As announced on November 1, 2022, similar audiences will be transitioned to more durable solutions. Starting May 1, 2023, there will be changes that will affect the way similar segments are used in Google Ads. Learn more about these Changes to audience targeting.
To provide a comprehensive and consolidated view of your Audiences and make audience management and optimization simpler, you’ll find the following improvements in Google Ads:
- New audience reporting
Detailed reporting about audience demographics, segments, and exclusions is now consolidated in one place, the “Audiences” tab within the left page navigation menu. You can also easily manage your Audiences from this report page. Learn more about Audience reporting - New terms
We’re using new terms on your audience report and throughout Google Ads. For example, “audience types” (these include similar, custom, in-market, and affinity) are now referred to as audience segments and “remarketing” is now referred to as “your data”. Learn more about the updates to Audience terms and phrases
Similar segments is a targeting feature that uses your data segments to help you expand your reach to new potential customers who have characteristics similar to your existing customers or people who have visited your website or used your app.
Before you begin
Similar segments targeting is available for the Display Network, the Search Network, YouTube, Gmail, Customer Match, and apps. Similar segments are automatically created and updated in real time when you have set up at least one eligible list in the Google Ads Audience Center library. It's recommended that you have your data strategy in place before running Similar segments. Learn more about targeting your data segments
Benefits
- More reach: Extend the reach of your campaign towards new users who are similar to your existing users, up to 5 times more than with your data segments alone.
- Improved conversions: Achieve 41% more conversions by combining similar segments with Display ad campaigns.
- Powered by machine learning: Similar segments are updated in real time using machine learning capabilities that analyze millions of signals to optimize towards performance.
- Easy to manage: Similar segments automatically appear in eligible Google Ads accounts. Eligibility is determined when there is enough statistically significant similarity found between new users and your website and app visitors.
- Target efficiently: Your existing customers and people who have visited your website or used your app are automatically excluded from Similar segments, making sure you target new users only.
Simplified audience targeting
Similar segments targeting takes the guesswork out of your search for new audiences by automatically finding new customers similar to your existing customers (such as ‘site visitors’, or ‘converters’’). Google Ads takes your existing user lists and uses machine learning to understand what they have in common to find more like-minded customers. You can use Similar segments to find new customers similar to your existing customers using criteria such as topics they like and sites they frequent.
Your Similar segments are automatically updated as your original customer list evolves and people change their browsing activity. So, you don't need to update the Similar segments when it's created.
Getting new potential customers
With Similar segments, you can boost the reach of existing campaigns created with the help of your data, and drive new users to your site who are also likely to be interested in your product or service.
Even if people in your Similar segments don't immediately purchase from your site, if you've added the Google tag to your site, and they visit your website or use your app, you can still reach them in the future. By adding Similar segments targeting, you can get more visitors to your website, and therefore more users added to your data segments in a way that's intuitive and easy to manage.
Case study: ZOOT
ZOOT is a Prague-based clothing retailer that aimed to increase audience prospectives and get more out of their campaigns. To achieve these goals, they implemented Similar segments and Dynamic Display Ads. The Dynamic Display Ads showed products that users were interested in, resulting in more purchases and an overall doubled return on investment when compared to their previous campaign settings. ZOOT used Similar segments to reach new users that their previous campaigns hadn't before. The Similar segments proved so effective that new users provided 63% of the ZOOT's website traffic over the course of the campaign.
How it works
Google Ads looks at browsing activity on Display Network sites over the last 30 days, and uses this in addition to the content the user is viewing to understand the shared interests and characteristics of the people in your data segments. Based on this information, Google Ads automatically finds new potential customers whose interests and characteristics are similar to those of people who visited your website, used your app, or are on your existing customer lists. The more characteristics and interests people who have visited your website and use your app share, the better Similar segments work.
To find Similar segments, Google Ads looks across the millions of apps and sites on the Display Network. As your data segments change, your Similar segments will change as well. Similar segments are created using machine learning and are built to adapt to changing markets and trends dynamically. These models find high-performing users who are similar to existing customers.
