Reaching Qualified Searchers With Google Ads Audience Targeting: A Layered Approach - Mary Hartman (MnSummit 2019)
•Download as PPTX, PDF•
1 like•416 views
In-Market and Affinity Audiences. Customer Match. Custom Intent. Demographic targeting. Google’s catalog of audience targeting options has exploded over the past couple of years. In this talk, we will cover how to make Google Ads’ ever-growing audience targeting options work for you on a layer-by-layer basis. Attendees will be given strategies for simple tests that can be rolled out right away, to campaign overhauls through data analysis and audience discovery. We will introduce new ideas for new audience layers including remarketing and customer match approaches. We will discuss some examples of tests gone wrong, such as the common problem of limiting audiences so severely that they halt spend altogether. And we will discuss how to overcome these key pitfalls for those who are just starting to use Google audience targeting in earnest across their search ad campaigns.
Report
Share
Reaching Qualified Searchers With Google Ads Audience Targeting: A Layered Approach - Mary Hartman (MnSummit 2019)
- 1. Reaching Qualified Searchers with Audience Targeting: A Layered Approach
- 2. • The Basics: • Sources • Key Benefits • Types of Targeting Available • Research: • Identifying Your Own Audience Testing Opportunities • How to Organize a Test: • Segmentation & Strategy • Common Pitfalls and how to avoid them Agenda
- 3. About Me: Mary Hartman @PPCHartman ● Industry Speaker ● Senior Account Manager ● Follow me @PPCHartman
- 4. Sources of Audience Data 1) Things Users Do while using Google properties ● Searches ● Website visits ● Ads clicked ● Videos viewed ● Location ● Device
- 5. 2) Things Users Create while using Google properties ● Emails sent and received via Gmail ● Contacts added ● Calendar events ● Photos & videos uploaded ● Docs, sheets, slides on Google Drive Sources of Audience Data
- 6. 3) Places Users Visit Online, where they confirm their identity. ● Demographic information gleaned from social media sites visited and attached to the user’s unique cookie. ● Demographic information gathered when a user sign up for a Google account: Birthday, gender, phone number, country, email, name. Sources of Audience Data
- 7. Benefits of Layering Search Audience Targeting with your Keywords @PPCHartman 1. Optimize: Reach “hotter” ready-to-buy audiences 2. Refine: Exclude unwanted traffic 3. Test: Expand broad match keyword targeting in a low risk way
- 8. @PPCHartman SEARCH AUDIENCE TARGETING IS THE FUTURE
- 9. @PPCHartman “The key to success in 2019…will be to create a detailed strategy of the various audience types and audiences lists and how you can layer them (with positive and negative bid types) to shape your paid search strategy.” -Christi Olson Head of Evangelism for Search,
- 10. “Customers aren’t keywords - They’re people” @PPCHartman -Mark Irvine Director of Strategic Partnerships, Named Most Influential PPC Expert of 2019
- 11. TYPES OF SEARCH TARGETING AVAILABLE @PPCHartman
- 12. In-Market Audiences @PPCHartman Ready-to-Purchase Audiences (mid-funnel searchers, browsers, media consumers) • Based on what users are actively researching and planning + =
- 13. Custom Intent Audiences @PPCHartman Ready-to-Purchase searchers across multiple platforms ● Used for YouTube ads only
- 14. Detailed Demographic Audiences @PPCHartman Audiences Based on Who Users Are ● Age ● Gender ● Parental Status ● Marital Status ● Education ● Homeownership status ● Life Events: Only available for YouTube, Gmail ads targeting ● ** Note that on average, 40% of Google Ads demographic info is “unknown”
- 15. Retargeting Audiences @PPCHartman Online interactions with your Business ● Website visitors ● YouTube users (Channel viewers / viewed a video from your channel. Membership duration is adjustable) Beyond Simple Cookie-Driven Retargeting ● Customer Lists: Email match, based on what you know about your users, 1st party data ● Similar Audiences: Allows Google to expand targeting
- 16. Use Google Analytics to Create Audiences @PPCHartman Import behavior-driven audiences into Google Ads by other channel interactions
- 17. @PPCHartman
- 18. IDNTIFYING YOUR AUDIENCE TESTING OPPERTUNITIES @PPCHartman
- 19. Leverage Google’s Audience Insights @PPCHartman
- 20. Or… Check Google Analytics Data @PPCHartman
- 21. @PPCHartman Hold Back - Observe First!
