Segment Sentiment Analysis
Friday, October 23, 2026
What it is
Segment Sentiment Analysis examines your calls, emails, and notes to generate important insights about your segment members. It surfaces signals such as competitor mentions, pricing objections, budget signals, and product feedback including the snippet pulled from the conversation that surfaced it.
Why it matters
Some of your most useful signals sit inside conversations: a prospect naming a competitor, a deal stalling on budget, a customer frustrated with support. Segment Sentiment Analysis brings those signals together, so you can spot risk early, decide who to follow up with first, and get the right message to the right people.
The Take
It's just the fact that it exists, great, but it should not impact sentiment analysis at all.
Casey picked this one, and it is the one we want to test first. Segment Sentiment Analysis reads the calls, emails and notes on the members of a segment, any contact, company or deal segment, and surfaces competitor mentions, pricing objections, budget signals and product feedback with the snippet that produced each one. Casey's starting point is the example in the update, contacts in an open deal, which is a worthwhile segment. Zach Hussion already runs a version of this on his own calls: his meeting note taker has an automation that scores each call after it ends, one call at a time, and it does not land in HubSpot. The other ways we have looked at this were a custom agent, or custom properties and contact scores you maintain yourself.
Read the fine print before you trust a report. Only records with calls, emails or notes get analyzed, and for busy records it looks at up to fifty of each type. Casey would hope it is the most recent fifty; it does not say that. Reports do not update on their own, you regenerate to see the latest activity. This is where the human belongs when AI is doing a ton of the work you used to do: the person who says this report is not accurate data, regenerate it, is now crucial if you are going to rely on it. And the cohort read matters as much as the islands. We had a client conversation this week about inconsistent sentiment on a single company record; five companies describing the same frustration in five different words should land in one bucket, not five, and that is what segment-level analysis should give you.
Here is the caveat, from a client a year and a half ago, and it still applies. If half the emails on a company are order confirmations and shipping notices, that is transactional text with no feeling in it, and a naive sentiment pass reads the relationship as cold while the humans on that account might love you. When you turn this on, take a beat and decide what an order confirmation should count for. It is just the fact that it exists; it should not impact sentiment analysis at all.
Who should act: Sales and Marketing Hub Professional and Enterprise teams with calls, emails and notes actually logged on records; start with a segment you already care about, like contacts in open deals. Who can skip it: portals where the activity is not in HubSpot yet, because there is nothing for it to read.
Watch Us Discuss This
The full segment from the show, us working through this update start to finish.
FAQ
Which records does it analyze?
Per HubSpot, any contact, company or deal segment, and only records that have calls, emails or notes logged. For records with a lot of activity it reviews up to fifty of each type.
Does the report stay current?
No. Per HubSpot, reports do not update on their own; regenerate the analysis to include the latest activity.
Who gets it?
Per HubSpot, Sales Hub and Marketing Hub Professional and Enterprise, in private beta as of October 21, 2026.
Produced by Value-First Media
