CRM analytics that score every account 0–100.
A logic-first look at how ICP analysis works — from a CRM export to a scored ideal customer profile you can act on.
Your CRM data already contains your ideal customer profile. You just can't see it yet.
Every company's best customers leave a pattern in their CRM data — how fast deals close, whether accounts expand, which contacts champion the deal, how long they stay. These signals exist in your data right now.
Most teams never surface these patterns because doing so requires aggregating across three separate data objects: Accounts, Contacts, and Deals. Manually correlating these is time-consuming and error-prone.
ICP Lens does that aggregation automatically, applies a transparent ICP scoring framework, and turns the output into a plain-English ideal customer profile your sales and marketing teams can act on.
ACCOUNT LAYER
Account layer — unit of scoring
Every company gets an ICP score 0–100. The account is the primary record the whole analysis builds toward.
CONTACT LAYER
Contact layer — unit of pattern
Which titles, departments, and countries appear in your best fit accounts. Multiple contacts per account.
DEAL LAYER
Deal layer — unit of evidence
LTV, sales cycle length, deal source, discount rate, expansion revenue. Multiple deals per account.
The ICP scoring framework: how every account is scored 0–100.
A transparent, four-component weighted formula for account scoring. No black box.
Lifetime value and total revenue — normalised against portfolio max. Accounts with expansion revenue receive a 1.2× boost.
Derived from churn status, last activity date, NPS, and open tickets. Proxied from deal recency if no retention field exists.
Requires the Contacts sheet. Best fit titles (VP Product, Head of RevOps) score 1.0. Low-fit titles (Procurement, IT Director) score 0.1.
Company size sweet spot: 50–200 employees scores 1.0. Industry fit is derived from your own data — top 2 best fit industries automatically score 1.0.
Score weights are editable in account settings after your first analysis run.
ICP analysis in 5 steps, from CRM export to scored accounts
Prepare your CRM data
Before uploading your CRM data, check that your file has the sheets ICP Lens needs. The Deals sheet is required and should contain at least 30 closed-won rows with company fields (name, industry, employee count) included. The Contacts sheet is optional but unlocks buyer personas and geography.
Upload & map your file
Drop a CSV or Excel file. ICP Lens auto-detects your CRM's column names using a synonym dictionary covering Salesforce, HubSpot, Pipedrive, and Zoho exports. Most fields map automatically — you review and confirm before proceeding.
Data quality check
Before running the analysis, ICP Lens scans your file for completeness and consistency. You see exactly which fields are missing and what impact that has — so the results never surprise you.
Overall
74%
With LTV
89%
Closed-won
312
Missing industry
61
With title
71%
Duplicates
0
CRM data analysis runs in your browser
The CRM analytics engine runs entirely client-side using a Web Worker. Your original file never leaves your device. Five stages complete in under 60 seconds: normalisation, LTV calculation, CRM account scoring, clustering, and narrative generation.
Explore your results
The results dashboard has four views. Account scoring shows every company's ICP score and tier. Buyer personas reveals which titles drive your best fit deals. Deal intelligence surfaces your winning deal patterns. Geography maps where your best customers sit.
Your best customers are 50–200 person B2B SaaS companies in fintech or devtools…
CRM-Based ICP Scorecard
Not sure if your CRM data is ready for ICP analysis? Score your team across data readiness, revenue signals, segmentation, alignment, and actionability — and get a clear recommendation on what to do next.
PDF · 9 pages · instant download
Sales analytics turned into a plain-English ideal customer profile.
Narrative template
Your best customers are SIZE_RANGE TOP_INDUSTRY companies GEOGRAPHY_CLAUSE, with BUYER_PERSONA, who BEHAVIOUR. They EXPANSIONNPS.
Each slot is populated with real computed values from your data — median employee count of best fit accounts, top industry by best fit density, most frequent best fit title. The template selects from multiple variants based on which pattern is most statistically notable in your specific dataset.
Example output
"Your best customers are 50–200 person B2B SaaS companies in Fintech or DevTools, with a VP of Product as primary best fit buyer, who activate within the first 14 days and integrate your product via API. They typically expand within the first year and report NPS above 8."