How Tacity works
Tacity builds a twin of each person you’re about to meet from evidence: your CRM, meeting notes, email and public news. Every point in the twin cites its source and is marked Sure, Likely or Guess. You rehearse the meeting with the twin, and the real meeting then checks its predictions.
Updated 7 November 2026
Capture: what Tacity reads
Only the sources your workspace connects or uploads. Inside your company: CRM records (Salesforce, HubSpot), meeting notes and transcripts (Fireflies, Fathom, Granola, Google Meet notes, or any transcript you upload), and email and calendar through Google Workspace. Outside it: the company’s own press releases and filings, and monitored news. Gong, Zoom and Microsoft 365 are coming soon.
Every signal keeps a link to where it came from. A person’s own words count most, then what they said in public, then summaries and your team’s notes, then what others wrote about them.
Infer: signals become readings
Signals combine into readings, such as “needs a number he can defend to the board”. Each reading is marked Sure, Likely or Guess, and Tacity’s readings are always labelled apart from what the person actually said. A reading with no evidence behind it is dropped.
Build: the twin
The readings make up one profile of the person: what moves them, the lines they won’t cross, the proof they need, how they decide and who they defer to, and the one question to ask next. The twin always sits inside its company’s situation: budget, pressures, recent news and the politics around the decision.
Deepen: from a first read to their own words
With nothing in their own words yet, Tacity starts with a First read: an educated guess from the person’s role, company and people like them, with every point marked a Guess. As meetings arrive the twin becomes Known from meetings, and with their own words in calls, email and talks, Known from their words.
Three measures sit side by side and are never blended into one accuracy score: Coverage (how much of the twin has evidence, with the gaps named), Evidence (Strong, Mixed or Weak, with the reason) and Track record (how many of its predictions real meetings have checked).
Respond: practice in context
You pick the meeting and what you want from it, then talk to the twin by voice or present your deck and see reactions slide by slide. The twin draws on the person, your relationship so far, their company, what you sell and the meeting you’re practising. In a group call each twin knows who else is in the room. A coach shows why the twin answered the way it did and what to try next.
Learn: the real meeting checks the twin
After the real meeting, Tacity compares what it predicted (objections, questions, who pushed on what) with what actually came up, adds the person’s new words as evidence, and suggests a drill for next time. Over time the twin’s track record shows how often it has been right.
Questions
Is a twin the real person?
No. A twin is a simulation built from evidence about a person. It is not the person, its readings can be wrong, and every simulation is labelled as one.
What if there’s no data about someone yet?
Tacity builds a first read from their role, company and people in similar roles. It is marked an educated guess, and it sharpens as meetings and their own words arrive.
How do I know where a point came from?
Every point links to its evidence: the source, the date and, where it is their own words, the exact quote.
Does practice change the twin?
No. Practice and tests never change the facts. Only real evidence, such as a new meeting or email, updates a twin.
Practise your next meeting with the people in it.
A 30-minute walkthrough with a live practice call on one of your own accounts.
Priced in US dollars by team size. We’ll quote on the call.