ABC analysis: how to organize your customers and products

ABC analysis answers the question of who really keeps the business running when your sales spreadsheet has 400 customers and 120 product lines, and you still cannot point to them. The problem lies in how the data is organized.ABC analysis answers one specific question: what actually drives results, and what merely consumes your team's time.

In our work as an agency partnering with B2B companies, we have seen the same pattern for years. Companies do not have too few reports. They have too many reports in which everything looks equally important. ABC analysis introduces hierarchy where there was previously a flat list.

What ABC analysis delivers when a company has too much data

Let us start with a situation rather than a definition. Mark, the owner of a service company, spent every Monday scrolling through a spreadsheet of 180 B2B customers without knowing where to begin. Salespeople complained they could not keep up. The support team was drowning in tickets. Leadership debated hiring another person. Yet nobody checked how many of those 180 customers accounted for 80 percent of revenue.

ABC analysis is a method for ranking items (customers, products, channels, SKUs) by their share of a chosen value. Most often that value is revenue, but it can equally be margin, order volume, or customer lifetime value. Items fall into three groups:

• Group A: the smallest in number, generating the largest share of value. Typically, around 10 to 20 percent of items account for roughly 70 to 80 percent of the result.
• Group B: the middle layer. Items that do not determine the company's survival but carry real significance.
• Group C: the most numerous, with the smallest share of value. This is often where the most operational work hides.

The value of ABC analysis lies not in the segmentation itself. It lies in what you do afterward. Group A receives a different service standard, different terms, and a different contact frequency than Group C. That sounds obvious, but in most companies we talk to, all customers are treated the same way. The result is predictable: your best customers do not feel recognized, and your weakest ones absorb resources that should go to Group A.

Before you move to implementation, it is worth checking whether your marketing and sales data even allow for such a split. If you do not know where your best customers come from, a sales-only segmentation will give you just half the picture. In that case, a sensible first step is a marketing audit, which organizes your data sources before you make segmentation decisions.

How customer and product classification by value works

The mechanics are simple, but the devil is in the decisions about what you measure. Classifying customers and products by value requires choosing a single axis of ordering. Not two, not three. One axis, one decision, one ranking.

Step by step, it looks like this:

1. Choose the value. Revenue is the simplest option, but in B2B gross margin is often better. A customer with high turnover and zero margin lands in group A only because they generate traffic. That is misleading.
2. Set the period. Twelve months is a reasonable horizon. Three months favors seasonality, three years loses changes in the portfolio.
3. Sort descending. From the highest value to the lowest.
4. Calculate the cumulative share. The first 70-80 percent of value is group A. The next 15-20 percent is group B. The rest is group C.
5. Record the size of each group. This is the most important number in the whole analysis. If group A is 12 customers and group C is 140, you already know where the team's time goes.

In the literature, this stage is simply called data analysis and is often treated as a formality. In practice it is the hardest part, because it requires deciding what not to measure. The quality of conclusions depends on whether the input data is consistent and complete. In ABC analysis this means one thing: if half of your customers in the CRM have zero revenue recorded because the salesperson did not fill in the field, no segmentation will be true.

ABC analysis of products works differently than for customers

With products you need to add a second dimension: margin. A product can be in group A by revenue and at the same time in group C by profitability. This is a classic trap for manufacturing and distribution companies in the region. The largest volume goes to a product that, after subtracting logistics and return handling costs, earns pennies.

That is why with products you run two analyses in parallel: ABC by revenue and ABC by margin. Only comparing both shows which items are truly worth promoting, and which are worth repricing differently or discontinuing.

How to start an ABC analysis in practice

The most common mistake at the start is trying to run an ABC analysis for the entire company at once. That guarantees the project will stall at the stage of agreeing on definitions. Start with one area where decisions are most urgent.

A practical sequence:

• Choose one area. Customers or products. Not both at the same time.
• Collect data from a single source. A CRM, an ERP system, an invoice spreadsheet. One source of truth.
• Do the segmentation manually in a spreadsheet. With 200 items you do not need a BI tool. You need two hours and sorting.
• Name the groups and assign rules. What exactly changes for a group A customer, and what changes for a group C customer.
• Decide when you will repeat the analysis. Once a quarter is a reasonable rhythm. Once a year is too rare, because the customer portfolio changes faster.

Only after the first cycle is it worth thinking about automation. By then you already know which CRM fields must be filled in for the analysis to make sense. Implementing a tool before the first manual segmentation is the most common way to burn through budget with no result.

If the ABC analysis reveals that group A consists of customers acquired through a channel you had not previously taken seriously, that is a signal that these findings should feed into an action plan. A B2B marketing strategy built on a real customer portfolio rather than on assumptions from a presentation helps with this.

What an ABC analysis will not show

An ABC analysis is a tool for organizing, not for diagnosing. This distinction matters, because many companies expect it to answer questions it does not resolve.

It will not show you why a group A customer is leaving. It will not show whether a group C customer has the potential to move into A within a year. It will not show whether the current split is simply the result of chance, for example one large contract signed three years ago that is about to end.

That is why an ABC analysis is worth supplementing with two elements. The first is customer potential, meaning how much a customer could buy from you if you expanded the relationship. The second is the cost to serve, which in B2B is often underestimated. A group B customer who generates 30 service requests a month may be genuinely less profitable than a group C customer who buys once a quarter and never calls.

An ABC segmentation is a starting point for decisions, not the decision itself. Treated as a final answer, it leads to wrong conclusions faster than no analysis at all.

The most common mistakes during implementation

From our experience with projects for companies across different regions, we see a recurring set of pitfalls. It is worth knowing them before you begin.

• Analysis without decisions. The team produces a nice chart, but nobody changes the service rules. A quarter later, everyone is back to old habits.
• Treating group C like garbage. Group C is often the source of clients who will be in A a year from now. The point is to serve them more cheaply and simply.
• No internal communication. If a salesperson finds out their client was moved to group C without explanation, they lose trust in the entire analysis.
• Analysis once and forgotten. The client portfolio is alive. Without regular repetition, the segmentation becomes outdated within a few months.
• Mixing criteria. One analysis by revenue, another by margin, a third by number of orders, all on one chart. This leads to conclusions nobody can defend.

The last mistake is particularly costly because it undermines trust in the entire method. Better to do one analysis well and return to it a quarter later than to do three superficially and lose credibility in the eyes of the team.

Conclusions

ABC analysis does not require advanced tools or a large budget. It requires a decision about what you measure and consistency in changing service rules after segmentation. Companies that do this regularly stop guessing where the team's time goes.

At Growth Group, we have been working with B2B companies for over 16 years. Growth Group generated over PLN 160 million in profit for more than 60 clients. ABC analysis is often the first step in projects where sales, marketing, and leadership previously lacked a shared language.