Qualifying for Similar segments
Google will automatically identify which of your data segments qualify for Similar segments based on a variety of factors, including but not limited to the following:
- Minimum of 100 visitors in the seed list (your data segments)
- How recently these visitors joined the original segment
- The types of sites that these visitors browsed
- List must be owned by your account, not shared
Many types of your data segments can be used to generate Similar segments, including but not limited to: Website visitors, app users, customer match, YouTube users, CRM, Google Analytics, and more. Custom lists aren't eligible. In some cases, there isn't enough information to generate a Similar segments. In these cases, the list may appear as “Unavailable” in the Audience manager.
Similar segments are created based on users that share the same characteristics as seed lists without any geolocation restriction. Which means that if a business is based in Canada for example, the users in the similar segments won’t be restricted to Canada only, they’ll be from different locations globally.
The list size available in the audience manager will show coverage for all regions possible. When creating a new campaign, apply geolocation restrictions if you want reach estimate only for a limited selection of desired locations.
Keep in mind
We don't support finding Similar segments for lists based on sensitive categories, such as race, religion, sexual orientation, or health. All use of Similar segments is subject to our Personalized advertising policy. People who don't want to be included in Similar segments can opt out through their Ads Settings.
It may take up to 72 hours for your list to start serving.
Best practices for Similar segments
- Use automated bidding: Similar segments cost per conversion can be up to 25% higher than targeting website visitors and app users, and existing customers since Similar segments target new users. Leverage tCPA or tROAS for better performance results. If the main success metric is "conversion," use tCPA. eCPC is a bidding model designed to optimize for clicks, rather than for conversions.
- Make your ads Responsive: Leverage Responsive ads design wherever possible so that your ads target the best customers with the closest accuracy. Try variations on 'call to action' text, images, etc.
- Modify bids: Depending on your campaign performance, a higher bid can help you stand out or a lower bid can save on your budget. Smart Bidding raises or lowers your bids for you with impression-based adjustments.
- Geolocations & languages: Evaluate whether you should make any changes to the geo & language due to particular aspects of your business. For instance, if your business only operates in specific locations or if your site is only available in certain languages.
- Communicate with third parties: If you find performance issues with your campaigns and use any third-party components, use the third party's support to check if they are working on any errors.
- Campaign structure: It is possible to add Similar segments to an existing campaign designed around your data in a separate ad-group. However, it's recommended to have a separate campaign for better control of budget, messaging and measurement.
- List strategy: Think through your data segments strategy. Your strategy will influence what kinds of Similar segments the system generates. Ideally, you should have your plans and goals in mind before setting up your initial data segments. It's recommended to start a campaign with your most successful segments. For instance, a "similar to converters” list or the ones with the largest volume of users.
Common issues with Similar segments
Not enough reach (or similar segments too small)
- Try increasing your bid and checking your impression score metrics. Low bidding can lose impressions while high bidding can use up your budget too quickly. Low budgets can restrict how often your ads show while high budgets can make campaigns expensive.
- Try adding an additional, or a bigger Similar segment to your targeting. You may have selected a segment that is too small.
Similar segments not generating
- Similar segments can take at least 48 hours to generate after the seed list is created.
- Your small data segments (<500 users), are less likely to generate Similar segments.
CPA too high
- Try to reduce your target CPA or bid. The recommended way to reduce CPA is to wait for the learning period to finish (about 2 weeks) and then start reducing 20% week over week following up on how the list performs. If your bid is already low, try raising it. Sometimes a low bid only gains you access to low quality inventory.
- Try new ad creatives and/or landing pages - perhaps your ads don't have enough information or aren't convincing enough to push sales. Keep in mind that every time you change creatives, the algorithm will need some time to adjust to the change and find new patterns.
- When planning for how you’re going to find new users, think about which attributes of those users are most important for your campaign as well as how much emphasis you want to put on view-through conversions.