- 22. @PPCHartman Hold Back - Observe First!
- 23. @PPCHartman While Observing, consider… 1. Are you maximizing audience depth? 2. Did you determine a point of statistical significance? • When does the test end? • How many clicks / conversions will secure your confidence in the observation? 3. Check your “surefire” converter stats vs. new audiences or prospecting audiences. • Prospect audiences on non-brand keyword searches • Retargeting audiences • All audiences on brand keyword searches Observation of audience behavior can drive strategy choices in CRO or other non-PPC marketing tactics!
- 24. Bid Modifiers @PPCHartman ● Aim to be more or less competitive compared to searchers not in the audience ● Take into account existing adjustments for geo, device, demographics. ● Ideal tests are with a 0% bid adjustment = clean data ● Determine increases or decreases based on previous results and value per conversion, your KPIs.
- 25. Bid Modifiers: Example Calculation @PPCHartman 5.6% ÷ 4.3% = (1.30%-1)*100 = 30.23% (Round up to 30%) Bid Adjustment Users aged 17-25 have a CVR of Compared to an ad group CVR of Subtract 1 and divide by 100 to get your bid adjustment %
- 26. Case Study: Challenges ● Leads being generated included college students, non-pro, and non-decision makers. ● Keywords: supply chain visualization, supply chain model, demand planning, logistics planning Solution ● After observing relevant in-market audiences, we increased bids for “ERP Solutions” enterprise software visitors: Result: 50% increase in quality conversions, and a drop in CPA over the course of 30 day test. @PPCHartman
- 27. SEGMENTING & LAYERING AUDIENCES: BEST PRACTICES @PPCHartman
- 28. SEGMENTING: Dos & Don’ts @PPCHartman Do Don’t Set to “Observe” setting initially • Use bid adjustments Use the “Target” setting without: • Having the data to back up your decision • Understanding potential overlap and dealing with exclusions • Being willing to take a cut in traffic Set targeting on the ad group level • Audiences may respond to different keyword groups Set targeting on campaign level • If you mix campaign & ad group targeting together, your changes will not post in Editor.
- 29. SEGMENTING: Dos & Don’ts @PPCHartman Do Don’t Create ”custom combination” audiences • Great for targeting multiple audience attributes at the same time. Assume having multiple audiences targeted means they are layered. They aren’t! • Automatic setting will be x audience OR y audience in the same ad group. Pick audience sizes that will spend your budget: • RLSA: 1,000 active users in last 30 days • This is why set to “observe” is important. Over-segment with too many “target” audiences • It can kill traffic. Avoid excluding “unknown” demographic traffic altogether – 40% of search traffic qualifies as ”unknown”
- 31. Case Study: Challenge ● Search efforts were resulting in low-quality leads from whitepaper offers. ● To improve lead quality, keyword targeting had been narrowed down over time to long-tail industry terms, reducing the amount of reach. Solution ● New test campaigns targeted “Enterprise Software” and “Network & Enterprise Security” In-Market audiences - Using “Target” audience setting, excluding consumers and younger demographic ages since decision-makers in IT tend to be older. ● Keywords in the new campaign were much broader (BMM and Exact match) than the norm, like ransomware data protection. Such a keyword would have been impossible to test in the past without drawing in loads of McAfee-type consumer searches. Result: Checking the client’s CMS, total qualified leads from Google Ads doubled, primarily from the new campaign test with more diverse keywords & the audience “safety net” put in place. @PPCHartman
- 32. Setting Up a “Safety Net” Search Campaign @PPCHartman Do Don’t Use the “Targeting” audience setting to narrow your audience. Keep keyword targeting as narrow as regular search campaigns, or you’ll risk not spending any budget at all. Test more daring keywords as broad match modified & exact match types. Don’t fail to cross-exclude keywords that are targeted in other campaigns with similar, bid- modified audiences. Avoid cannibalization! Monitor your SQRs carefully – do not rely upon the audience targeting 100% Fail to check mid modifiers that may skew test data. Personalize your ads to address the audiences & use pre-qualifying language. Use the same generic ad copy as regular campaigns in your “safety net” campaign.
