How to Size Your B2B Market Bottom-Up and Score Target Accounts Without Inflated Analyst Numbers

When I review a pitch deck that claims a $40 billion market by taking 1% of a global analyst line item, I ask one question: how many companies in your target geography can sign your average contract value this year? Count that number for your own market before you open the deck. It changes the hiring plan, the territory design and the board conversation.

In 24+ years of B2B go-to-market, the plans I trust start from a count of real accounts. I was VP Global Performance Marketing at Locus from February to October 2022, and we built $4.2M of pipeline in H1 CY22. Locus was acquired by IKEA (Ingka Group) in October 2025.

Bottom-up sizing also gives you something top-down never does: a customer count. A revenue target turns into a number of deals, a number of deals into pipeline, and pipeline into the accounts your team must work this quarter.

The TAM, SAM and SOM tool is a calculation framework, not a data source. It takes four inputs and shows the arithmetic for each line, as in the table alongside.

From the SOM it works back to customers, deals per month and the pipeline needed at a stated win rate, and it projects two more years on the assumption that share doubles each year. That growth path is marked as an example, because it is one.

The rule that matters most sits in its sector notes for fintech spend and expense: count only companies where the buying roles exist (CFO, finance controller, head of accounts payable and the others it lists) and the problem is measured, meaning someone tracks days to close the books and reconciliation effort. A company that fails either test is in your TAM on paper and nowhere near your pipeline.

Where the three inputs come from

• Company count: run an account search in LinkedIn Sales Navigator with your industry, headcount and geography filters and note the count, then cross-check it with a second source, as the tool's own checklist asks.
• ICP match rate: take a random sample of 100 accounts from that count and check each by hand for the champion role and the measured problem. The share that passes is your rate.
• Year 1 share: divide the deals your team closed in this segment last year by the number of companies in your SAM, because the tool applies the share to the SAM, not the TAM. Raise it only for sales capacity you have already hired.

Step 1. Describe your best customers

Input to the tool that finds the patterns your best customers share: "Mid-size manufacturers with 300 to 3,000 employees and large field sales teams; the finance controller buys and the pain is a slow month-end close. Also IT services firms with heavy travel spend, and distributors with many branches."

What came back: it read the sector as fintech, spend and expense, kept the company size exactly as typed, and said plainly that the stage was not stated. It listed who usually decides (the CFO signs, the finance controller champions, internal audit and compliance review controls, IT checks ERP integration and security, HR joins when employees are affected), the metrics to ask about (days to close the books, reconciliation effort, policy breach rate, approval cycle time) and the proof that lands with this buyer: close time or reconciliation effort before and after for one entity, signed off by the finance controller.

How to read it: a text description gives you the sector frame, not patterns. The tool shows the row format it wants (industry, size, ACV, sales cycle days, tech stack, buying trigger, champion title). Export your closed-won deals in that shape and run it again before you trust any pattern.

Step 2. Size the market bottom-up

Input: 9,000 Indian mid-size companies with an ERP, ACV $24,000, 25% ICP match, 1% Year 1 share.

From the SOM it worked out about 23 customers, about 2 deals a month, and $1.6M of pipeline at a 1 in 3 win rate, with the win rate labelled as an example.

How to read it: the number to defend is 23 customers, not $216M. The tool's own check is the right one: compare 23 with the deals your team closed last year. Remember that the share applies to the SAM, which here is 2,250 companies (25% of 9,000). If you closed 8 of them last year, that is 0.36%, and a 1% target needs a reason it can almost triple.

Step 3. Turn the ICP into a scoring model

Input: four criteria with an importance level each, plus the pattern "customers with an ERP and a finance controller renew most".

The maximum is 65, and the tiers fall at 80%, 60% and 40% of it: A from 52, B from 39, C from 26, and D below that for marketing nurture only. The tool says what these points are: weights preset by importance level, not calculated from your data. It also noticed that the renewal pattern I gave matches two of my criteria, and suggested making both critical if closed deals confirm it.

How to read it: list each criterion's values from best fit to worst, because the first value scores the full weight, the last scores zero and the values between score even steps. That is why 100 to 299 employees scores 13, half of 25 rounded. The tool also suggests turning its sector questions into criteria you can score before the first call, such as how many days the month-end close takes or what the last audit said about spend controls, when the answer can be found in an annual report, job posts or your CRM. Then check the model against your recent closed-won and closed-lost deals before you let it route leads. The same output lists the objections that should hold a high score back in this sector: "our ERP already does this", the security and compliance review, the fear that migration will disrupt the close, and regulatory caution.

Step 4. Map who decides

Input: champion Finance Controller; known stakeholders CFO, Head of IT and Head of Accounts Payable; $24,000 ACV; 300 to 3,000 employees.

How to read it: the reviewer is the role teams forget. I did not name internal audit, and the tool added it as the usual blocker in this sector. In finance software that review runs inside the sales cycle, so bring it in before the proposal, not after. Use its discovery question for the technical evaluator early: which ERP or ledger must every transaction post into, and who owns that integration? A company on Tally and a company on SAP will give you very different answers, and so will their integration effort. Ask too whether the group has several legal entities: that changes both the ACV and which entity the tool's suggested one-entity pilot should run on.

