Short answer: An ideal customer profile (ICP) for B2B lead generation describes the type of company most likely to buy, succeed and stay profitable: its industry, size, location, operating context, technology, problem, buying trigger and disqualifiers. Build it from your own won and lost deals, enrich it with Indian public data such as MCA21 company master data, GST taxpayer records and IndiaMART activity, and re-test it every quarter against win rate and deal value. A buyer persona describes a person inside that company; the ICP describes the account.
What is an ideal customer profile in B2B, and how is it different from a persona?
An ideal customer profile is a description of the accounts worth your scarce sales and marketing effort. It answers "which companies?" A buyer persona answers "which people, and what do they need to believe?" You need both, in that order: a perfect persona inside the wrong company still produces a lost deal.
| Ideal customer profile (ICP) | Buyer persona | |
|---|---|---|
| Unit | Company / account | Individual role in the buying group |
| Typical fields | Industry, size, turnover band, location, ownership, tech stack, problem, trigger, disqualifiers | Job title, goals, KPIs, objections, information sources, proof they trust |
| Main data sources | CRM won/lost deals, MCA and GST records, IndiaMART, LinkedIn company pages, job postings | Customer interviews, sales call notes, support tickets |
| Used for | Target account lists, ad audience and exclusions, fit scoring, territory planning | Messaging, content, discovery questions, sales enablement |
| Example | Private limited auto-component makers in Maharashtra and Tamil Nadu, ₹50–500 crore turnover, multi-plant, on Tally or SAP Business One, adding a plant | Plant head, 15+ years' experience, measured on OEE and on-time delivery, distrusts software that needs IT staff |
A segment sits between the two: a group of companies that buy in a similar way. Most B2B firms have one primary ICP and one or two secondary segments. If you have six "ICPs", you have a market description, not a targeting decision.
Why should you build the ICP from win/loss data first?
Founder intuition usually describes the customers you want; CRM history describes the customers who actually buy, pay and stay. Start there.
- Export every closed opportunity from the last 12–24 months, won and lost, with deal value, created and closed dates, lead source and lost reason.
- Add account attributes: industry, employee band, turnover band, state, company type (private limited, LLP, partnership, proprietorship), years in business, main software used.
- Add post-sale outcomes for won deals: gross margin, time to go live, renewals or repeat orders, support load, whether they would act as a reference.
- Pick your best customers on value and effort, not revenue alone. A large account that needed heavy customisation and pays at 120 days may be a worse fit than a mid-size one that went live in three weeks.
- Interview five or six best customers and three or four lost prospects. Ask what was happening in their business when they started looking, who was involved and what nearly stopped them.
If you have fewer than about 20 closed deals, treat the ICP as a hypothesis and test it with a narrow campaign rather than a full rebuild of your targeting.
Where can you find firmographic data on Indian companies?
Paid databases are useful, but India has free public records that are more reliable for verification. Each source below shows something different.
| Source | What it shows | ICP use | Limits |
|---|---|---|---|
| MCA21 View Company/LLP Master Data | Master details of any company or LLP registered with the Registrar of Companies, searchable by name or CIN: status, date of incorporation, registered office, authorised and paid-up capital, directors | Confirm a company is active, its age, legal form and scale of capital; filter out struck-off firms | Covers companies and LLPs only, not proprietorships or partnerships; no headcount |
| Corporate Identification Number (CIN) | 21 characters encoding listed/unlisted status, a 5-digit industry code, state, year of incorporation and company class (e.g. PTC for private limited) (structure explained) | Quickly sort a list by industry code, state and age without opening each record | Industry code reflects the activity declared at incorporation, which may be out of date |
| GST portal: Search Taxpayer | Without login: legal and trade name, constitution of business, registration date, GSTIN status, taxpayer type, principal place of business, nature of business activities, goods and services dealt in, return filing details. After login: annual aggregate turnover, compliance rating and e-way bill history | Verify any business (including proprietorships), see what it trades, whether it files returns regularly and, after login, its turnover band | One search at a time; GST registration is state-wise, so a multi-state company has several GSTINs |
| MSME classification (Udyam) | Budget 2025-26 (1 February 2025) raised the limits: micro up to ₹2.5 crore investment and ₹10 crore turnover, small up to ₹25 crore and ₹100 crore, medium up to ₹125 crore and ₹500 crore | Define size bands in rupees rather than headcount, which suits Indian manufacturers and traders | Self-declared registration; not every eligible firm registers |
| IndiaMART supplier listings and BuyLeads | Supplier catalogues by product and city; BuyLeads show the buyer's products of interest, requirements posted to date, member-since date and verified mobile/email marks | Find SMB manufacturers and distributors by category and city; spot which categories post frequent requirements | Paid seller packages are needed for BuyLead volume; listings are self-described |
| LinkedIn company pages and job postings (LinkedIn, Naukri) | Employee range, locations, open roles | Headcount bands and hiring triggers | Smaller firms are under-represented on LinkedIn |
Two practical signals come from tax rules. Since 1 August 2023, businesses with aggregate turnover above ₹5 crore in any financial year from 2017-18 must issue e-invoices for B2B supplies (CBIC Notification 10/2023), so a supplier that sends you e-invoices has very likely crossed that line. And a company holding GST registrations in several states runs operations in each of them, which matters if you sell multi-location software or services.
Company-level facts are not personal data, but directors' names and individual contact details are. Collect and use those under the DPDP Act, 2023, which our cold email guide covers in detail.
Which technographic and trigger data should an ICP include?
Firmographics tell you who a company is. Technographics and triggers tell you whether your product fits their setup and why they might act now.
- Website technology: tools such as BuiltWith show which analytics, CRM, ecommerce or chat tools a website runs. Useful if you integrate with, or replace, a specific platform.
