Strategy

What a Customer Is Worth Over the Whole Relationship

The average customer value is the least useful number in your business, because nobody is average. Segment-level numbers are what you act on. Here is the simple calculation for a small B2B company, and how it changes what you are willing to spend to win the right one.

November 2, 2026 · 8 min read · By José Cabal

Every owner has a rough sense of which customers are the good ones. Ask them to put a number on it and the conversation stops.

That missing number decides more than people realise. It sets what you can afford to spend winning a customer, which customers deserve a proposal and which deserve a price list, whether a discount is a smart move or a slow loss, and what your business is worth if you ever sell it. Without it, every one of those calls is made on a feeling about who is nice to deal with.

This is the value side of a pair. The cost side, what you pay to acquire a customer and how quickly you get it back, is covered in the post on acquisition cost and payback period. Here the question is the other half: what is the thing you bought actually worth.

The calculation for a small B2B company

Three inputs, multiplied.

What a customer pays you in a year. Add up a full year of invoices for a customer and use the real total rather than the contract value, because add-ons, extra jobs, and price changes all sit in the real number.

Your gross margin on that work. What is left after the direct cost of delivering it, before your overhead. If you do not know this per customer, use your overall gross margin and note that you did.

How many years they stay. This is the input people guess at, and you do not need to guess. Pull your customer list from five years ago and count how many are still buying. If 60 of 100 customers from 2021 still buy today, your average relationship is running past four years and you can work with that.

Multiply the three. A customer paying $30,000 a year at 35 percent margin for four years is worth $42,000 in gross profit over the relationship.

For a project business where customers buy occasionally, change the middle input. Count the average number of projects a customer commissions over their relationship with you and multiply by the gross profit per project. A customer who does three $20,000 projects at 30 percent margin is worth $18,000. If most customers do exactly one project, your answer is the profit on one project and that itself is the finding.

Ignore the discounting formulas you will find elsewhere. Adjusting for the time value of money changes the answer by a small amount and stops most owners from ever finishing the calculation. Get the plain number first.

The average hides everything useful

Here is the trap. You do the calculation across all your customers, get $28,400, write it down, and nothing about your business changes.

The reason is that no customer is worth $28,400. In a typical small B2B firm, the top fifth of customers are worth five to ten times the bottom fifth, and the average sits in a gap where almost nobody actually lives. Acting on it means spending the same to win a $90,000 customer as a $6,000 one, which is how a marketing budget produces a full pipeline and a flat profit line.

The blended number also hides the direction of travel. If your best segment is shrinking while your worst is growing, the average can hold steady for two years while your business quietly gets worse. I have seen that pattern in more than one company where the owner was confident things were stable.

So calculate the blended figure once, as a reference point, then stop using it for decisions. Everything that follows happens at the segment level.

Cutting your customers into segments that mean something

Export your customer list with a year of revenue against each name. Then cut it four ways and see which cut produces the widest spread in value.

By industry. Do engineering firms stay longer than retailers? This is often the sharpest split in B2B, because different industries buy at different rhythms and have different budget stability.

By company size. Larger customers usually pay more and stay longer, and they also cost more to serve and to win. Check whether the extra revenue survives the extra cost, because at some firms it does not.

By how you acquired them. Referrals commonly stay longer and buy more than customers won from advertising, and that gap has direct implications for where the budget goes. Measure it rather than assuming it.

By what they first bought. Customers who start with your larger service often behave differently from those who start with a small trial job. If one entry product produces long relationships and another produces one-off buyers, that changes what you promote.

The cut that matters is whichever one produces a clear difference between segments. If splitting by industry gives you $52,000 against $9,000, that is a finding you can act on. If every cut gives you numbers within 20 percent of each other, your customer base is uniform and you should spend your attention elsewhere.

Keep segments large enough to mean something. Eight customers in a segment is a story, not data. With a customer list under 40, use two segments rather than five.

How this changes what you will spend to win a customer

This is where the number earns its keep.

Most small companies set one marketing budget and one idea of what a lead is worth. Once you know that a manufacturing client is worth $60,000 in gross profit over their relationship and a small retail client is worth $7,000, you can justify completely different spending on each.

