August 2026 · the phase NOC is actually in

One placement made. The month is the metric.

Seven clients on the books, fourteen role folders open, forty scored candidates, and a delivery side of one person. The question this month is not what NOC is worth. It is whether €135k can be contracted by 4 September without selling more than the delivery side can finish.

Running without revenue booked in 2026, cash in the BV, monthly tool spend. Anything looking backwards on this page is dark until those are filled in on variables.

The month

4 August to 4 September 2026. Contracted, not collected, unless block A says otherwise.

Contracted against target

€0 / €135k

0% of the number, day 4 of 31

Even pace would be €17k by today. The tick is par.

Gap left to close

€135k

€5.000 a day for the 27 days left

Weighted plan against target

€74k

Short by €61k before anyone makes a call

Pipeline coverage

40%

€409k of qualified pipeline needed at a 33% close rate

Placements the target implies

at €7.500 average per role

Or packaged deals

at €45k for nine roles

A warm route converts

4,0×

better than a cold one. Thirty-day windows are won on warm lists.

Can it be delivered

The ceiling is screening hours, not demand. When the target implies more roles than the delivery side can finish, selling successfully becomes a delivery failure.

Roles the target implies

inside the sprint window

Roles the window can finish

3,5

7,7 roles held at once, 6 week cycle

Blocked

The target needs 18 roles and the delivery side can finish 3,5 in the same window. Selling per hire cannot reach the number even if every conversation converts. Packaged deals with staged start dates land the cash inside the month and spread the delivery past it, which is the only shape that resolves this.

Capacity ceiling, a full year

€303k

40 closed roles at 70% fill

People who can run a screen

Named as a live risk in both the deck and the standup

Delivery hours per role

23,4 human hours after the stack absorbs 10%

Roles held at once

7,7

derived from hours, not guessed. 30 delivery hours a week per operator.

What a role is worth

Everything downstream is a multiple of this. The previous version of this model assumed €15.000 a role, which is roughly double anything NOC has ever charged, and every number built on it was wrong by the same factor.

Average fee per role

€7.500

0% of roles sold inside a package

Contribution per role

€5.820

78% margin with founder time costed at market

Margin once the stack carries delivery

98%

€120 of data and inference per role is the only genuinely marginal cost

Per role inside the nine-role package

€5.000

against €7.500 per hire. One sales cycle instead of nine.

Fee per hire

The wedge. Low risk, gets NOC embedded.

€7.500

Three roles

€20.000 · €6.667 each

Six roles

€38.000 · €6.333 each

Nine roles, the Q3 package

A third off, in exchange for one sales cycle and payment at signature.

€45.000 · €5.000 each

Percentage model, where used

14% of €95.000. The ivee deal runs on this shape, and it prices €5.800 above the flat fee.

€13.300

Sourcing only

Depth of vetting and handover still undecided, so it does not feed revenue here.

€1.500

Where the fee goes

Delivery hours€1.170 16%

Intake and brief rewrite€390 5%

Data and inference€120 2%

Acquisition, allocated€173 2%

Left over€5.647 75%

Founder time is costed at market rate even though no founder is paid. Leaving it out is what makes a services business feel like software.

The proof, and the instruments that would show it

Fit is not measured by whether clients arrive unprompted. That measures direction of arrival, which any advertising budget manufactures. These five signals were chosen because spend cannot fake them.

Fit score

25%

weighted across five signals that money cannot buy

Instrument coverage

4 of 5

signals currently measurable at all

The five signals

Repeat purchase without a new sales cycle30% wt33%

1 of 3 clients repeating. One client buying 9 times is real demand, not fit.

Unprompted referral converted25% wt0%

2 offered, none worked. The signal spend cannot buy, and nothing produces it.

Paid before seeing the work15% wt100%

100% pay upfront. Revealed preference, and it clears.

Twelve-month placement retention20% wt0%

no instrument

0 of 12 placements old enough to measure. No instrument yet.

Sells and delivers without a founder10% wt0%

0%. Both founders are still load-bearing on both sides.

Inbound leads are tracked on the variables page and given zero weight here on purpose. NOC also has a structural reason the arrival test fails that has nothing to do with quality: the need is episodic. A client who loved the work does not come back until they hire again.

Placements made

one card carries status hired in candidates.ts

North star, twelve-month retention

0 / 12

placements old enough to measure, against the count needed before the number can be stated

Referrals offered, converted

2 / 0

the cheapest unbuilt signal in the company

Counted work behind all of it

scored candidates across 12 role processes, 25 screens on record, 7 clients

Background

True, and not decision-relevant this month. Kept live so it stays honest, kept quiet so it does not compete with the four bands above.

Cash burn a month

Tool spend plus founder cash. Currently unknown, so it reads as zero.

€0

Economic burn a month

€312k a year of founder opportunity cost that never appears on a statement

€26.000

Cash runway

Unbounded while roles close and nothing is being spent. The real clock is founder patience.

unbounded

Average revenue a month

Dark until revenue booked in 2026 is filled in.

€0

Acquisition cost per logo

€780 of selling time plus tooling

€780

Lifetime value per client

4,5 roles across 1,8 years, gross profit

€26k

Value against acquisition cost

Meaningless while spend is unknown and capacity is the constraint.

33,6×

Logo churn a year

Structural. The customer leaves when the need is filled, not when the service disappoints.

56%

Category market

21.204 addressable roles a year at €7.500

€159,03M

Serviceable

28% of the category survives the geography, role family and no-in-house-talent filters

€44,61M

Capacity ceiling as a share of it

Demand is not what caps this business.

0,68%

Three-year path against the €40,00M target

Year	Operators	Agent share	Packaged	Closed roles	Revenue	Recurring	Multiple	Value
2026	1	10%	20%	40	€283k	0%	1,1×	€311k
2027	2	35%	50%	112	€795k	12%	3,3×	€2,59M
2028	3	55%	70%	242	€1,63M	15%	3,7×	€6,09M

On this path year three lands at €6,09M, short of the target by €33,91M. The target needs €2,11M of revenue if NOC is read as vertical AI, and €36,36M if it is read as an agency. The share of revenue that recurs is what decides which.

Change an assumption Run it as a scenario