How to Create a Lookalike Audience from Subscriber Events

How to create a lookalike audience starts with an Airbridge paid-subscriber seed, synced to an ad platform that decides eligibility and builds the lookalike.

How to Create a Lookalike Audience from Subscriber Events

Build a lookalike seed from paying subscribers.

  • Define the seed with an event that means a customer became a paid subscriber.
  • Keep trial starters in a separate audience when the campaign needs trial users.
  • Build and save the event-based audience in Airbridge, then connect it to a supported ad platform.
  • Let the destination platform decide whether the received audience qualifies as a lookalike seed.
  • Airbridge updates connected audiences daily; its send-size guidance is 1,000+ for Meta Ads and TikTok for Business, and 5,000 for Google Ads.

How to create a lookalike audience: define a seed from users who completed a paid-subscription event, sync it to a supported ad platform, then build the lookalike in that platform. Airbridge Audience Manager handles the event-based segmentation and, on Core Plan, audience sync to Meta Ads, Google Ads, and TikTok for Business.

A first paid subscription, a trial start, and a renewal are different customer states, so use the event that matches the campaign's intended audience. A trial start, a first paid transaction, and a renewal describe different customer states. Choose the event that fits the campaign goal, then keep the platform handoff separate from conversion measurement.

Conversion feedback and subscriber seeds do different jobs

A conversion event tells an advertising system that a chosen outcome happened. A subscriber seed is a group of existing users, selected because they reached that outcome, which the destination platform can use for a separate audience action.

Airbridge provides the seed-building step: its Audience Manager can create audiences from attribution events such as subscribers, trial starters, churned users, and high-value users. Core Plan supports automatic audience sync to Meta Ads, Google Ads, and TikTok for Business.

For web conversion tracking, Google enhanced conversions capture first-party details such as email, name, address, or phone in conversion tags, hash that information, and use it to match customers to Google accounts they were signed into when engaging with an ad. When an advertiser opts in, Google conversion-based customer lists use hashed first-party data supplied for enhanced conversions to build customer lists. Google requires a tag-based enhanced-conversions implementation for conversion-based customer lists.

Choose the event that represents a paying subscriber

Use a collected event that corresponds to paid subscriber status, and select the matching event in Audience Manager. Keep the event definition aligned with how the app records billing state. If the app sends one event when a trial begins and a different event when a first payment succeeds, use the payment event for a paid-subscriber seed and the trial event for a trial-user seed.

A trial-start event can be a suitable source when the objective is to find prospects who resemble trial starters. It is a poor fit for a seed intended to represent customers who have actually paid. A renewal event can identify returning payers if that recorded event is the criterion the team wants to model.

Use separate audiences for paid subscribers and trial starters when the campaign needs to model those customer states separately. For example, a consumer app might create one segment of users who performed its paid-subscription event in the last 90 days, and a separate segment of users who started a trial in the last 30 days. A shorter window can emphasize recent behavior; a longer one can include more people when the seed is small.

Google recommends uploading available first-party customer data, including email, mobile device, phone number, and physical address, in the correct format to create the largest possible Customer Match list.

Google's Demand Gen guidance recommends high-intent seeds that closely match recent converters, such as frequent purchasers. For a subscription campaign, choose the paid-subscription event or a repeat-payer group when the collected events support that distinction. A larger trial audience answers a different prospecting question, even when it gives the campaign more seed members.

Build the event-based audience in Airbridge

An event must have been collected at least once before it can be used as an Airbridge audience condition. The Airbridge audience creation guide also requires Owner or In-house Marketer permission to use Audience Manager. Use this sequence to define a seed:

  1. Open Integrations > Audience Manager, then select Create an audience in the Audience or Request History tab.
  2. Choose whether users performed or did not perform an event. For a subscriber seed, choose the collected event that marks paid subscriber status.
  3. Add an event-property condition when a property helps isolate the intended subscription event. For example, if the collected event carries a property that distinguishes a subscription product or payment state, use that property to select the right records.
  4. Set event aggregation and date range. Airbridge supports event counts, sums, or unique days as aggregation criteria, along with a date range for when the event occurred.
  5. Review the matching-user estimate shown in the dashboard as conditions are added. Give the audience a clear name that states the customer state and time period, then save it.

Treat the audience filters as separate controls. The event rule selects people who performed or did not perform an action; an event property narrows by an attribute recorded with that event; aggregation measures the event by count, sum, or unique days; and the date range sets when it occurred. For example, a subscription app could start with users who performed its recorded successful-payment event during a recent period, then add a plan property if the campaign is specifically for one plan. The live matching-user count changes as conditions are added, so remove a property or aggregation rule when it excludes customers outside the campaign's intended audience.

Keep the audience definition as specific as the campaign question requires. A basic subscriber seed might use the paid event and a recent date range. A higher-value subscriber seed could add a collected event property or an aggregation condition that reflects the value rule the app already records. Every added filter narrows membership, so the estimate helps the team see how the chosen definition changes the available seed before sending it.

Use a separate audience for each state the growth team may want to activate. One audience can contain new paying subscribers; another can represent trial starters; a third might represent churned users for a win-back campaign. Naming the audience with its event meaning and date window helps the person managing ads choose the correct seed later.

If the paid event is missing from the selector, confirm that Airbridge has collected it at least once and that your role is Owner or In-house Marketer. If the audience count is lower than expected, review the performed-event choice, event properties, aggregation, and date range against the subscriber definition you intended.

Check the seed before sending it

Airbridge's audience delivery guide gives destination-specific recommended sizes as a delivery checkpoint in the table below. Airbridge notes that audiences below its recommended size may not be delivered properly.