A practical rule for an owner-funded business is to be willing to spend up to a quarter of a segment's lifetime gross profit to win one, provided the payback period stays inside your cash tolerance. On a $60,000 segment that is $15,000 per customer, which opens options that were unthinkable under a blended budget: direct outreach with real research behind it, trade shows in that industry, hosting a small event, hiring someone who knows that sector.

The same rule on a $7,000 segment gives you $1,750, which means that segment has to be won through channels that cost almost nothing per customer. Search visibility, referrals, and a website that converts on its own.

Both are correct. What is wrong is applying one figure to both, which either starves the valuable segment or overspends on the cheap one. Before you act on the number, check the payback stays survivable, because a segment that is worth $60,000 over six years still has to be funded out of this year's cash.

Deciding which customers to chase more of

Build a small table: segment, number of customers, lifetime gross profit each, what it currently costs you to win one, and how hard they are to serve. Five columns and usually five rows.

Read it in a specific order. Where value is high and acquisition cost is low, do more immediately, and be specific about how: more outreach to that industry, a case study written for them, presence at their events. Where value is high and acquisition cost is high, the work is to bring the cost down before scaling, usually through better qualification and a sharper message. Where value is low and cost is low, keep it running and stop thinking about it. Where value is low and cost is high, stop, and expect that to be the segment somebody on your team likes most.

The uncomfortable outcome shows up regularly: the segment producing the most enquiries is the least valuable one. That happens because cheap-to-reach customers are cheap to reach for everyone, so they arrive in volume. Volume is not the goal.

One caution. The difficulty-to-serve column is not decoration. A high-value segment that consumes your team's patience and produces no referrals can be worth less than the arithmetic says. Keeping the customers you already have in the valuable segments is usually cheaper than winning replacements, which is the case made in the post on customer marketing and retention. Lifetime value tells you what one customer is worth; net revenue retention tells you whether your whole existing base is growing or shrinking, and the two numbers answer different questions.

The danger of over-modelling this

I have watched owners disappear into this calculation for three weeks. It is a comfortable place to hide, because the spreadsheet always has one more refinement available and none of them require you to talk to a customer.

Three specific traps. Precision beyond your data, where a number carried to two decimal places rests on a retention estimate you guessed at. Predicting the future from a small sample, where a segment with eleven customers gets a confident lifetime value that will not hold. And modelling scenarios instead of acting, where three versions of the spreadsheet exist and nothing in the business has changed.

The test is simple. Would a 25 percent error in this number change the decision you are about to make? If a segment is worth $60,000 and another is worth $7,000, being wrong by a quarter changes nothing about which one you chase. That is the level of accuracy this calculation needs.

Redo it once a year. More often than that and you are watching noise.

Connecting the number back to who you target

The finding is only worth having if it changes what your marketing says and who it says it to.

Three concrete changes. Rewrite your main pages to speak to the valuable segment specifically, naming their industry and their problem, so that a visitor from that segment recognises themselves. Change the questions your team asks early in a sales conversation so you know within five minutes which segment a new enquiry belongs to and can spend your time accordingly. And write your proof for that segment, because one case study about a company like them beats six about companies unlike them.

This is also the input that makes an ideal customer definition real instead of aspirational. Most definitions are written from a description of who the company would like to serve. Yours now comes from which customers have actually been worth the most, which is a better foundation. The process for turning that into a working target is in the ideal customer definition playbook, and the positioning framework covers how to say it once you know it.

The takeaway

Do the calculation once at the whole-company level so you have a reference point, then throw it away and do it by segment. The gap between your best and worst segment is the finding, and at most small B2B firms it is a factor of five or more.

Then let that gap set different acquisition budgets rather than one. Give yourself a single annual update and a rule that a 25 percent error would have to change the decision before you refine anything further. The point of knowing what a customer is worth is spending differently on different customers, and if your spending looks the same next quarter, the calculation did not do its job.

