Clay logo, go to homepage
Clay GTM guide

What Is Clay Audiences? Centralize and Act on Your GTM Data

Clay Audiences is the unified data layer that connects your CRM, warehouse, signals, and 200+ providers into one always-current home for your GTM data.

June 23, 20269 min read

Your CRM was supposed to be the source of truth. But a record decays the moment it lands, and the rest sits scattered across a warehouse, a dozen tools, and exports nobody trusts. Worse, Clay tables capped at 50,000 rows, so anything bigger meant splitting the work. Audiences removes that ceiling.

You hold millions of records as live profiles and slice them into audiences with simple filters. Enrich and monitor at scale, then sync clean data back to your CRM. This guide explains what Audiences is, how it works, and what teams run on it.

What Clay Audiences actually is

Audiences turns data work from maintaining snapshots into querying a database that maintains itself. Most GTM data work is fighting decay: a rep cleans a list, it goes stale, someone re-exports it, and the cycle repeats across a dozen tables. Audiences replaces that with two global lists, All People and All Companies, that hold every record you import. Each record is one persistent profile combining CRM fields, warehouse data, signals, and enrichments. You do not rebuild the list, you query it. An audience is a filtered slice of that database, and because the profiles update continuously, the slice is always current.

Five partial source records merge into one live profile. Click a source to see what it adds.

JL
Jordan LeeOne profile · deduped · updated in real time
Salesforce: Title, owner, stage
Snowflake: Product usage
HubSpot: Marketing activity
200+ providers: Verified email, firmographics
Signals: Recent job change

One persistent profile per contact and account, updated continuously

Auto-playing · click to hold

Audiences collapses the scattered, partial copies of a record across your CRM, warehouse, signals, and providers into one persistent profile that stays current.

How Clay Audiences works: Segment, Enrich, Act

Every play in Audiences runs the same three steps: Segment, Enrich, Act. Before the three steps even start, one thing runs in the background: your data syncs. The sources you connect keep importing records into the universe behind everything in Clay. Segment, Enrich, Act is what you run on top of that universe when you are ready to move. Segment slices your database to the records that matter. Enrich fills the gaps with bulk enrichment and AI research written back to each profile. Act routes the records: to a rep, a sequencer, a Slack alert, or back to your CRM.

Auto-plays the Sync → Segment → Enrich → Act cycle on an always-on sync. Click a node to hold it and read what happens there.

Sync (always on)

Records import from your connected sources in the background, continuously, so every play below works from current data.

Auto-playing · click to hold

Audiences runs a continuous Segment, Enrich, Act cycle on top of an always-on data sync, so every play works from current data.

Step 1: Connect your sources

Nothing works until your data lives in Audiences. Setup is a set of decisions you make once that shape everything downstream, in two directions: pulling data in and writing it back out.

In Audiences, click Add Data and Clay takes you to Settings, where Sources and Destinations sit side by side. Connecting Salesforce gives you four object toggles, all off on a fresh account: Accounts, Contacts, Leads, and Opportunities. Turn on what you want to work with. You can also import from Snowflake, HubSpot, and BigQuery, or pull net-new records from Clay's own company and people database. The same screen sets your export sync, so cleaned data flows back to the CRM without a manual pull. To watch where your data stands, open the Data Hub, which shows the fill rate for every field across the whole audience.

Step 2: Build an audience, a segment that never goes stale

An audience is a live query over your whole database, so it never goes stale. Everything you import lands in All People or All Companies. From there you build an audience by filtering on anything in the profile: firmographics, CRM fields, enrichment results, and signal flags. Because the underlying profiles update continuously, records flow into and out of the audience on their own as they start or stop matching. A single contact exists once in All People but can sit in multiple audiences at once. They are not copied, just filtered differently.

Auto-narrows 12M live profiles to a target audience as filters stack, then hands you the controls. Toggle a chip to recompute.

All People12,000,000
live

Records join and leave this audience on their own as profiles change.

Auto-playing · click to hold

An audience is a dynamic segment over your full database, not a saved list: filters narrow millions of live profiles to a target set that updates itself. Illustrative counts; Audiences holds far more than a 50,000-row table.

Clay Audiences vs. Clay tables: when to use which

A table is the laboratory; Audiences is production. The two look similar because the enrich action inside an audience shows the same table view you know from a workbook. The difference is scale and role. A Clay table caps at 50,000 rows and is where you build and test a workflow. Audiences is your source of truth, where you run that workflow across millions of records and keep the CRM current through an export sync.

When to use a Clay table versus Clay Audiences

Clay table / workbookClay Audiences
ScaleUp to 50,000 rowsMillions of records
RoleLaboratory: build and test a workflowProduction: the source of truth
HoldsRows for one workflowPersistent profiles in All People / All Companies
Best forDesigning an enrichment, testing on a sampleRunning enrichment, signals, and segments across the whole base

Build and prove a workflow in a table, then point the finished version at an audience to run it at scale.