Step 5. Find lookalikes and rank the accounts

Lookalike input: manufacturing, IT services and distribution; 201 to 5,000 employees; India; SAP, Tally or Oracle NetSuite; champions Finance Controller and Head of Accounts Payable; triggers new CFO hire, audit finding and ERP migration. It returned the filters for a Sales Navigator company search and lead search, a Boolean title string, and one signal per trigger.

How to read it: check the output against your intent before you run it. The lead search covers only the titles you gave, so add Chief Financial Officer if a new CFO is one of your triggers; the tool's signal for that trigger watches for a new finance controller or related leader. The Boolean string carries titles and industries only, so set geography, headcount and technology as filters. Headcount comes in bands in Sales Navigator, which is why the input was 201 to 5,000 for an ICP of 300 to 3,000: select the three bands, then score down the accounts outside your real range. Desktop ledgers such as Tally may not show up in technology filters at all, so confirm the ERP on the first call.

Ranking needs a fit score from 0 to 100. Take it from the scoring model: points scored divided by 65, times 100. In this example, Example Distribution Co has 300 to 3,000 employees, an ERP and a finance controller but no travel-heavy workforce: 55 of 65, a fit of 85. Example Manufacturing Co has the size, an ERP and a travel-heavy workforce but no controller: 50 of 65, a fit of 77. Example IT Services Co has the size and the travel spend but neither an ERP nor a controller: 35 of 65, a fit of 54.

For intent I use my own working points, not the tool's: 50 for one of the three triggers (a new CFO, an audit finding or an ERP migration) in the last 90 days, 30 for a reply or meeting from the finance controller or head of accounts payable, and 10 for each other sign of engagement, up to 100. Example IT Services Co announced an ERP migration, its controller took a call and someone attended a webinar: 90. Example Manufacturing Co had an audit finding and one webinar visit: 60. Example Distribution Co had no trigger, but its controller replied to an email: 30.

With the default weights (fit 40%, intent 30%, relationship 15%, timing 15%) the three example accounts came out as in the table alongside.

How to read it: the account with the weakest fit ranked first, because intent added 27 of its 74 points. The tool says so, and advises confirming the problem and budget before writing a proposal. If your team has room for only a few deals, move weight from intent to fit rather than only raising fit, and keep the four weights adding up to 100: the tool does not rescale them, and its tier bands are set from their sum. With fit 60, intent 20, relationship 10 and timing 10, Example Manufacturing Co moved to first with 69, ahead of Example IT Services Co with 67, and Example Distribution Co rose from 52 to 63, so all three sat in Tier B. When the tiers bunch up like that, use the rank order to decide where account executives spend this week.

Read the labels in the output with the numbers beside them. In that second run the tool still called Example IT Services Co high intent while showing that most of its points now came from fit, and the buying group map can list a role as missing that you already named. The scores and the role table are the parts to trust.

Step 6. Take the map into discovery with Revenue Enablement

Sizing and scoring tell you which accounts to work. Discovery decides whether the account is real. I take the champion from the map into the discovery question bank in Revenue Enablement: framework MEDDPICC, stage discovery, industry "Mid-size manufacturing", contact Finance Controller, solution "Spendrill, spend management software". It writes for one contact role per run, so run it again for the CFO and the Head of IT.

What came back: an opening that asks where the problem starts, who it touches and what it costs in days to close the books, reconciliation effort or policy breach rate; questions for the finance controller about which numbers are late or reworked each period and what the last audit asked for; and a line on what this buyer needs to see before saying yes: how it posts to the ledger, what controls it adds and which manual steps the team stops doing. The CFO and the head of IT appear in its questions because of its sector notes, not because I passed the buying group in.

How to read it: cut the lines that put Spendrill where the problem should be, such as "What is not working today in Spendrill", and the shop-floor questions, which suit a plant manager rather than a controller. Then add the pain from Step 1 ("slow month-end close") as a known pain point and run it again. Ask the sector questions on the call and write the answers into the account's scoring fields, so the next ranking runs on facts instead of guesses. Treat the economic buyer questions as the gate: if the controller cannot bring the CFO into the next conversation, the account drops a tier whatever its score.

1. Get your company count from a filtered account search, and check it against a second source.
2. Hand-check a sample of 100 accounts for the champion role and the measured problem, and use the pass rate as your ICP match.
3. Run the sizing with your own ACV and last year's share, and compare the customer count with last year's closed deals.
4. Build the scoring model, then test it against your recent wins and losses before it routes a single lead.
5. Score your top 30 target accounts, rank them with weights that add up to 100, and put the first five in front of your account executives with the buying group map for each.
6. Run the MEDDPICC bank for the finance controller of each of the top five, and drop a tier on any account where the controller cannot bring the CFO to the second call.

1. Counting every company in a size band. Count only those where the buying roles exist and the problem is measured.
2. Keeping the example ICP match rate or Year 1 share. Both defaults are labelled; replace them with your own conversion history.
3. Pitching the TAM. Boards and buyers respond to the customer count and the pipeline it needs.
4. Treating preset weights as evidence. Validate the scoring model against closed deals before it routes leads.
5. Meeting the reviewer late. In finance software, internal audit and IT review sit inside the cycle.

• How I Pick the Lead B2B Go-To-Market Motion With EPIC, Then Size and Anchor It
• Account-Based Plays Without Burning SDR Hours: Account Plan, Competitive Traps, Lookalikes and Roundtable Follow-Up
• B2B Go-To-Market Playbook for Logistics Tech: Find the Buyer, Land One City, Expand on Proof
• All 15 guides