DestinationAirbridge recommended send sizePractical use of the figure
Meta Ads1,000 or moreAim for at least 1,000 estimated users before connecting the seed.
Google Ads5,000Use 5,000 as Airbridge's recommended delivery checkpoint.
TikTok for Business1,000 or moreAim for at least 1,000 estimated users before connecting the seed.

Check three different counts before assuming the audience is ready. Second, account for Airbridge's rule that devices without a collected GAID or IDFA are excluded when audiences are sent to ad channels. Third, let the destination platform determine how many received users it can match and whether that audience qualifies for the requested campaign feature.

Airbridge's audience delivery guide estimates audience size by sampling 10% of matching users, so actual size can differ. To refresh the estimate, open the menu beside the audience in Audience Manager and select Recalculate.

Under Google's Customer Match policy, a list stays eligible when at least 100 members have been added or updated within the past 540 days, and membership lasts no longer than 540 days. For Customer Match targeting, Google requires 90 days of Google Ads history and more than USD $50,000 in lifetime spend; all policy-compliant advertisers can use Customer Match for Observation and Exclusions.

Treat subscriber records as customer data with a defined collection and use purpose. Google's Customer Match policy requires a privacy-policy disclosure when customer data is shared with service providers, and consent when applicable law or Google policy requires it. Use only customer data collected and shared under the permissions and notices that apply to the app and campaign.

Google's Customer Match policy limits uploaded customer information to data collected in a first-party context, such as through an app, website, or physical store, where customers shared it directly with the advertiser. In Customer Match best practices, Google explains that uploaded customer information is hashed and matched with codes of Google accounts; after matching, the codes are deleted immediately, and Google says it does not keep or use the data for another Google product. For a subscription app, use records collected directly from its own customers and keep the subscriber event definition aligned with the campaign's purpose.

Sync the seed, then create the lookalike in the destination

Airbridge sends audiences through an authenticated channel connection. The channel must be authenticated before sending, and an audience must already exist. After a connection is added, Airbridge starts sending the audience and updates it daily, per its audience delivery guide.

  1. In Audience Manager, open the saved subscriber audience and add the connection for the intended channel.
  2. Use the audience the destination reports as available when choosing a seed for its campaign.
  3. For Google Demand Gen, choose a customer list from the seed lists available in the campaign setup.
  4. Review campaign delivery and conversion reporting separately. The seed is the group used to guide prospecting; the subscription event remains the outcome to measure in the campaign.

Google's Demand Gen Lookalike guide allows one to 10 seed lists for a segment. Google lists customer lists, website visitors, app users, Google Analytics audiences, and YouTube audiences as seed types. To create a segment in a new Demand Gen campaign, add an audience at the ad-group level, choose New Audience, open the Lookalike Segment option, and select New Segment. Choose the seed list, select the countries or regions to target, and choose Narrow, Balanced, or Broad for the segment. Google recommends creating the campaign and segment 2 to 3 days before the planned start so the segment has time to collect enough data.

For a Google workflow, use a seed that Google presents as eligible in Demand Gen. Google's Demand Gen Lookalike guide says Lookalike segments refresh continually as activity updates.

Choose reach or similarity in Demand Gen

In Google's Demand Gen Lookalike guidance, Narrow aims to include 2.5% of users in the campaign's target location, Balanced aims for 5%, and Broad aims for 10%. Narrow prioritizes similarity to the seed, while Broad aims for more reach and may include people less similar to the seed. For a paying-subscriber seed, choose the setting that fits the campaign's balance between resemblance to existing payers and audience size.

Google also identifies Lookalike segment statuses as populating, complete, inactive, or failed. Its guide says a failed seed-list requirement can take up to 3 days to appear in the segment status, so use that status when diagnosing a newly created Demand Gen audience.

Airbridge supports audience sync to Meta Ads and TikTok for Business, and its audience delivery guide says connected audiences update daily. For these destinations, open the synced audience in the ad account and choose the audience options available for the campaign.

For a subscriber seed with a low usable count, widen a reasonable date range or create a separate seed from another meaningful paid-subscriber event. Keep trial users in a separate seed unless the campaign is intended to model trial behavior.

Quick readiness checklist

  • The selected event represents a successful paid subscription, and Airbridge has collected it at least once.
  • The audience uses an intentional date range and any event properties or aggregation rules have a clear purpose.
  • The saved audience estimate meets Airbridge's destination-specific recommendation where practical.
  • The connected ad channel is authenticated, and the audience has been added to that connection.
  • The app's collection notices and applicable platform policies permit the customer-data use.
  • The destination has received and marked the seed eligible for its lookalike workflow.
  • The campaign measures paid subscriptions as outcomes, separately from audience delivery.

FAQS

FAQ

How do I create a lookalike audience on Facebook?

Build an Airbridge audience from the paid-subscription event and sync it to Meta Ads. Then continue in Meta Ads with the seed and campaign options available in that account. Airbridge supports Meta Ads audience sync, updates connected audiences daily, and recommends a send size of at least 1,000; use Meta Ads' account workflow for the destination-side setup.

Can you do lookalike audiences on Google?

Yes. Google Ads offers Lookalike segments for Demand Gen campaigns, and Google's official guide lists customer lists as a seed type. Airbridge recommends a send size of 5,000 users for Google Ads.

What is a lookalike audience?

A lookalike audience is a prospecting audience a destination platform builds from an existing seed of users, such as a customer list. For a subscription app, a seed defined by paid subscribers gives the platform a different customer group to model than a seed of trial starters.

Get Started Free

See how these criteria hold up on the real thing.

Get Started Free