Build-vs-Distribute Split
An hour is distribution only if what you made reached somebody. A draft is not distribution. An unpublished video is not distribution. This logs the week honestly, bands it twice, and reads the direction across periods rather than defending any single one.
In the worked example, week two declares 31% distribution against a 30% floor — it clears — and qualifies at 19%, because four of those ten hours produced drafts nobody has seen. Over eight weeks the qualified share moves from a 24% mean to a 10% mean. That is not a bad week. That is a direction.
instant download · .xlsx · yours to keep
deterministic · offline · grades your own hours, not a person
The problem
You cannot spend the same hour twice, and you have been.
One week in the sample clears a 30% floor on the hours it called distribution and lands at 19% on the hours that reached anyone. The verdict takes the worse one.
A 24% mean qualified share in the earlier half, 10% in the later half. No single week looks like a crisis. The direction does.
The floor is yours and the movement threshold is yours. If eight straight weeks read starved against a floor you set, either the floor was wrong or the weeks were.
Every founder who has ever said “I need to do more marketing” believes they already do some. Most of what gets counted is production: writing the post, recording the video, building the page. That work is necessary and it is not distribution, and the gap between what you call distribution and what actually reached a person is usually where a whole quarter goes. Nothing here judges how much you got done. It only asks where the hours went, and whether that is moving.
See it work
Move four hours out of drafts and watch the band change.
W-02 is selected. It declares 31% distribution — over the 30% floor — and qualifies at 19%, because four of those ten hours produced drafts nobody has seen. The verdict takes the worse one.
| Period | Declared | Qualified | Leak | Verdict |
|---|---|---|---|---|
| W-01 Week 1 — launch week | 33% | 30% | −3 | BALANCED |
| W-02 Week 2 — two posts drafted | 31% | 19% | −12 | BUILD-HEAVY▼ |
| W-03 Week 3 — back on the roadmap | 25% | 22% | −3 | BUILD-HEAVY |
| W-04 Week 4 — bundle build | 26% | 23% | −3 | BUILD-HEAVY |
| W-05 Week 5 — deep in the engine | 17% | 14% | −3 | DISTRIBUTION-STARVED▼ |
| W-06 Week 6 — shipping features | 14% | 11% | −3 | DISTRIBUTION-STARVED |
| W-07 Week 7 — more features | 11% | 11% | — | DISTRIBUTION-STARVED |
| W-08 Week 8 — assets made, never sent | 13% | 5% | −8 | DISTRIBUTION-STARVED |
W-02 — Week 2 — two posts drafted
On the hours you declared
31%
BALANCED
On the hours that reached anyone
19%
BUILD-HEAVY
RECLASSIFICATION — the hours you called distribution put this week one band better than the hours that actually reached someone. The verdict takes the worse one.
Week verdict
BUILD-HEAVY
32.0 productive hours · admin excluded · leakage 12 points
Weeks
1 balanced · 3 build-heavy · 4 starved
reclassified in 2 of them
Hours that never left
12.0
called distribution, reached nobody
Movement
-14
24% → 10%
Trend
SLIDING BACK TO BUILDING
Direction, not level. Any single week can be defended.
It grades your own recorded allocation of hours — not a person, not anyone else's time, and not how much you got done. Your floor, your threshold, no benchmark. Not financial advice.
W-02 is selected. Raise its reached someone hours from 6 to 10 — the same total, just actually sent — and the week goes from BUILD-HEAVY to BALANCED without a single extra hour worked. Then drop your floor from 30 to 20 and watch how many weeks change their mind, which is the honest way to find out whether the floor or the weeks were the problem.
The model
One counting rule, two bandings, and a direction.
An hour is distribution only if the output of it reached a person outside your company during that period. Admin hours are excluded from the split entirely — invoicing is neither building nor distributing. Bands: BALANCED at or above your floor, BUILD-HEAVY at or above half your floor, DISTRIBUTION-STARVED below it.
