Connect Gong in Settings → Integrations → Gong → Connect. NEXT AI then automatically imports new Gong calls into your chosen Teamspace as Recordings (with transcript and speakers). You can filter which calls to import and even sample a percentage to control volume.

Each source adds a new dimension of customer truth. Gong contributes the sales perspective (discovery, objections, negotiation) that complements survey/review signals, and support conversations already flowing into NEXT AI. Together these streams build a richer, cross-functional picture of customer needs—so product and GTM teams automate the right workflows with the evidence attached.

Once your Gong account is connected, NEXT can automatically import new Gong calls into one of your Teamspaces.

When setting up Auto-import, you can define a filter so only calls matching specific criteria are imported. This is useful for bringing in only certain segments, industries, regions, or opportunities.

The filter is a single text field that lets you describe which calls to import using Gong and Salesforce call data. You can filter on any field from Gong call metadata or linked Salesforce data (e.g., Account, Opportunity).

Examples of field names you might use:

For the full search syntax, see the Lucene Query Syntax documentation.

When connecting your Gong account, you may find that you have a very large number of calls. Importing all of them can be expensive and often unnecessary, especially if you just want to analyze a representative set. That’s why we offer sampling. Sampling allows you to choose what percentage of your Gong calls to bring into our system - for example, 5%, 20%, or 50%. This way, you can keep costs under control while still working with enough calls to get meaningful insights.

The way sampling works is by selecting calls at random rather than just taking the first calls in the list. This ensures that your imported calls represent the whole range of conversations across days, teams, and topics, rather than being clustered in one period. In the background, each call is given a unique identifier, and we use that to decide whether it falls into the percentage you selected. The end result: if you choose 10%, we will consistently import roughly one out of every ten calls, spread evenly across your Gong data.

One important thing to know: if you change the sampling percentage later and re-run the import from the same time period, you may see different calls appear. For example, lowering from 10% to 7% doesn’t just mean you get fewer calls - it can also mean you get a different set of calls. In most cases, this results in even more calls overall than before. This is not an error but simply how random sampling works. Keep this in mind when adjusting your percentage so you can plan your imports accordingly.

A Recording per Gong call in the Teamspace where you connected Gong, with transcript, speakers, and timing.

NEXT AI relies on the Gong transcript.

Yes—use the filter and optional sampling to control scope and volume.

No, NEXT AI only supports polling.