What teams use Audiences for

The same database powers four motions. Because every play runs Segment, Enrich, Act, one Audiences setup covers what used to be four tools and a sprawl of tables.

CRM enrichment runs a loop that fills the gaps faster than reps create them, writing clean data back so your system of record stays current. TAM sourcing builds the market that should be in your CRM but is not, pulling net-new accounts from Clay's database, filtering to your ICP, and pushing verified records in. Signal-based plays watch every contact for job changes, buying-role moves, and ICP fit, and route them the moment something changes. AI outbound combines those signals with automated personalization and hands enriched, triggered contacts to your sequencer.

Auto-plays signals firing across your database over a week and routing automatically. Click a lit contact to see its signal and route.

Every contact in All PeopleSignals routed: 1
Changed jobsrouted to a rep

Without Audiences: found out weeks late

With Audiences: routed the same day

Auto-playing · click to hold

Audiences watches every contact continuously, so a job change or buying-role move routes to a rep the day it happens instead of weeks late.

Managing scale without burning credits

Running plays across millions of records raises the obvious worry: am I about to burn tens of thousands of enrichments by accident? Audiences builds in safeguards so you do not.

Start at the Data Hub to see the fill rate for every field, so you enrich what is actually missing rather than re-running good data. Then reuse enrichments across audiences. A bulk enrichment behaves like a function: build it once and point it at more than one audience instead of paying to rebuild it. Field mapping is also optional on an action, so you can fire a downstream step like a sequencer handoff without writing back data you do not need.

Where to start

Pick one of the four motions and run it on a real slice of your database, then let it run. The most common starting point is the CRM enrichment loop. Connect Salesforce, import your contacts, enrich the gaps, and turn on the export sync so cleaned data writes back on its own. Once one loop keeps a slice of your CRM current, layer on a signal play or a net-new TAM build using the same database.

Clay has become the orchestration layer for everything GTM. Salesforce for record-keeping, Snowflake for product data, and Clay for turning it all into automated action.

When the data lives in one place and updates itself, the team stops re-exporting and starts acting.

3x

data enrichment coverage Anthropic reached after consolidating its GTM stack on Clay

Read the full story

Common mistakes teams make with Audiences

A few misconceptions trip teams up early.

  • Treating an audience like a saved list: An audience is a live query, not a frozen export, so build it on filters and let records flow in and out as profiles change.
  • Running enrichment in scattered tables: A dozen separate tables recreates the decay problem. Centralize records as profiles first, then enrich at the database level and reuse the enrichment across audiences.
  • Skipping the export sync: Enriching in Audiences without syncing back leaves the CRM stale. Turn on the export so cleaned data and segment membership reach the tools reps use.
  • Importing everything before you have a use case: Pulling your entire CRM up front spends credits on data nobody acts on. Start with the records one motion needs, then expand.

Outcomes teams report after centralizing their GTM data with Clay

What teams report after centralizing their GTM data with Clay

CompanyOutcomeStory
IntercomMultiple data sources consolidated; +140% outbound-sourced pipelineRead
OpenAI2x enrichment coverage across the sales teamRead
Vanta80%+ enrichment coverage held across CRM contactsRead

Make your GTM data one living database

Centralize your CRM, warehouse, signals, and 200+ providers into live profiles, then segment and act at scale.

Frequently asked questions

What is Clay Audiences?

Clay Audiences is Clay's unified data layer. It connects your CRM, data warehouse, signals, and 200+ enrichment providers into one continuously updated profile per contact and account. You hold your whole database as two global lists, All People and All Companies, build dynamic segments on top, enrich and monitor at scale, and sync clean data back to your CRM. It removes the old 50,000-row table cap, so you can work across millions of records.

How is Clay Audiences different from a Clay table?

A Clay table caps at 50,000 rows and is the laboratory where you build and test a workflow. Audiences is production: your source of truth, holding millions of records as persistent profiles. Use a table to design and prove an enrichment on a sample, then point the finished workflow at an audience to run it across your entire base.

How does Clay Audiences work?

Every play follows Segment, Enrich, Act, running on top of an always-on data sync. Sync imports records from Salesforce, Snowflake, HubSpot, BigQuery, or Clay's own database in the background. Segment filters the universe to the records that matter. Enrich fills the gaps with bulk enrichment and AI research written to each profile. Act routes records to a rep, a sequencer, Slack, or back to the CRM.

What can you use Clay Audiences for?

Four common motions. CRM enrichment fills gaps faster than reps create them. TAM sourcing builds and maintains the market that should be in your CRM. Signal-based plays route a contact the moment something changes. AI outbound pairs signals with personalization and hands contacts to a sequencer.

Does Audiences cost a lot of credits to run at scale?

It has safeguards. The Data Hub shows the fill rate for every field, so you enrich only what is missing. A bulk enrichment behaves like a function you build once and reuse across multiple audiences. Field mapping is optional on an action, so you can fire a downstream step without storing data you do not need.