- Back-office systems: job postings often name them ("Tally Prime experience required", "SAP B1 administrator"). In India this is often the best technographic signal for ERP, accounting and HR software vendors, because those systems leave no trace on the website.
- Hiring triggers: a manufacturer hiring its first digital marketing manager, or an NBFC hiring collections staff in three new cities, has a new problem and a budget owner.
- Growth and change triggers: new GST registration in another state, a new plant or warehouse, funding announcements, a leadership change, a compliance deadline or an export licence.
- Negative signals: GSTIN cancelled or returns not filed, company status "strike off" or "under liquidation" on MCA, recent layoffs.
ICP worksheet: what should you fill in?
Fill one worksheet per segment. The right-hand column is a worked example for a hypothetical Noida company selling energy-monitoring software to factories.
| Field | Question to answer | Example answer |
|---|---|---|
| Industry | Which sectors produced our best customers? | Plastics, auto components, food processing (high power cost as a share of production cost) |
| Size | Turnover and headcount band of best customers | Small and medium enterprises under the 2025 limits: ₹10–500 crore turnover; 100–1,000 employees |
| Legal form and age | Company type and years since incorporation | Private limited, 5+ years (CIN year and MCA status checked) |
| Geography | Where can we sell and support? | Industrial clusters in UP, Haryana, Gujarat, Maharashtra |
| Operating context | What setup must exist for us to succeed? | At least one plant with 500+ kW connected load; maintenance team in place |
| Technology | What must or must not be installed? | Uses Tally or SAP B1; no existing energy management system |
| Problem | What costly problem do we solve? | Power bills rising, no visibility of which machines waste energy |
| Trigger | What event makes it urgent? | Tariff revision, new plant, ESG reporting request from an export customer |
| Buying group | Who uses, champions, approves, blocks? | Maintenance head (user), plant head (champion), MD (approver), accounts (procurement) |
| Disqualifiers | What makes a deal unwinnable or unprofitable? | Single-shift units; rented premises with no control over wiring; GSTIN inactive |
| Evidence | Which data supports each line, and how confident are we? | "Industry" from 18 of 26 won deals; "trigger" from 5 customer interviews |
The Evidence row is what makes an ICP trustworthy. Any line without evidence is an assumption to test, not a rule.
How do you turn an ICP into account tiers and campaigns?
| Tier | Definition | Treatment |
|---|---|---|
| Tier 1 | Meets every ICP field and shows a current trigger | Named-account outreach, personalised research, LinkedIn matched audiences, sales owns |
| Tier 2 | Meets core fields (industry, size, geography); no trigger seen | Sequenced outreach and content; monitor for triggers |
| Tier 3 | Partial fit | Inbound and search capture only; no outbound effort |
| Excluded | Hits any disqualifier | Exclude from ad audiences and outbound lists |
The same fields then feed your lead scoring model as fit points, your Google Ads keyword and negative lists, and your LinkedIn targeting. For SMB manufacturing ICPs, the IndiaMART categories in the worksheet tell you where to list; our channel comparison covers when that beats outbound.
How do you validate an ICP with win/loss data?
An ICP is correct only if accounts that match it win more often, at better value, than accounts that do not. Compare segments on four numbers. Illustrative example for the energy-software company above:
| Segment | Opportunities | Win rate | Average deal | Median cycle | Value per opportunity |
|---|---|---|---|---|---|
| ICP match (all fields) | 30 | 30% (9 won) | ₹6 lakh | 75 days | ₹1.8 lakh |
| Core fit, no trigger | 40 | 15% (6 won) | ₹5 lakh | 120 days | ₹75,000 |
| Outside ICP | 25 | 8% (2 won) | ₹3 lakh | 140 days | ₹24,000 |
Value per opportunity (win rate × average deal) is the number to watch: in this example, an ICP-match opportunity is worth more than seven times an outside-ICP one, which justifies putting outbound and sales time only on Tier 1 and Tier 2. If the gap were small, the ICP would not be predicting anything and would need rework.
Add the lost side. Tag every lost deal with one reason (price, chose competitor, no decision, poor fit, timing) and review them by segment. If "poor fit" losses cluster in one industry or size band, add it to the disqualifiers. If "no decision" dominates inside the ICP, the problem is your trigger or business case, not the profile. Run this review every quarter, and whenever your offer, pricing or delivery model changes. Our B2B lead generation service starts every engagement with this analysis.
What are the most common ICP mistakes?
| Mistake | Fix |
|---|---|
| "SMEs in India" as the ICP | Add industry, turnover band, geography, trigger and disqualifiers |
| Built from the biggest logos, not the most profitable accounts | Rank customers on margin, time to value and retention |
| Headcount used as the only size measure | Use turnover bands from GST or MSME classification for Indian manufacturers and traders |
| Persona and ICP merged into one document | Keep the account definition separate from role-level messaging |
| Purchased lists used without verification | Check status on MCA and GST before outreach; drop inactive entities |
| Never updated | Quarterly win/loss review with the segment table above |
Sources
- Ministry of Corporate Affairs, About View Company/LLP Master Data: public view of master details by company name or CIN.
- Razorpay Rize, Corporate Identification Number explained: 21-character CIN structure.
- GST portal user guide, Search Taxpayer: fields shown before and after login.
- PIB, revised MSME investment and turnover limits (1 February 2025).
- Taxmann, e-invoice threshold reduced to ₹5 crore (Notification 10/2023, from 1 August 2023).
- IndiaMART Help, Information in a BuyLead: buyer details visible on BuyLeads.
- BuiltWith: website technology profiling.
- Digital Personal Data Protection Act, 2023: rules for processing personal contact data.