Where your revenue breaks and what it costs you: a B2B revenue diagnostic, stage by stage

What is a revenue diagnostic?

It's an analysis of the whole chain that turns a stranger into a client who pays and comes back. It looks at each stage with your data (invoicing, opportunities, proposals, analytics), finds where the most value is lost, puts a figure on it and sets the order in which to fix things.

What you get isn't a report to read, but a decision: which stage is holding you back, what it costs and where to start. At RevOps Studio it's delivered in under four weeks, with scope and timeline fixed before we begin.

How is it different from a CRM report or a sales review?

A CRM report tells you what happened inside the CRM. If the CRM doesn't record why opportunities are lost, neither does the report. A sales review checks whether the team does what it should. A diagnostic asks a different question: where does the chain break, and what does that break cost?

The practical difference: a review finds shortcomings, and almost every company has plenty. A diagnostic finds which of them is holding back growth right now, and what leaving it unfixed costs.

What is the revenue chain, and how do you read it stage by stage?

The revenue chain is the eight stages every B2B client goes through, whether you track them or not:

Each stage gets a figure and a status:

• Broken: it doesn't perform, or it doesn't exist.
• Limited: it works, but it's at its limit or depends on a single resource.
• Healthy: it works and isn't a priority for intervention.

The most useful reading rule: the symptom is rarely where the cause is. A full sales pipeline where deals don't close is usually a qualification problem (stage 4), not a closing problem (stage 7). Before investing in the stage where you see the symptom, check the two before it.

How do you put a figure on a bottleneck?

Every bottleneck is described the same way: where it is, what happens, why it happens, how much it weighs, what evidence supports it and how confident we are. The figure doesn't have to be exact. It has to be defensible and say where it comes from.

Three real examples, in relative figures:

1. Capacity arithmetic (a creative studio). Reaching its original target through project work needed around thirty wins a year. Each win took about ten hours of the managing director's selling time, and she had about two hours a month to spare. Thirty wins at ten hours each is about 25 hours a month: roughly twelve times her available selling capacity. The target wasn't failing for lack of demand, but for lack of hours.

2. Retention scenarios (a B2B marketing services company). More than six in every ten euros of the year came from new clients, and 73% of the previous year's clients hadn't come back, with the year still open. If the share of revenue retained from one year to the next stayed at around 40%, reaching the target meant growing new business by 73% to 98%. If it rose to 60%, 41% to 65% was enough. The figure doesn't say "retain more": it says what each point of retention is worth against each point of new business.

3. Contracted versus invoiced (the same creative studio). Only around 28% of the value contracted in the year's projects had been invoiced. The monthly report counted as closed about 2.2 times what had been recognised as revenue. All four figures were correct by their own definitions, yet the report showed progress in a year of contraction.

What if your data isn't clean?

It almost never is, and it doesn't need to be. Every figure in the diagnostic carries a label that says where it comes from:

The assumption with a test is the important part. If you don't know why you lose opportunities, the diagnostic doesn't stop: it assumes a cause, states it and says which data point would disprove it. Sometimes that data arrives within weeks, and the assumption holds or falls.

Something you can't measure is also a finding. In a digital services agency, the margin on the capacity it bought from others wasn't recorded anywhere. The diagnostic didn't make it up: it flagged it as blocking any growth decision and made it the first thing to measure.

Confidence is stated separately for the fact and for the cause. You can be very sure of a figure and much less sure of why it happens. The sample diagnostic closes with an overall confidence of 3 out of 5, and says what would raise it.

Where does revenue usually break? The patterns that repeat

Eight patterns repeat across the diagnostics we've run. No company has all of them, and almost every company has three or four.

