Skip to main content
An Ideal Customer Profile (ICP) describes the kind of company most likely to buy from you and succeed after they do. Doris builds these profiles automatically from your closed deals — so instead of a generic marketing persona, you get an evidence-backed picture of who you actually win with. ICPs live in the Library under ICPs.

​
How Doris builds your ICPs

Once you have enough closed-deal history connected (won and lost), Doris runs a discovery agent that:
  1. Analyzes your closed deals — firmographics, deal size, sales cycle, stakeholder patterns, and the qualification evidence Doris already captured from conversations
  2. Clusters them into archetypes — groups of deals that look and sound alike, rather than forcing everything into one average profile
  3. Builds a profile for each cluster — with firmographics, qualification patterns (MEDDPICC and SPICED), common objections, key stakeholders, and messaging that resonated
  4. Scores every active and closed deal against each profile — so you can see which opportunities fit which archetype and how confident Doris is
Auto-discovery runs automatically once you have at least 25 closed deals in Doris. Profiles then refresh automatically every 10 new closed deals so the picture stays current as your pipeline evolves.

​
What’s in a profile

Each ICP profile has a dedicated detail page with:
At-a-glance win rate, average deal size, average sales cycle, and confidence score for the profile. The confidence gauge shows how much evidence Doris has to back the profile — profiles built from more closed deals score higher.
Company size, industry, geography, technology fit, and other structured dimensions that characterize the archetype. Each dimension shows the typical values from matching won deals and where the gaps are when deals lose.
For each qualification dimension, Doris shows what a winning deal looked like (“typical summary”) and where lost deals fell short (“typical gaps”). Every claim is cited back to specific deals and conversations so you can click through to the evidence.
Who you tend to sell to when you win — titles, seniority, and how many stakeholders are typically engaged. Helps reps know who they need in the room.
Which messages, pain points, and value statements showed up in winning conversations versus losing ones. Grounded in what your team actually said on calls, not hypotheticals.
A side-by-side of matched deals that won versus lost, so you can see where the profile breaks down and why.
A ranked list of the deals assigned to this profile, sorted by fit score. Filter by outcome (won, lost, open) to see which open deals look most like your best customers.
Point-in-time copies of the profile so you can see how it has evolved. Doris takes a snapshot automatically on every auto-refresh.

​
Deal fit scoring

Every deal in your pipeline gets scored against every ICP profile. The fit score is a blend of firmographic match, qualification pattern match, and stakeholder pattern match — all grounded in the profile’s evidence. Reps use this on open deals to quickly see which archetype an opportunity fits and which playbook from that archetype’s win patterns applies. Managers use it at the pipeline level to flag deals that don’t look like anything you’ve won before.

​
Offers and ICPs

If you have offers configured, you can filter ICP profiles by offer — so you can see the ideal customer for your enterprise tier separately from your mid-market tier. Doris links profiles to offers via the deals and meetings they’re grounded in.
ICPs are a living picture, not a one-time exercise. As you close more deals and your conversations get better, Doris refreshes the profiles automatically — no quarterly offsite required.

​
Getting started

  1. Connect your CRM (HubSpot, Salesforce, Attio, or Affinity)
  2. Connect a transcript source (Gong, Zoom, or another supported provider)
  3. Let Doris process your historical closed deals — qualification scores, risks, and stakeholder extraction all feed the ICP agent
  4. Once you cross 25 closed deals, auto-discovery runs automatically and profiles appear in the Library under ICPs
You can also trigger discovery manually from the ICP list page if you’d like to kick it off sooner.