About the author

José Cabal, AI Marketing Consultant

AI Marketing Consultant · HubSpot Certified Trainer · Vancouver, BC

José Cabal has 20+ years of B2B marketing experience across Canada, the US, and Latin America, working with companies in tech, real estate, professional services, health, education, and manufacturing. He is a HubSpot Certified Trainer, holds an MBA from University Canada West, and founded Expresa3 in 2010 — one of the first inbound marketing agencies in Latin America. He has trained hundreds of marketers in English and Spanish, served on the HubSpot LATAM Partner Advisory Council, and spoken at conferences and universities across North America. He currently consults for SMBs building AI-powered marketing systems alongside his role as Head of Marketing and Business Systems for a growing company.

More about José Cabal · LinkedIn

Frequently asked questions

How do I calculate customer lifetime value for a small B2B company?

Multiply three things: what a customer pays you in a year, your gross margin on that work, and how many years they stay. A customer paying $30,000 a year at 35 percent margin for four years is worth $42,000 in gross profit. Skip the discounting formulas, because adjusting for the time value of money changes the answer a small amount and stops most owners from finishing.

How do I find out how long customers actually stay?

Pull your customer list from five years ago and count how many still buy today. If 60 of 100 customers from 2021 are still active, your average relationship runs past four years. This takes twenty minutes and it replaces the input that most people guess at.

Why is average customer lifetime value misleading?

Because the top fifth of customers at a typical small B2B firm are worth five to ten times the bottom fifth, so the average lands in a range where almost nobody sits. Acting on it means spending the same to win customers of completely different value. Calculate it once as a reference, then work at the segment level.

How should I segment my customers to calculate value?

Cut the list by industry, by company size, by how you acquired them, and by what they first bought. Use whichever cut produces the widest spread between segments, because that is the one with a finding in it. If every cut produces numbers within 20 percent of each other, your customer base is uniform and your attention belongs elsewhere.

How much should I be willing to spend to win a customer?

A workable rule for an owner-funded business is up to a quarter of that segment's lifetime gross profit, provided the payback period stays inside what your cash can absorb. On a $60,000 segment that is $15,000 per customer. On a $7,000 segment it is $1,750, which means that segment has to be won through low-cost channels.

How do I calculate lifetime value for a project business?

Replace the years input with the average number of projects a customer commissions over the relationship, then multiply by gross profit per project. Three $20,000 projects at 30 percent margin is $18,000. If most of your customers do exactly one project, that answer is itself the most important thing you learned.

How many segments should a small company use?

Two or three if your customer list is under 40 names, and up to five above that. Eight customers in a segment is an anecdote rather than data, and splitting too finely produces confident numbers built on nothing. Keep segments large enough that one unusual customer cannot move the result.

How often should I recalculate customer lifetime value?

Once a year. Recalculating more often means watching noise, since the inputs move slowly and the retention figure needs years of history to shift meaningfully. Set an annual date, tied to your fiscal year end if that is easier to remember, and use the intervening months to act on what the last calculation told you rather than rerunning the number.

When is the customer lifetime value calculation not worth doing?

When a 25 percent error in the answer would not change the decision in front of you. If one segment is worth $60,000 and another $7,000, refining either number further changes nothing about which one you chase. That test also tells you when to stop refining a spreadsheet you have been in for three weeks.

What if my most valuable segment is also the hardest to serve?

Add difficulty to serve as a real column in your table rather than treating it as a footnote. A high-value segment that consumes your team's patience and produces no referrals can be worth less than the arithmetic suggests. Price for the difficulty or narrow the segment to the parts of it that work.

How does lifetime value change my marketing?

Three ways. Rewrite your main pages to name the valuable segment's industry and problem so they recognise themselves. Change the questions your team asks early so you can tell within five minutes which segment an enquiry belongs to. And write your case studies about that segment, because one relevant example beats six irrelevant ones.

Should referrals be treated as a separate segment?

Yes, and it is one of the more revealing cuts. Referred customers commonly stay longer and buy more than customers won through advertising, and the size of that gap has direct budget implications. Measure it in your own data rather than assuming it, because the difference varies a great deal by industry.