Every week is banded twice — on declared hours and on qualified hours — and the verdict is the worse of the two. The gap between the two shares, in percentage points, is the leakage. The verifier proves leakage alone can move a week two full bands, from BALANCED on what you declared to DISTRIBUTION-STARVED on what reached anyone.
The rollup compares the mean qualified share of the earlier half of your periods against the later half, against a movement threshold you set. With an odd count the middle period is ignored, because assigning it to either half would let the answer depend on which half you chose. Fewer than four periods reads TOO FEW PERIODS TO READ A TREND.
There is no benchmark anywhere in this kit, and that is not a gap. If eight straight weeks read DISTRIBUTION-STARVED against a floor you set yourself, either the floor was wrong or the weeks were — and only one of those is comfortable. A number imported from someone else’s business would let you argue with the number instead of with the weeks.
Verified, not asserted
The workbook is checked against a reference engine.
You get the .xlsx — that is the product. It is built against a separate reference implementation of the same model, and every formula in the sheet is asserted equal to it before release: 100 cell-level checks on the worked example, 15,453 share pairs, a 270-row hours grid, and 1,053 ordered trend sequences. One of those checks rebuilds the whole workbook with only three periods, because a formula that divides by half the row count is exactly the one that explodes on a short sheet. Below is that engine’s verbatim output on the same eight weeks the workbook ships with, captured programmatically and diffed against this page.
==============================================================================
BUILD-VS-DISTRIBUTE SPLIT - BDS-039 - RedHub AI LLC
==============================================================================
Hours count as distribution only if the output of them reached someone
outside your company that period. Your floor, your threshold, your hours.
It grades your own recorded allocation - not a person, and not anyone else.
Your distribution floor 30%
Half your floor (the lower band) 15%
Movement threshold for a trend 5 points
------------------------------------------------------------------------------
PERIOD W-01 [Week 1 - launch week]
------------------------------------------------------------------------------
Build 20.0h declared distribution 10.0h of which reached someone 9.0h
Admin 4.0h (excluded from the split) productive 30.0h
Share you declared 33% -> BALANCED
Share that reached anyone 30% -> BALANCED
Leakage 3 points of distribution time
that never left the building
VERDICT BALANCED
------------------------------------------------------------------------------
PERIOD W-02 [Week 2 - two posts drafted]
------------------------------------------------------------------------------
Build 22.0h declared distribution 10.0h of which reached someone 6.0h
Admin 3.0h (excluded from the split) productive 32.0h
Share you declared 31% -> BALANCED
Share that reached anyone 19% -> BUILD-HEAVY
Leakage 12 points of distribution time
that never left the building
RECLASSIFICATION GATE FIRED the hours you called distribution
put this period one band better than
the hours that actually reached anyone
VERDICT BUILD-HEAVY
------------------------------------------------------------------------------
PERIOD W-03 [Week 3 - back on the roadmap]
------------------------------------------------------------------------------
Build 24.0h declared distribution 8.0h of which reached someone 7.0h
Admin 4.0h (excluded from the split) productive 32.0h
Share you declared 25% -> BUILD-HEAVY
Share that reached anyone 22% -> BUILD-HEAVY
Leakage 3 points of distribution time
that never left the building
VERDICT BUILD-HEAVY
------------------------------------------------------------------------------
PERIOD W-04 [Week 4 - bundle build]
------------------------------------------------------------------------------
Build 26.0h declared distribution 9.0h of which reached someone 8.0h
Admin 2.0h (excluded from the split) productive 35.0h
Share you declared 26% -> BUILD-HEAVY
Share that reached anyone 23% -> BUILD-HEAVY
Leakage 3 points of distribution time
that never left the building
VERDICT BUILD-HEAVY
------------------------------------------------------------------------------
PERIOD W-05 [Week 5 - deep in the engine]
------------------------------------------------------------------------------
