Build Retargeting Audiences From Your Data

Most retargeting audiences are built by hand. Someone in growth opens the CRM, filters for "trial users," exports a CSV, and uploads it to Facebook Ads Manager. Then they do it again for Google. A week later the list is wrong: half those trials converted, a few churned, and the new signups never made it in. The ads keep running against a snapshot of who your customers were last Tuesday.

The data to build accurate audiences already exists. It's just scattered. Subscription status and cart abandonment live in your billing tool. Trial stage, feature adoption, and last login sit in your product database. Deal stage and churn flags are in the CRM. Pull those signals into one place and the audiences build themselves from live data instead of stale exports.

What retargeting audiences are and which customers to retarget

A retargeting audience is a list of people you already know, uploaded to an ad platform so you can show them ads. Retargeting re-engages people who showed intent but didn't finish: a trial user who never activated, a free-plan user who hit a limit, a deal that went quiet. It's distinct from remarketing, which leans toward winning back existing or lapsed customers, though the two overlap enough that people search "remarketing retargeting" as if they're one thing and use the terms interchangeably.

It's also distinct from lookalikes. If you want to expand reach to brand-new prospects who resemble your best customers, that's a different workflow. We cover it in the lookalike audiences guide and won't repeat the seed-list mechanics here. Retargeting points the other direction: people already in your data who haven't converted yet.

The segments worth building usually fall into a few buckets:

• Trial and free users who stalled. Signed up, poked around, never reached the activation moment. High intent, low effort to convert.

• Cart and checkout abandoners. Got to the payment step and bailed. The classic retargeting use case, and one of the highest-converting.

• Open deals that went cold. A deal in your CRM stuck at "proposal sent" for three weeks is a retargeting candidate, not just a sales follow-up.

• Recently churned accounts. Canceled in the last 30 days. A win-back audience with messaging that references what they used works far better than a generic "we miss you" ad.

You don't need all of these. Pick the two or three that map to where you actually lose people.

What customer data to sync to build retargeting audiences

The whole approach depends on getting the right fields into one place. An ad platform can only match on identifiers like email and phone, but the segment logic that decides who goes into which audience runs on your business fields. Those fields come from different systems.

Sync these onto a single set of contact records keyed on email. Now a segment like "free plan, logged in 5+ times, no payment" is one filter instead of a join across three tools you can't actually join. That segment is your upgrade-retargeting audience.

One field deserves a callout: whatever marks a customer as converted. Active subscription, closed-won deal, completed purchase. You'll use it twice. Once to build retargeting audiences (exclude converters from the trial audience) and once to build suppression audiences (stop advertising to converters entirely). Keeping that field fresh is what separates a retargeting setup that wastes money from one that doesn't.

How to build retargeting audiences and sync them to Facebook and Google Ads

The setup runs about 30 minutes. Most of the time goes into deciding which segments matter, not the mechanics.

1. Connect your sources. In Oneprofile, add your CRM, billing tool, and product database as sources. Authenticate each with OAuth or an API key. Oneprofile reads the available fields from each tool at connect time, so you map from a real list, not a guess.

2. Sync the segment fields onto one set of records. Map billing, product, and CRM fields to a shared contact record using email as the matching key. For fields that don't exist yet in the destination, Oneprofile creates the property automatically with the right type. Property-level change tracking means only the fields that actually changed get written, so a status flip doesn't trigger a full record rewrite.

3. Define your segments. Build each retargeting audience as a filter on the synced fields. "Trial users who stalled" might be trial_stage = started and features_activated < 2 and subscription_status != active. Save it as a named segment.

4. Connect Facebook and Google as destinations. Add Facebook Custom Audiences and Google Customer Match. Map the segment's email and phone fields to the audience. Both platforms hash identifiers on their side before matching against their user base, so you're not shipping raw PII into an ad account.

5. Choose a sync schedule. Daily is the right cadence for most of these segments. Run it more often if your conversion volume is high enough that same-day staleness costs you. Each run pushes the current segment membership, so people who converted overnight drop out on the next sync.

6. Run the first sync. The initial push populates each audience in Facebook and Google. From there it's incremental: only records that entered or left the segment get sent.

Funnel-based retargeting audiences for each customer stage

Blanket retargeting (one ad to everyone who didn't convert) leaves a lot on the table. Funnel-based retargeting splits the audience by where the person actually is and shows each group a different ad.

Because the segment logic runs on synced product and billing fields, you can be specific about stage in a way that pixel-only retargeting can't. A pixel knows someone visited your pricing page. Your synced data knows they're on a free plan, activated three features, and logged in yesterday. That's a far better signal for an upgrade ad.

A workable funnel split for a SaaS team:

• Top of funnel: signed up, never activated. Ad reinforces the core value and links to a setup guide.

• Middle: activated, on a free plan, hitting usage limits. Ad pushes the upgrade with a specific benefit the paid plan adds.

• Bottom: trial expired without converting, or deal stalled late. Ad offers a nudge: a call, a discount, a deadline.

Each stage is its own segment, its own audience, its own campaign. When someone's product behavior changes, the daily sync moves them from one audience to the next without anyone editing a list.

Suppression audiences and smarter remarketing campaigns

The same synced data that builds retargeting audiences also builds the list of people you should stop paying to reach. An ad suppression audience is an exclusion list: existing customers you don't want acquisition campaigns spending budget on.

Sync every active paying customer to a separate Facebook and Google audience, then set that audience as an exclusion on your prospecting campaigns. The payoff scales with freshness. Someone converts today, the suppression list refreshes tonight, and tomorrow's acquisition spend skips them. With manual CSV uploads the suppression list is stale within days and you're paying to advertise to people who already bought.

Suppression and retargeting are two sides of the same setup. The retargeting audience says "this person hasn't converted, show them an ad." The suppression audience says "this person did convert, stop showing them the acquisition ad." Both read from the same converted flag. This is also where remarketing campaigns aimed at existing customers come in: a separate audience of current customers for cross-sell or loyalty messaging, distinct from the acquisition ads you're suppressing them out of.

Retargeting technology has historically leaned on the pixel and the cookie to do this. As third-party cookies fade, audiences built from first-party customer data you already own are more durable than tracking-based ones. The data lives in your tools, not in a browser.

What changes after your retargeting audiences go live

The manual export ritual stops. No more Monday CSV pulls for each platform. When a trial user activates, a deal closes, or a customer churns, the relevant audience updates on the next sync and the ad spend follows the change.

A few things to keep an eye on once it's running:

• Match rates. Facebook and Google only match the identifiers they recognize, typically 40-70% of a list. If yours is lower, check email formatting at the source. Synced fields tend to match better than hand-maintained ones because they come straight from the system of record.

• Minimum audience sizes. Both platforms enforce a floor (Facebook needs at least a few hundred to a thousand matched users). A tightly scoped segment can fall below it. Widen the filter or combine adjacent stages if an audience is too small to activate.

• Failed records. When a record fails to sync, Oneprofile surfaces it instead of dropping it silently. Fix the cause, reprocess, and it lands on the next run.

One honest caveat: this approach is built for customer-data retargeting, the audiences you can define from CRM, billing, and product fields. It doesn't replace site-level pixel retargeting for anonymous visitors who never gave you an email. If anonymous-visitor retargeting is your main channel, you still need a pixel for that piece. The synced-data approach covers the known-customer audiences the pixel handles poorly.