Audience Intent Power Audience on Search May 3

Audience + Intent= Power Audience on Search May 3 2016

Proprietary + Confidential Agenda Intro to Google Search Audiences Audience + Intent Google Search Audience Products

Proprietary + Confidential Intro to Google Search Audiences Audience + Intent Google Search Audience Products

Proprietary + Confidential Audience at Google has robust audience products across all platforms t

Proprietary + Confidential Today, we will focus only on Audiences on Search t

Proprietary + Confidential Audience + Intent t Intro to Google Search Audiences Audience + Intent Google Search Audience Products

Proprietary + Confidential What is Audience + Intent? Your Data Customer Match Remarketing Lists for Search Ads (RLSA) t Google Data Similar Audiences Demographics for Search Ads + All the power of Google Search

Proprietary + Confidential Keywords are the most powerful intent signal available

Proprietary + Confidential But audience data had dramatically changed things Bidding up Bidding down For example, a gluten-free, vegan, organic granola company cannot afford to bid on the keyword ‘granola, ’ unless they know that the user searching for that keyword is a potential customer.

Proprietary + Confidential How we got here RLSA Reach higher-value audiences for more effective Ad. Words campaigns. 2013 DFSA (beta) Add Demographic signals and focus on a specific gender and age on Search. 2014 Shopping Remarketing Lists Bring the power of audience data to Shopping campaigns. 2015 Google Analytics RLSA Increase segmentation and list capabilities dramatically through integration with GA. Customer Match Help strengthen connections with your known customer base. 2016 Similar Audience (beta) Reach new users with similar search behavior.

Proprietary + Confidential Audience + Intent is valuable, and we can prove it 53% average CPA decrease for conversions from an RLSA user list vs. entire campaign 20% average increase in conversions rate 10% average increase in conversions for advertisers that use RLSA vs. advertisers that do not use RLSA

Proprietary + Confidential Google Search Audience Products t Intro to Google Search Audiences t Intent + Audience Google Search Audience Products

Proprietary + Confidential Remarketing Lists for Search Ads (RLSA) Reach and prioritize past site visitors with RLSA t To implement RLSA, you must place a RMKT/GA/Container tag on your website. Users already familiar with your site, products or brand are 20% more likely to convert

Proprietary + Confidential Improve performance with RLSA + DSA Virgin Experience Days increased ROI by 34% using DSA with RLSA Compared to DSA campaigns not using RLSA, DSA+RLSA delivered: 281% increase in CTR 230% 13. 7% t increase in conversion rate increase in total flights booked

Customer Match Use your offline / CRM data to target customers Reach new and existing customers across devices when they want to hear from you. . . Just Purchased t Lapsed Customer Loyal Buyer Contract Expiring Enrolled in Contest Created an Online Account . . and deliver the right message to build relationships and drive action.

Proprietary + Confidential Customer Match Use your offline / CRM data to target customers “ Customer Match empowered us to tailor our messages to both our clients and prospective clients. The strategy has increased our revenue and decreased our cost per sale. ” — Didem Namver, Digital Marketing Manager, Pegasus Airlines Customer Match proving to perform! 13% Lower cost per acquisition 12% Higher revenue per conversion 52% t Higher Clicking conversion rate 32% Higher Clickthrough rate

Proprietary + Confidential Improve performance by optimizing search campaigns for all audiences The combination of all audience data will give you the largest reach and impact Remarketed traffic Similar User data This box represents all of your search traffic Customer Match traffic (CRM) Your Data Demographic data Google Data No Audience Data

Proprietary + Confidential Vertical level example: Retail

Proprietary + Confidential RLSA is already making a big difference for Retail Text Ads 56% average CPA decrease for conversions from an RLSA user list vs. entire campaign Shopping Campaigns Ads 35% average CPA decrease for conversions from a Search Audience list vs. entire campaign Source: Google Internal Data - May 2015

