You Don't Have A Lead Problem.

Ask a room of independent consultants to name their biggest problem and 59.6% say finding clients. Six percent say keeping them. Those are real figures, from the Frak Conference State of Fractional report, 250 professionals across 29 US states. I believe both numbers. I also think the second one is what produces the first.

A month with no new work announces itself. A client leaving does not. There is no notice, no cancellation and no meeting. There is a last email, and then there is a quiet you read as a busy quarter.

WHAT WORKS

Count last year's money by where it came from.

Open every invoice you sent in the last twelve months, one line per client. Mark each client A if they were already paying you in September last year, B if they were not. Total the two columns.

Now the third number, which is the one nobody keeps. List every engagement that ended in those twelve months where nobody ever said why. Leave out the ones that closed on a stated decision: budget cut, in-house hire, scope genuinely finished. You want only the ones that stopped. Next to each, write what that client billed you in the year before it stopped.

Subtract that total from column B.

What is left is growth. The rest is replacement, and replacement is the most expensive revenue you will ever bill, because you already had it once.

Here is the arithmetic that makes it expensive. A new $2,000 engagement costs you an intro call, a proposal, two follow-ups and a scoping call. Call it eight hours before the first invoice, none of them billable. Keeping a client you already have costs one honest thirty-minute call about what they need next. So every quiet exit sets you back about eight unpaid hours to stand exactly where you were already standing.

Mine last year was a retainer that ran for eleven months and then did not. There was no complaint. Their replies went from one a day to one a week, then to a file I had to ask for twice, and I read all of it as their busy season. It showed up in my invoices as a gap I filled in June with somebody new, after about nine hours of calls and proposals, and I remember counting that June as a good month.

The client you win in March to cover the one who went quiet in January is not a new client. That is the same revenue, bought twice.

Which is why retention comes last on the survey and first on the bank statement. Finding clients hurts in a way you feel on a Tuesday. Losing one hurts in a way that only ever shows up in a column.

ONE PRACTICAL IMPLEMENTATION

Here is a year filled in. Round numbers, invented, but the shape is the one you will find.

Twelve months, 96,000 billed.

• →Column A, clients already yours last September: 54,000.
• →Column B, clients who were not: 42,000.
• →Engagements that ended: five. Three of them on a stated decision. Two that simply stopped.
• →What those two billed in the year before they stopped: 18,000.

Column B is 42,000, and 18,000 of it replaced revenue you already had. The year grew by 24,000. You ran a full year of pipeline for 42,000 and kept a little over half.

Two things follow from that. Your win rate was never the problem, so working the pipeline harder fixes nothing. And the 18,000 was recoverable at a fraction of what the 42,000 cost you, at a point in the year when nobody was looking at it.

Replacement revenue looks exactly like growth on an invoice. It only looks different on the year.

Now the automation question, because this newsletter is supposed to answer it. The counting is worth automating. It is the same arithmetic every quarter, on invoices and threads you already hold, and the whole value is running it four times a year rather than once while you re-forecast in January. Noticing is worth automating, which is what last week's message ledger does. The outreach at the far end is not, and it is the first thing people reach for. A scheduled "just checking in" aimed at a client who has gone quiet is the exact message that taught them to stop replying.

WHAT DOESN'T WORK

• →Measuring the pipeline every week and the base once a year. The pipeline is the part you already worry about, which is why it is the part that gets managed.
• →Reading a finished project as a closed loop. Most engagements do not end on a decision. They end with a last email that nobody recognised as the last one.
• →The exit survey. You are asking someone who already left to explain a choice they made months ago, and you will get the polite version of it.
• →Ranking your problems by how they feel in the week you are in. That week is always the week with the invoice gap, so the ranking always comes out the same.
• →Buying lead generation to cover a base that is quietly draining. It works, at roughly eight unpaid hours per replaced client, for as long as you keep paying.

ONE THING TO DO THIS WEEK

Thirty minutes with last year's invoices.

• →Open every invoice from the last twelve months, one line per client.
• →Mark each client A if they were already yours in September 2025, B if they were not.
• →Total both columns.
• →List every engagement that stopped with nobody saying why, and what it billed the year before.
• →Subtract that total from B. The number left is what your year actually grew by.

---

Every quiet exit was visible months before it finished, in a thread that got shorter, and a thread going quiet leaves nothing behind to remind you it is there. That is the job Alacrio does for me: it watches the client conversations I am not currently thinking about and tells me which ones have gone still.

Do the subtraction first. If column B turns out to be mostly replacement, book a free consultation and we will work out where it went.