ICP Inference — Find Your Ideal Customer Without ICP Research
ICP inference turns a plain-language product description into a structured ideal customer profile:
target industries, company size bands, decision-maker roles and technographic signals. It replaces
the spreadsheet-and-guesswork step that usually precedes any outbound campaign.
What is an ideal customer profile, and why infer it?
An ideal customer profile (ICP) describes the companies and people most likely to buy. Teams
usually build one by hand, which is slow and easily biased toward the customers they already have.
Mailor infers the profile from what you sell, then hands back a definition you can inspect, edit and
reuse as the search input for lead discovery.
How does inference work?
Inference runs as a single API call against your product description and returns a structured
profile. Each dimension comes with its reasoning so you can tell why a field was chosen.
- Describe your product — one or two sentences in plain English. No persona
worksheet, no industry taxonomy to learn.
- Mailor infers your ICP — industry, roles, geography and keywords, each with
visible reasoning.
- Confirm or adjust — edit any dimension; corrections feed back so the next
inference returns a sharper profile.
What does the inferred profile contain?
- Target industry — for example e-commerce / retail, or B2B SaaS.
- Company size — employee and revenue bands where your product fits.
- Buyer roles — for example Head of CX or VP Operations, rather than a job-title
list to guess from.
- Technographic signals — for example Shopify Plus, DTC, or support volume,
which narrow a broad industry into a real segment.
Can I edit the inferred profile?
Yes, every dimension is editable. Editing matters because inference is a starting point, not a
verdict: the correction loop is what makes the second search tighter than the first.
Related
- AI email writing — turns the leads into personalized
emails.
- Auto outreach — schedules follow-up until they reply.
- API endpoint:
POST /api/icp/infer