Key points
- CTM's Lisa Salvatore and Brian Barranger outlined a webinar framework for mining call, chat, text and form conversations for content and ad signals.
- Step one is a shared lead-quality definition: sales and marketing review a sample of calls together, then turn agreed traits into scoring criteria.
- Barranger said AI-agent calls and human calls can both be tracked, but should be tagged differently because intent may differ.
- The presenters advise validating scoring and testing conversion feedback before making major budget changes.
Why it matters for marketers: Call and chat data can show which campaigns bring qualified buyers and which customer questions your pages miss, giving content and ad teams better inputs than raw call volume.
In a webinar covered by Search Engine Journal, CTM's Lisa Salvatore and Brian Barranger argued that the conversations a business already has with customers are an underused source of marketing insight. Their starting point was not a new dashboard but an agreement between marketing and sales on what counts as a good lead.
Salvatore called the gap between lead volume and lead quality the MQL treadmill. She suggested both teams review a representative set of calls, note the traits of a qualified one, and turn those into scoring criteria that can be applied manually or with automation. From there, repeated questions and objections can shape FAQ content, comparison pages, headings and ad copy. She cautioned that matching customer language supports discoverability but does not guarantee an AI recommendation, since credibility and brand recognition still matter.
Barranger focused on connecting the marketing source, whether the caller was qualified, and what happened after the call. Together, those can separate a targeting problem from a call-handling problem. He said scored outcomes can be sent back to ad platforms through supported conversion workflows, but urged teams to validate criteria and test before changing budgets. On AI tools, he said, "the customer conversations are what make those ideas actually relevant."
The presenters also addressed calls placed by AI agents. Barranger said they can be tracked and analyzed, but an agent may only be checking hours or pricing, so those interactions should be tagged separately instead of counted as sales leads.
For teams that want to start small, the advice is to pick one use, such as improving a page around a recurring question, and compare results over time. Frequency of a request can signal demand, but a small sample is not a market forecast.
Sources
This brief is our own summary. Read the full reporting at the source.
