Why agencies with full pipelines still run out of money

Pipeline is a measure of possible future revenue. It is not a measure of cash, margin, or capacity — and agencies get into trouble on all three while the pipeline number looks reassuring.

The failure has a specific shape. Sales is going well. The board or the spreadsheet says $1.2m of opportunity. Everyone relaxes about selling and concentrates on delivery. Four months later there's a payroll problem that nobody saw coming, because nothing in the pipeline number was ever about the next four months.

Timing: the gap between "won" and "banked"

Take a pipeline of $1.2m across four stages.

Stage	Value	Win rate	Weighted	Typical weeks to cash
Early conversations	$240,000	10%	$24,000	28
Proposal sent	$480,000	25%	$120,000	20
Verbal yes	$300,000	70%	$210,000	14
Contract out	$180,000	90%	$162,000	10
Total	$1,200,000		$516,000

A blended 43% conversion on $1.2m. Nothing wrong with that.

Now ask the only question that matters for the next quarter: how much of it becomes cash within twelve weeks? Only the contract-out stage clears that window. That's $162,000 of weighted cash arriving in the next three months.

Against a cost base of $85,000 a month, twelve weeks costs $255,000.

$255,000 cost − $162,000 expected cash = $93,000 short

The pipeline is healthy and the quarter is $93,000 underwater. Both are true at once, and the pipeline number cannot tell you the second thing because it has no time dimension at all.

The gap is structural, not a sales failure. A deal at "verbal yes" today still needs a contract, a start date, a first milestone, an invoice and 60 days of payment terms. Fourteen weeks is optimistic, not pessimistic.

Margin: a pipeline can be full of work worth having and not worth doing

The second failure is that pipeline is measured in revenue, and revenue is the number least connected to whether you survive.

If your pipeline is weighted toward competitive pitches won on price, you are forecasting revenue at a margin you can't operate on. Selling $500,000 at 12% margin instead of $350,000 at 35% is more revenue, more delivery risk, more headcount, more management, and less money.

Two habits fix most of this:

• Record expected margin at proposal stage, not just value. You made an assumption about the delivery cost when you priced it. Write it down.
• Weight the pipeline by expected margin as well as expected revenue. A pipeline that looks strong on one and weak on the other is telling you something urgent about how you're winning work.

Capacity: revenue you can't deliver isn't revenue

The third failure is arithmetic nobody does. If twelve delivery people generate around $150,000 of revenue each per year, the agency can deliver about $1.8m annually — roughly $450,000 a quarter.

Put that next to the pipeline above. If a surprising amount of the $1.2m landed in the same eight weeks, the constraint stops being sales and becomes delivery, immediately. What follows is predictable: freelancers at short notice and poor rates, seniors doing junior work because it's faster than briefing, overtime, quality slipping on the accounts you already had.

Winning more than you can deliver is more expensive than winning less, and the pipeline number actively encourages it.

Concentration: one deal is not a pipeline

$1.2m across four opportunities and $1.2m across twenty behave completely differently. In the first case a single client's budget freeze removes your quarter. Weighted values assume the probabilities are independent, and in a concentrated pipeline they aren't — the same market conditions that kill one deal kill the others.

The tell is simple. If removing your largest opportunity changes the weighted figure by more than about a fifth, you don't have a forecast, you have a bet.

What to track instead

Keep the pipeline. Add three numbers next to it.

Cash coverage  = expected cash receipts (next 13 weeks) / committed costs (next 13 weeks)
Weighted margin = Σ (deal value × win probability × expected margin %)
Capacity load   = weighted revenue in the window / deliverable revenue in the window

Cash coverage below 1.0 means the quarter is short regardless of how the pipeline looks, and it's the earliest reliable warning an agency gets. Capacity load above roughly 0.9 means your next problem is delivery, not sales, and you should be recruiting or subcontracting now rather than in six weeks.

Also track one boring thing: the age of each opportunity. Deals don't usually get rejected, they get old. An opportunity sitting at "proposal sent" for four months is at a far lower win rate than the stage average, and leaving it in at 25% is how a pipeline stays flatteringly large while producing nothing.

Why the number stays wrong

Because nobody is incentivised to correct it. Salespeople are measured on pipeline. Owners find a large pipeline reassuring during a difficult month. Dead deals are removed slowly because removing them feels like admitting something.

The single most useful hygiene rule is a hard age limit — anything with no client-initiated contact in 60 days moves to a separate list and stops counting. It will shrink the number substantially the first time. That shrinkage is information, not a loss.

This week

Take every opportunity in your pipeline and add one column: the week you'd realistically expect the cash to arrive. Not the close date — the date the money lands, including start delays and payment terms.

Then sum only the ones inside thirteen weeks, weighted by probability, and put that next to your committed costs for the same period. It's a twenty-minute job and it answers a different question than the pipeline does — which is the question of whether you can pay everyone.