- 33. Remember to Set Exclusions! @PPCHartman Keep your tests clean and tidy ● Be aware of broad match modified keywords / query overlap in other campaigns & exclude accordingly. ● Layering audiences with a “Targeting” setting will cut spend, so a natural solution is to open up to new keyword targets. ● Keep negative keyword lists strong, but review what’s necessary within the audience.
- 34. Closing Shout-Out to Microsoft (formerly Bing) Ads also provides audience testing: ● Custom & Remarketing audiences ● 361 total In-Market audiences ● Demographic targeting including company, industry, and job function from LinkedIn data! @PPCHartman
- 35. ● Share the love! Follow me @PPCHartman ● Link to the Deck: HanapinMarketing.com/
- 36. @PPCHartman Appendix: A Link to Google Ads Audiences
- 37. Bonus: Audiences Available only on Google Display, YouTube, Gmail ads Custom Affinity Audiences ● People with interests aligned with your brand ● Target by relevant keywords, URLs, apps, and places you select Custom Intent Audiences ● People actively researching products and services ● Target by in-market keywords Life Event Targeting ● Graduation ● Marriage ● Moving @PPCHartman
Editor's Notes
- For people who aren't signed in to their Google Account, we sometimes estimate their demographic information based on their activity from Google properties or the Display Network. For example, when people browse YouTube or sites on the Display Network, Google may store an identifier in their web browser, using a “cookie.” That browser may be associated with certain demographic categories, based on sites that were visited. Source: https://support.google.com/google-ads/answer/2580383?hl=en Example: Sarah's favorite hobby is gardening. Many of the gardening sites and blogs on the Display Network that she visits have a majority of female readers. Based on this, Sarah's browser (when she’s not signed in from her Google account) could be added to the "female" demographic category. As a result, Google may show Sarah ads from advertisers who have chosen to show their ads to women. When Sarah’s signed in from her Google account, Google may show her ads based on selections she made in her Ads Settings, including her demographic information.
- Reference: https://www.searchenginejournal.com/google-demographic-data-ppc-account/228931/#close How Google determines demographic information When people are signed in from their Google Account, we may use demographics derived from their settings or activity on Google properties, depending on their account status…. In addition, some sites might provide us with demographic information that people share on certain websites, such as social networking sites. For people who aren’t signed in to their Google Account, we sometimes infer their demographic information based on their activity from Google properties or the Display Network. So even if you aren’t signed in, Google may make educated guesses about your gender, age, location, etc.
- ”…What will continue to separate the best-in-class search marketers from the average Joes will be how audience data is are segmented and implemented via an audience targeting strategy…
- Source: https://www.wordstream.com/blog/ws/2016/09/12/demographic-targeting-for-search-campaigns
- Full Chart data is in the appendix of this presentation.
- The difference between in-market audiences and affinity/custom affinity is that the former is tied to more temporary behavior whereas the latter type is for getting to audiences based on their general interests and behavior. Remarketing lists for search ads (RLSA) audiences and similar audiences also show some indication of being close to conversion, but unlike those, in-market audiences may be entirely unfamiliar with your brand name and site – and they’re better for acquisition. In-market audiences use machine learning to analyze trillions of search queries and browsing activity to predict purchase intent. This means that you can actually target people based not only on their present search activity but also on recent browsing history. As a result, you can be sure to display ads that are relevant to audiences at the right time. Past browsing behaviors show the kinds of products and services that users are actively interested in, allowing you to reach people who have displayed high purchase intent signals. You, therefore, have access to audiences who are likely to be incredibly valuable to businesses, as they are precisely curated to be in the market for specific products and services, and so are expected to have a higher likelihood of converting compared to other audiences. For example, if a user is reading camera review websites and watching camera comparison videos, Google may deem them to be keen to purchase a camera and put them in the relevant in-market audience. Without needing keyword targeting or targeting based on interaction with the brand, advertisers can target users who demonstrate these kinds of signals. It’s an amazing illustration of what machine learning can do, and only a glimpse into what the future holds for search.
