TAM analysis: mapping your real market before you sell

Almost every company entering a new market has a market number. It came from a report, it is impressively large, and it is useless on a Monday morning. Nobody can call a number. What a sales team needs is the answer to a different question: which companies could actually buy this, what are they called, and in what order do we approach them.

Why market reports do not help sales

Top-down market sizing works by estimating a total and shrinking it through assumptions. The result is an order of magnitude that is useful in an investor conversation and useless for territory planning.

The reason is simple: a number contains no addresses. It does not tell you whether those companies are reachable, whether they are already locked into a competitor's framework agreement, whether they meet the technical preconditions for your product, or whether they sit in your size band. It also overstates the market systematically, because it counts companies you could never realistically serve.

A bottom-up analysis reverses the direction. It starts from individual, real companies, tests each against your fit criteria, and totals them at the end. The number it produces is usually far smaller than the report's and considerably more useful, because every unit has a name behind it.

A list is not a market map

A database export is not a TAM analysis. Databases describe companies by the attributes that are easy to record: industry code, headcount, revenue, country. Whether a company is actually a candidate for you almost always depends on attributes that are not recorded anywhere.

Fit criteria no database captures:

• Process and technology. Which manufacturing process, which machine class, which software environment is in use.
• Operating model. Own production or contract manufacturing, own warehouse or 3PL, in-house IT or outsourced.
• Maturity. Whether the role your product serves exists at all.
• Trigger. New facility, certification, site expansion, regulatory deadline, leadership change.
• Lock-in. Whether a framework agreement is running, and how long it has left.

These are researchable but not filterable. They live on company websites, in job postings, in registers, in permit filings, in trade press and in trade fair exhibitor lists. That research component is the entire difference between a market map and a purchased list.

Three things people call market sizing

Count sites, not corporate entities

The most common and most expensive mistake in B2B market sizing is the wrong counting unit. Databases list companies, but B2B purchases very often happen at the site.

A group with eleven plants is, in practice, eleven customers with eleven sets of equipment, eleven budgets and eleven contacts. Count it as one and you understate your market by an order of magnitude, and you also write to the wrong address, since head office does not have the problem.

The opposite error is just as real. Project vehicles, holding structures and dormant entities inflate the count without a buyer behind them. In construction and property, that noise makes up a substantial share of register data.

Deciding what gets counted is therefore not an administrative detail. It is the single most consequential decision in the whole exercise.

Segment by priority, not by size

A finished target list should be ordered by how it will be worked, not alphabetically. Three tiers work well:

• Tier 1 meets every fit criterion and shows an identifiable trigger. Worth individual research and a personal approach.
• Tier 2 meets the fit criteria with no visible trigger. Structured outreach in series.
• Tier 3 fits in principle but sits outside the preferred size band or region. Monitor, do not work.

The practical value of tiering lies less in the ordering than in the fact that it fixes the effort per contact. Without it, sales teams treat every account identically, which means the most valuable targets receive the same bulk email as everyone else.

What the finished analysis contains

• a named list of qualifying companies or sites
• per entry, why it qualifies and the source for that judgement
• the priority tiering
• the relevant roles per entry, not a switchboard number
• a documented criteria definition, so the list can be extended later in a traceable way
• an honest statement of which part of the market could not be established

That last item is rarely supplied and matters. Every research exercise has gaps, and a provider presenting a list as complete has either not looked hard enough or is not telling you something.

When the research effort pays and when it does not

The arithmetic is unglamorous: research cost per account pays off when customer value is high and the number of target accounts is manageable.

Worth it for capital equipment, industrial plant, high-contract-value software, specialist services, and anywhere a single won customer carries for years. Between a few hundred and a few thousand target companies, careful analysis is almost always the cheaper option, because it stops an expensive sales team from working the wrong market.

Not worth it at low contract values against very large audiences. If you address tens of thousands of small companies, broad outreach with fast disqualification beats per-account research.

There is a second case where the effort is premature: when nobody at your company has ever sold this product to this audience. A market map answers who you should approach. It does not answer whether your offer lands. That order cannot be reversed, and mapping first simply measures the wrong market very precisely.

Common failure modes

The market is defined by industry code instead of behaviour

Industry codes are coarse and often badly maintained. Two companies sharing a code can run entirely different processes. Conversely, your best customers frequently sit under codes you would never have searched.

Existing customers are never analysed

The cheapest source of fit criteria is your own successful customers. What do the best ten have in common that the hardest ten do not? That analysis takes an afternoon and replaces a great many assumptions.

The map is treated as a deliverable rather than a state

Sites close, framework agreements expire, roles change. Without a set maintenance rhythm the list is worthless within a year.

What is easy to count is what gets counted

When a criterion is hard to establish, it tends to quietly disappear from the definition. The result is a list that reflects exactly the attributes databases hold, which makes it a database export again.

How Ripe Leads builds them

We build market maps bottom-up from public sources: commercial and national registers, site directories, exhibitor lists, association and certification registers, job postings and company sites themselves. We count the unit that actually buys, which is usually the site.

You receive the named list, the qualifying rationale and source per entry, the priority tiers, the relevant roles, and a documented criteria definition so the list stays extendable. What could not be established is stated too.

Whether we then run the outreach is your decision. The map is deliberately built to be usable by your own sales team or another provider. Our outbound pricing is published: EUR 3,750 for the first month including setup, then EUR 2,850 per month, cancel anytime.

Fit boundary: we do not produce market studies with revenue forecasts, competitive analyses or strategy consulting. If you need a total-market valuation for an investor conversation, a market research firm is the right supplier and we are not.

More on this: how to define your ICP, B2B contact list building, territory and account planning, B2B data sources in Europe.