1. The relationship ends at delivery. Nobody owns bringing the client back. In a B2B marketing services company, only 27% of the previous year's clients had bought again by mid-year.
2. Lead source and loss reason aren't recorded. Decisions run on perception. In a digital services agency, a channel credited with 10% of new business had brought in 5 of 12 accounts.
3. The first stage doesn't exist. Everything comes through personal networks. In our own diagnostic, visibility on the topic that decides the purchase was about four times lower than the most visible competitor's.
4. Revenue sits in a few accounts. In a creative studio, almost all recurring revenue came from a single account.
5. All selling runs through one person with no time. The same studio needed about twelve times the selling hours it had available.
6. The offer isn't packaged, and price is set deal by deal. In the marketing services company, 100% of invoice lines went under a single generic item, so nobody knew what was really being sold.
7. Qualification is informal and stages are set by feel. In that same company, 58% of the year's opportunities were still open at mid-year.
8. Closing is rarely the bottleneck. In three of the four diagnostics, closing worked; in the fourth it couldn't be measured reliably. In the creative studio, the negotiation stage converted 84%. The problem sat earlier or later.

What is not the problem?

This section exists to stop you investing in the wrong place. Many companies arrive with an explanation (too few leads, poor closing, the wrong tool, prices too high) and the data says something else:

Ruling out a cause with data is a decision that saves money: the CRM you don't replace, the salesperson you don't hire, the campaign you don't launch.

Why does my sales forecast never come true?

Usually because it's built on stages set by feel. If nobody agrees what moves an opportunity from one stage to the next, the forecast adds up hopes. In one diagnostic, 58% of the year's opportunities were still open. In another, the monthly report counted contracted value as closed, and invoicing told a different story.

Before you change how you forecast, decide what each stage means and what evidence moves an opportunity forward. The forecast improves on its own when the stages describe how you really sell.

How do you decide what to tackle first?

With an impact-by-effort matrix, in three groups:

• Quick wins: high impact, low or medium effort. Done now.
• Structural investments: high impact, high effort. Planned.
• Background: the rest. Noted and parked.

It nearly always produces a two-speed sequence: what pays back in weeks (setting up the CRM properly, a renewal routine with an owner) and what matures over quarters (visibility, offer, structure). And one rule: whatever costs money comes after the checkpoint, not before.

A named example: our own diagnostic

We also ran the Diagnostic on ourselves. The finding with the highest confidence was the most uncomfortable one: against a target of 10 to 12 qualified newcomers, 1 had come in. Cold outreach was answered at a normal B2B rate (3.4–4%) but produced no meetings, and 85% of the contact base bought for it was still untouched.

It also said what wasn't the problem. The measured sales cycle ran from 19 to 63 days, not the three to six months we had assumed, and two of every four priced proposals were won. The bottleneck was at the way in, not at the close.

At the checkpoint we accepted part of the recommendation and decided against it on other points. The riskiest decisions were written down as hypotheses, with the test that would disprove each one. Read the full case →

What do you decide at the end?

The diagnostic ends at a checkpoint, which is a validation meeting, not a presentation. You arrive having seen the work and decide with it in front of you:

• whether you confirm the diagnosis and the order of priorities;
• which assumptions you accept and which you want checked first;
• which segments and what scope the work continues with, if it continues.

You can decide against the recommendation. Then nobody insists: the risk is stated, and it's written down how you'll know if you were wrong. And if the diagnostic finds no actionable bottlenecks, it stops there.

Can I do this myself, with my team or with AI?

You can do a good part of it. This guide gives you the chain, the labels and the matrix. The hard part isn't the method, but three things: cross-checking sources that don't talk to each other, accepting findings that contradict what you believe about your own business, and having no stake in any conclusion.

An AI organises what you give it. The Diagnostic finds what contradicts itself between what you declare, what you invoice and what you measure.

Frequently asked questions

How long does a revenue diagnostic take? RevOps Studio's takes under four weeks, with scope and timeline fixed before we start. The clock starts once we have your data.

What data does it need? What you already have, as it stands: invoicing by client and month, your record of opportunities (a CRM or a spreadsheet), the proposals you've sent, contracts with recurring clients and your web analytics, if you have them.

Does it work without a CRM? Yes. One of the diagnostics in this guide was done without a CRM, from spreadsheets. Whatever isn't recorded is stated as an assumption.

Why a paid diagnostic rather than a free audit? A free audit is a sales pitch; a paid diagnostic is work you can decide on.

What if the diagnostic finds nothing? If there are no actionable bottlenecks, it stops there, and you know it from the data. That's a useful answer too.

When is the right moment for one? When there's a significant investment on the table (a hire, an agency, a new CRM) and no figure to decide it with.