B2B market analysis: how to read data so you don't make a bad decision

We start with a situation we see at client companies regularly. At one manufacturing company we worked with, the management board spent three hours analyzing whether to enter a new segment or add another salesperson to the current one. On the table lay a spreadsheet with sales data from the last two years, a paid industry report, and a note from a trade fair. After three hours, no one could say which decision made more sense given the available budget. The topic was shelved until the next quarter, and in the meantime a competitor won two key clients in that segment. Market analysis is about building a model that lets you compare two decisions and say which one makes more sense given the available budget. Collecting as much data as possible on its own achieves nothing.

Where to start with market analysis that does not end up in a drawer

The first step is to write down the decision the analysis is meant to support. It sounds trivial, but in practice it eliminates half the work. If the decision is "we are considering growth," no analysis will help because there is no evaluation criterion. If it is "should we move 30% of the sales budget from the Kujawsko-Pomorskie region to serving clients in Germany in 2027," it is clear what data we need and when to stop collecting it.

The second step is to separate three layers that constantly get mixed up in conversations: market size, trends, and segments. These are not three chapters of one report, they are three different questions. Market size tells you how much money is in play at all. Trends tell you whether that pool is growing or shrinking and in which direction demand is shifting. Segmentation tells you where in that pool there is room for us, given our production capacity, sales channel, and price. Without that separation, the report grows to forty pages and no one uses it.

The third thing that is easy to forget: B2B market analysis almost never relies on data as good as in B2C. There is no consumer panel, no POS data. There are industry reports, national statistics office data, competitors' financial statements in the company register, data from suppliers, and client interviews. That is enough if the model is built on assumptions rather than apparent precision. The analysis does not have to be accurate to the third decimal place. It is enough that it shows the order of magnitude and the direction.

It is also worth establishing right away who the audience is. Research on how academic work differs from practitioners' work shows a simple thing: reports for companies must be easy to read and decision-oriented, not methodologically rigorous (A Concise Guide to Marketing Research, ch. 1.3). If the analysis requires a translator, it will not be used. That is why in practice it is better to show three numbers with a clear assumption than twenty tables with footnotes. We write more about how to translate analysis findings into concrete marketing actions in the article on marketing audit.

Market size: TAM, SAM, SOM

Market size, TAM, SAM, and SOM, is the most commonly confused set of terms in strategy discussions. TAM is the entire market we could theoretically operate in, if there were no geographic, production, or budget constraints. SAM is the portion of that market we can realistically access: with our product range, within our logistics reach, through our channel. SOM is the share we are able to capture within the planning horizon, given our current resources and competitors who are not standing still.

The number one mistake is calculating TAM and treating it as a target. A company producing packaging for the food industry in a single region can calculate the European packaging market and arrive at a number that has no connection to its actual capabilities. That number looks good in an investor presentation and works terribly on the team, because it shows a target that cannot be addressed. TAM only makes sense as context: it shows whether expansion is worth considering at all, whether the market is large enough that even a small share generates meaningful revenue.

The number two mistake is calculating SOM as a percentage of SAM. It is convenient, but it adds nothing. SOM has to be built from the bottom up: how many clients we can serve with the current number of salespeople, what the average sales cycle is in this segment, what a typical share of a client's wallet looks like, how many of those clients are even willing to switch suppliers. Only the product of these assumptions gives a number that can be compared with the sales plan. That is when you can see whether the plan is ambitious or fantastical.

Third: TAM, SAM, and SOM must be calculated in the same unit. Most often this is a monetary value, but in B2B the number of clients or the number of installations is often better. If we mix units, the model stops being verifiable. It is also worth recording where each of the three numbers comes from and how certain they are. In practice, it is assumed that with industry data from external reports the approximate uncertainty is on the order of 20 to 30 percent, while with data from company registries for specific firms it is much smaller. This is not a measured value, just an order of magnitude that helps assess how heavily you can rely on a given number. This distinction matters when someone later asks why the forecast did not hold up.

B2B market trends

B2B market trends have one feature that makes the work harder: they are slower than in B2C, but because of that they are harder to spot. A change in consumer behavior shows up in sales data within weeks. A change in how an industrial client places orders plays out over quarters, sometimes years. This means that a single reading, even a clear one, is almost never a trend. You need at least three points in time and a check of whether the change appears across many clients or just one large one.

A practical method is to divide trends into three categories. The first is structural trends that change the rules of the game: production automation, a shift toward smaller and more frequent orders, growing requirements around documentation and carbon footprint. The second is cyclical trends that result from the economic cycle and will reverse on their own. The third is noise, single events with no consequences. Most strategy mistakes come from confusing the third category with the first.