Proprietary + Confidential 3 Q Digital Search Audience Success

RLSA Strategies @ 3 Q Amplification - Boost bids for groups more likely to convert. Layer with audience data. - Customer Match- Upsell to existing customers, loyalty/retention assist (esp. on competitor terms) - Correlate CVR to GA segments (Time on Site, Page Depth) to access KWs otherwise not affordable - Combine with Dynamic Search ads, Broad Match KWs for relevantly testing into scale A Harte Hanks Company 21

RLSA Strategies @ 3 Q ▪ Suppression - ▪ Customer Match- remove existing customers from search targets (hello, brand campaigns) Decrease bids for groups less likely to convert. Layer with audience/demographic data. Conquesting - Aggressively position on competitive terms in research deep verticals Leverage GA lists based off research deep pages (ie: features, pricing, etc) Target KWs only your competition does with purpose Very deliberate ad copy for those shopping around. Be creative A Harte Hanks Company 22

Customer Match ▪ Leading online fitness retailer targeted CRM list of past purchasers who had not engaged on the site in 6 months ▪ Resulted in a 76% increase in CVR relative to average brand non brand performance A Harte Hanks Company 23

Customer Match + Similar Users In GSP ▪ ▪ ▪ 180% Increase in ‘Opportunities’ 82% increase in Lead to Opportunity rate 22% Decrease in CPA A Harte Hanks Company 24

Proprietary + Confidential Thank You

Proprietary + Confidential Practitioners guide to search audience products t Intro to Google Search Audiences t t Audience +Intent Google Search Audience Products Practitioners guide to search audience products

Proprietary + Confidential Combine audiences to improve performance Google’s audience types should be complementary when used together People who’ve interacted with your brand Customer Match traffic (CRM) People who’ve interacted with your website Remarketed traffic Customer Match traffic (CRM) People who’ve yet to interact Similar User data Demographic data No audience data yet No Audience Data

Proprietary + Confidential Combine audiences to improve performance There are many different list types now. How should I implement them and how do they work together? Add all audiences to all Search campaigns as bid only, and bid based on value. List Example Bid adjustment Clicks Conversion s CPA RLSA Cart Abandoners +100% 10, 000 5 Similar to: RLSA Cart Abandoners +50% 8000 300 7 Customer Match: Lapsed purchasers (>180 days) +10% 2000 80 5 Demo: Female +5% 1000 50 8 • Ad. Words will only consider the most valuable list eligible in the auction. • Ad. Words determines list value by the bid adjustments set for each list. • Use initial bid adjustments in accordance to your best estimate of value, and then adjust as you go.

RLSA Strategies @ 3 Q Amplification - Boost bids for groups more likely to convert. Layer with audience data. - Customer Match- Upsell to existing customers, loyalty/retention assist (esp. on competitor terms) - Correlate CVR to GA segments (Time on Site, Page Depth) to access KWs otherwise not affordable - Combine with Dynamic Search ads, Broad Match KWs for relevantly testing into scale A Harte Hanks Company 29

RLSA Strategies @ 3 Q ▪ Suppression - ▪ Customer Match- remove existing customers from search targets (hello, brand campaigns) Decrease bids for groups less likely to convert. Layer with audience/demographic data. Conquesting - Aggressively position on competitive terms in research deep verticals Leverage GA lists based off research deep pages (ie: features, pricing, etc) Target KWs only your competition does with purpose Very deliberate ad copy for those shopping around. Be creative A Harte Hanks Company 30

Customer Match ▪ Leading online fitness retailer targeted CRM list of past purchasers who had not engaged on the site in 6 months ▪ Resulted in a 76% increase in CVR relative to average brand non brand performance A Harte Hanks Company 31

Customer Match + Similar Users In GSP ▪ ▪ ▪ 180% Increase in ‘Opportunities’ 82% increase in Lead to Opportunity rate 22% Decrease in CPA A Harte Hanks Company 32