Build 30.0h declared distribution 6.0h of which reached someone 5.0h
Admin 3.0h (excluded from the split) productive 36.0h
Share you declared 17% -> BUILD-HEAVY
Share that reached anyone 14% -> DISTRIBUTION-STARVED
Leakage 3 points of distribution time
that never left the building
RECLASSIFICATION GATE FIRED the hours you called distribution
put this period one band better than
the hours that actually reached anyone
VERDICT DISTRIBUTION-STARVED
------------------------------------------------------------------------------
PERIOD W-06 [Week 6 - shipping features]
------------------------------------------------------------------------------
Build 32.0h declared distribution 5.0h of which reached someone 4.0h
Admin 3.0h (excluded from the split) productive 37.0h
Share you declared 14% -> DISTRIBUTION-STARVED
Share that reached anyone 11% -> DISTRIBUTION-STARVED
Leakage 3 points of distribution time
that never left the building
VERDICT DISTRIBUTION-STARVED
------------------------------------------------------------------------------
PERIOD W-07 [Week 7 - more features]
------------------------------------------------------------------------------
Build 34.0h declared distribution 4.0h of which reached someone 4.0h
Admin 2.0h (excluded from the split) productive 38.0h
Share you declared 11% -> DISTRIBUTION-STARVED
Share that reached anyone 11% -> DISTRIBUTION-STARVED
VERDICT DISTRIBUTION-STARVED
------------------------------------------------------------------------------
PERIOD W-08 [Week 8 - assets made never sent]
------------------------------------------------------------------------------
Build 33.0h declared distribution 5.0h of which reached someone 2.0h
Admin 3.0h (excluded from the split) productive 38.0h
Share you declared 13% -> DISTRIBUTION-STARVED
Share that reached anyone 5% -> DISTRIBUTION-STARVED
Leakage 8 points of distribution time
that never left the building
VERDICT DISTRIBUTION-STARVED
==============================================================================
THE TREND
==============================================================================
Periods logged 8
Balanced / build-heavy / starved 1 / 3 / 4
Periods where the gate fired 2
Hours called distribution that
never reached anyone 12.0
Earlier half, mean qualified share 24%
Later half, mean qualified share 10%
Movement -14 points
TREND SLIDING BACK TO BUILDING
The rollup reads direction, not level. A single period is noise; the
question is whether the share is moving, and in which direction.
==============================================================================What it is — and isn't
It grades your own hours. It is not a productivity score.
- A log of where your own hours went, with one strict counting rule.
- Deterministic and offline — same hours, same verdicts, every time.
- A named model you can argue with: the counting rule, both bands and the trend rule are all on this page.
- Dateless. Add a row a week; the trend gets more useful the more rows it has.
- Not a productivity score. It has no opinion on how much you got done.
- Not for evaluating anyone else’s time. It grades your own recorded allocation.
- Not connected to anything — no time tracker, no calendar. You log four numbers a week.
- Not financial advice, and it scores no people.
Scope note: this grades an allocation of hours you recorded yourself, against a floor you set yourself. It is not a measure of output, effort or performance, and it is not built for managing other people’s time. It is not a timekeeping, payroll or wage-and-hour record. No verdict guarantees any audience, traffic or revenue outcome. This is not financial advice.
Pairs well with
The Split says you need the hours. These three are where they go.
The recovered hours have to go somewhere. This is an operating system for spending them on one channel rather than five.
ViewThe other honest move is to make a unit cost fewer hours. This is the drafting system, with a drift guard so volume does not cost you voice.
ViewThe cheapest distribution hours you will ever spend are the ones your existing customers spend for you. Worth building before you buy more of your own.
ViewCommon questions
The mechanics, stated plainly.
One spreadsheet
Log the week honestly.
- Three tabs, 26 rows — half a year of weeks — live formulas, a worked eight-week example.
- A row-validity check, so qualified hours cannot quietly exceed declared ones.
- Opens in Excel, Google Sheets or Numbers. Nothing to install.
- Deterministic and offline. It grades your own hours, not a person.
Sold by RedHub AI LLC · Secured by Stripe · redhub.ai