It is also worth remembering that a trend seen from the demand side and from the supply side can look different. Clients may declare they are looking for local suppliers while simultaneously choosing a cheaper offer from abroad. This is not hypocrisy, it is the difference between a declaration and a behavior. That is why a trend has to be confirmed in transactional data, not just in conversations. If we do not have such data, we need to state clearly that it is a hypothesis, not a fact.

Trend analysis also benefits from looking at how audience attention is distributed across different information sources. Research on advertising in new formats shows that audiences are set on avoiding advertising messages and pay attention primarily to content that is useful to them (Advertising in New Formats and Media, ch. 8). In B2B this works the same way: if we describe a trend only as a slogan, the client will ignore it. If we show what specifically changes in their purchasing process, then the conversation makes sense.

B2B market segmentation

Market segmentation in B2B most often stops at industry codes, and that is its biggest limitation. Two companies in the same industry can have a completely different purchasing method, a different decision cycle, and a different sensitivity to price. One buys through a tender and looks only at the specification, the other buys through a relationship with the technologist and values speed of response. Segmentation by industry will not show this.

Better criteria are how the purchase happens and how the product is used. In practice, four dimensions work well: who makes the decision (owner, technical department, purchasing department), how long the cycle is (days, weeks, quarters), how the customer evaluates a supplier (price, delivery time, technical support, documentation), and how large the typical volume is. These four dimensions usually produce three to five segments that can be served through different channels and different messaging. We cover how to translate segments into communication in more detail in our article on market segmentation.

One thing is worth keeping in mind here, as the literature on structural models in marketing emphasizes: a model describes the behavior of actors and the relationships between variables, not just a division into categories (Handbook of Marketing Decision Models, chapter 6). In other words, segmentation makes sense when we can say how a customer in a given segment makes a decision and what changes it. A division into groups without a description of behavior is just labels that add nothing to the sales plan.

The practical consequence: segmentation should end with a decision about resource allocation. If after the division we do not know which segment should get more salespeople, where to invest in marketing, and where to limit service, then segmentation is descriptive rather than decision-driving. It is worth asking how much revenue each segment generates and what the cost of serving it is. That shows where we actually make money.

Competitive analysis in B2B

In B2B, complete data on competitors almost never exists. There are financial statements in the company register, there are press releases, there are offers that customers show during conversations, there are changes in employment visible in job postings. That is enough to build a picture of the competition at a level that supports decisions, as long as we do not try to pretend we have data we do not have.

The most underrated source is lost deals. If a company systematically records who won a tender and at what price, after a few quarters it has a map of competitors' positioning that cannot be bought in any report. This is data on specific transactions, not estimates. It is worth collecting even when we are not running an analysis on an ongoing basis, because the value only appears over time.

The second source is changes in the offer. A new product, a new service, a new sales channel at a competitor signals that they see a gap somewhere. There is no need to react immediately, but it is worth understanding why they are doing it. Sometimes it is a response to a trend, sometimes to a change in costs, and sometimes simply an attempt to take a place that opened up after another player left.

The third thing: competition in B2B is rarely all-or-nothing. The same customer often buys from two or three suppliers, splitting orders. This means the goal is to increase our share of the customer's wallet rather than eliminate the competitor. Market analysis should therefore answer the question of what our current share is with key customers and what would have to change for it to be larger.

When market analysis delivers results, and when it only costs money

Market analysis delivers results when it is tied to a decision, has defined assumptions, and is updated. The most common mistake is a one-off project that ends with a report, and then nobody returns to it. The market changes, assumptions age, and decisions are made based on a document from two years ago.

The second thing is scope. Market analysis does not have to be a six-month project. Often a few days of work are enough to organize the data the company already has and calculate TAM, SAM, and SOM on reasonable assumptions. The value lies not in volume but in the fact that the numbers are consistent and their origin is known.

The third thing is combining market analysis with an analysis of our own marketing. Even the best market model will not help if sales and communication channels do not reach the chosen segments. That is why it is worth checking in parallel whether current marketing activities respond to what the analysis shows. If not, the first step may be a marketing audit, which will show where we are losing budget and where there is room for growth.

At Growth Group, we work with B2B companies across the region. For over 16 years we have served more than 60 clients, and we have generated PLN 160 million in profit for clients. This experience shows one thing: companies that regularly return to market analysis and compare it with their own sales data make decisions faster and backtrack less often. Not because they have better data, but because they know what they are looking for.

If you want to check how this looks in your company, start with three numbers: TAM, SAM, SOM. Calculate them on current assumptions, write down where they come from, and compare them with the sales plan. If the numbers do not match, that is exactly the moment to do a proper market analysis and market segmentation before the next planning cycle.