage with audiences at scale 22 Confidential | Xaxis Capabilities | June 2011
Notes de l'éditeur
Planners (used to) buy an audience through data provided by the publisher about their audience. If I’m a car advertiser (in this case Lexus), I would place my ad on a car site because people on car sites tend to be interested in cars.Nothing wrong with this approach, but there is waste (interest in cars does not necessarily = being “in the market” for car / let alone a lexus (expensive / luxury).
Today, we want to target audiences directly. Using all the data we have (cookies, third party data etc), we can determine an audience based on key characteristics that would indicate that that person is high net worth and in the market for a car etc.We will then use my technology to ensure that my ad appears when that person is surfing. If we am doing this effectively, we care less (though we still care) about what actual site they are on. The audience drives the value of that impression, not the site.
Today, we want to target audiences directly. Using all the data we have (cookies, third party data etc), we can determine an audience based on key characteristics that would indicate that that person is high net worth and in the market for a car etc.We will then use my technology to ensure that my ad appears when that person is surfing. If we am doing this effectively, we care less (though we still care) about what actual site they are on. The audience drives the value of that impression, not the site.
Planners (used to) buy an audience through data provided by the publisher about their audience. If I’m a car advertiser (in this case Lexus), I would place my ad on a car site because people on car sites tend to be interested in cars.Nothing wrong with this approach, but there is waste (interest in cars does not necessarily = being “in the market” for car / let alone a lexus (expensive / luxury).
So what do we need to enable this approach to buying?The first thing is data.
But what data? There are so many providers and so many segments (especially in the US, but also increasingly in Europe), we need to understand how to specify the segment in a way that Allows us to differentiate between the relevance and likely performance of segments and therefore which ones to buy AND Negotiate the price for that data (usually quoted as a CPM). All data loses its value at some point as the price for that data increases
To ensure that the right ad is in front of the right audience at the right time and at the right price we also need to have the technology to engage with those audiences and place the ad.
In the US, the 72% of display spend goes to the top 10 publishers and 92% goes the top 50.