- Note that in-market audiences are being added without fanfare all the time. They were first made available for search advertisers in 2017 but a LOT of advertisers do not keep up with how detailed the segments have been getting. In-market audiences are users who have shown intent to compare and purchase a specific category of product. This can be valuable for marketers looking for high-intent shoppers that are looking for a particular item. However, this in-market intent is temporary and doesn’t paint an overall picture of a user. Time limited – High intent means that this audience has a fast churn. Expanded to search for the first time in 2018
- https://support.google.com/google-ads/answer/2580383?hl=en
- “Similar audiences helps you find people who share characteristics with your site visitors. By adding ‘Similar audiences’ to your campaign or ad group targeting, you can show your ads to people whose search behavior is similar to those of your site visitors. These people are more likely to be potential customers.” Essentially, Google will review your remarketing lists and work to create a list of users who exhibit similar search behavior. They’ll then put them in a list and let you target them.
- A 2014 study from Kenshoo showed that paid search audiences which were already exposed to Facebook advertising generated 30 percent more return on ad spend, and a 7 percent uplift in CTR. https://kenshoo.com/wp-content/uploads/2014/03/The-Kenshoo-Guide-to-Search-and-Social-Intersections.pdf
- Full Chart data is in the appendix of this presentation.
- A quick recap: audience depth is the spend through all audiences divided by total spend. Modified audience depth is the spend through all audiences with a bid modifier applied divided by total spend. Maximizing audience depth gives you more data points to optimize toward. Even if an audience has a poor CPA, applying it should still be considered a success because it lets you identify an opportunity for optimization. Research has shown that audiences have been chronically underused in the past, although benchmarks vary a lot from one industry to another.
- If you choose to show your ads to everyone without excluding any demographic groups, you can still place higher or lower bids for certain demographic groups. But, before making bid adjustments, it’s best to let the data accrue with a 0% bid modifier. For example, say the conversion rate for users in the 17-25 age group is 5.6 percent and the ad group conversion rate is 4.3 percent, your bid modifier would be 30.23 percent but this needs to be rounded up to the nearest whole figure i.e. 30 percent.
- Calculate your bid modifier by dividing the age conversion rate by the conversion rate of the ad group. Subtract the answer by one and then multiple by 100 to obtain a percentage. For example, say the conversion rate for users in the 17-25 age group is 5.6 percent and the ad group conversion rate is 4.3 percent, your bid modifier would be 30.23 percent but this needs to be rounded up to the nearest whole figure i.e. 30 percent.
- Full Chart data is in the appendix of this presentation.
- Always set audiences to “Observe” initially – not just for gathering data, but to set bid adjustments. Only use ”Target” setting if you are willing to take a cut in traffic, and the audience has been demonstrated to be extremely valuable. In special cases: Exclude audiences between ad groups for a clean test.
- Always set audiences to “Observe” initially – not just for gathering data, but to set bid adjustments. Only use ”Target” setting if you are willing to take a cut in traffic, and the audience has been demonstrated to be extremely valuable. In special cases: Exclude audiences between ad groups for a clean test.
- Full Chart data is in the appendix of this presentation.
- https://instapage.com/blog/adwords-demographic-targeting With the data from your bid adjustments, consider creating a unique “Target and Bid” campaign. If you’re simply layering different demographics bids into a single campaign, it’s considered “bid only” — meaning you pay more or less to target a select demographic, rather than exclusively showing ads to them. “Target and Bid,” on the other hand, allows you to only show ads to searchers who meet your demographic criteria, and to control the settings and ad spend separately from other demographics. Using the “Target and bid” strategy, ROI Revolution saw a 42% increase in conversion rate and 21% decrease in cost-per-acquisition in the first month the campaign was live.
- https://instapage.com/blog/adwords-demographic-targeting