05.25 Sell-Out: Three-Ledger Reconciliation and Six-Disease Root-Cause Diagnosis

Governance metadata | Owner: Head of Sales | Last reviewed: 2026-10-22 | Next review: 2027-01-31

Core principle: reconcile first, diagnose second, and only then prescribe. Selling to a distributor is not the same as selling to a customer; the "false prosperity" of tracking only sell-in explodes all at once in the month of delisting. And once you have confirmed that the product really is not selling, jumping straight to promotion is shooting blind — with the wrong diagnosis, the stronger the medicine, the faster you lose money. This document joins the "three-ledger reconciliation" and the "six-disease root-cause diagnosis" into a single chain. Channel health dashboard: see 05.03; rescue actions: see 05.20; payment terms and rebates: see 05.22.

1. Why the Order Cannot Be Reversed

The two most common mistakes in the industry happen to sit at the two ends of this chain:

Mistake	Symptom	Consequence
Watching only the sell-in ledger	Shipment numbers look great while channel inventory piles up	Concentrated returns and price erosion at delisting, wiping out profit in one go
Skipping diagnosis and going straight to promotion	Discounting the moment sell-through slows	Gross margin is destroyed, the problem is unsolved, and customers learn to wait for discounts

The correct sequence is three steps:

① Reconcile: confirm that "it is not selling" is real, not a bookkeeping mismatch
        ↓
② Diagnose: confirm which disease it is
        ↓
③ Prescribe: targeted remedy (promotion is only one of them)

The "rescue triad" in 05.20 (Scan-down / geofencing / weekend in-store presence) is part of the remedy for step ③. This document supplies steps ① and ② — without them, the triad is shooting blind.

2. The Three Ledgers: Splitting "Sold" Into Three Layers

Ledger	English	Who keeps it	What it answers
Sell-in ledger	Sell-In	Our own shipment system	How much we sold to the channel
Channel-inventory ledger	Channel Inventory	Distributor / retail KA inventory reports	How much the channel is still sitting on
Sell-out ledger	Sell-Out	Store POS / platform back office / end-device scanning	How much customers actually bought

The identity (the basis of reconciliation):

Opening channel inventory + sell-in during the period − closing channel inventory ≈ sell-out during the period

If any side does not reconcile, the difference must be attributed; it must never simply be smoothed away.

Five Categories of Variance Attribution

#	Variance source	Discriminating signal	Handling
1	In transit, not yet received	Logistics documents show the goods in transit and the channel has not signed for them	Extend by the in-transit cycle; do not treat it as slow-moving stock
2	Channel inventory glut	Sell-in is normal, sell-out is falling, and channel inventory keeps rising	Move to step ③ diagnosis; ship new goods cautiously
3	Slow sell-out at the end	Channel inventory is normal but store sell-through is falling	Move to step ③ diagnosis
4	Distorted data	Channel-reported numbers clearly disagree with POS / platform back office	Fix the data definition first, then talk about sell-out
5	Returns written back but not deducted	The sell-out ledger does not deduct returns, so it is inflated	Recalculate on the returns basis (see 06.03)

Three-Ledger Reconciliation Table (Template)

The value columns are measured data and must be updated weekly; never enter estimates.

Week	Opening channel inventory	Sell-in this period	Closing channel inventory	Implied sell-out	Actual sell-out	Variance	Variance attribution	Action
W1
W2
W3
W4

Rules of use:

Variance > 5% for 2 consecutive weeks with no attribution → trigger a data audit;
Channel inventory turnover (weeks) rising for 3 consecutive weeks while sell-out falls → move to step ③;
Any ledger missing → do not start a promotion decision (with incomplete data, promotion is gambling).

3. Four Forms of False Sell-Through

"It sold" can also be fake. Identifying false sell-through is the core value of the three ledgers.

Form	Mechanism	Identification signal	Harm
Channel-stuffing form	Policy or rebate incentives drive the channel to hoard	Sell-in rises, channel inventory rises, end sales stay flat	Concentrated returns at delisting
Fabricated-data form	The channel inflates sell-out to qualify for rebates	Reported numbers disagree with POS	Decisions built on fake numbers
Promotion pull-forward	A mega-promo consumes future demand early	A cliff 2~4 weeks after the mega-promo	Misdiagnosed as "the product does not work"
Return write-back	Sell-out does not deduct returns	Sell-out is high while the return rate rises in step	Profit is eaten away

Rebate design is the main driver of channel stuffing. If rebates hook only onto purchase volume, the channel will inevitably stuff inventory. Rebates should hook onto sell-out or end-retail merchandising execution, see 05.22 Rebate Design.

4. Six-Disease Root-Cause Diagnosis

Once you have confirmed that the product really is not selling, rule them out one by one using the table below. The six diseases overlap heavily and may occur together, so they must be worked through in order.

#	Disease	Discriminating signal	Often misdiagnosed as	Targeted remedy
1	Wrong price anchor	In-store conversion is low, but shoppers leave after asking the price; comparable competitors are cheaper	"Not enough brand power"	Adjust the price structure or supply value evidence, not a discount
2	Missing display and demonstration	Stock is there but there is no display and no demo; staff cannot present the product	"The product does not work"	Fix display and demonstration, see 05.20
3	Messaging-audience mismatch	The in-store audience does not match the target customer; staff present the wrong selling points	"Not enough ad spend"	Fix the selling-point order (04.01) + staff training (05.16)
4	Channel-stuffed false sell-through	The three ledgers show high channel inventory and flat end sales	"The market has no demand"	Stop shipments + clear channel inventory, see 05.22
5	Inventory mismatch	Best-selling SKUs are out of stock while slow-moving SKUs pile up	"Nothing sells overall"	Break down by SKU, see 03.05 PSI
6	The product itself	Sell-out is still flat after ruling out 1~5; return reasons cluster and relate to the product	—	Go to 06.02 VOC and product iteration

Diagnosis Order (No Step May Be Skipped)

① Do the three ledgers reconcile?
   No → Fix the data first (go back to Section 2); do not diagnose
   Yes ↓
② Is it false sell-through? (stuffing / data / pull-forward / write-back)
   Yes → Handle the false sell-through; do not promote
   No ↓
③ Break it down by SKU: are all SKUs static, or only some of them?
   Some → Inventory mismatch (disease 5)
   All ↓
④ Is in-store conversion low?
   Yes → Price anchor (disease 1) or messaging mismatch (disease 3)
   No ↓
⑤ Is there display / demonstration? Can staff present it?
   No → Missing display and demonstration (disease 2)
   Yes ↓
⑥ Rule out the above → the product itself (disease 6)

5. Targeted Remedy Mapping

Disease	Remedy	Prohibited action
1 Price anchor	Adjust the price structure / supply value evidence / change the SKU mix	Across-the-board price cuts
2 Display and demo	Display rebuild, in-store demonstration, staff incentives	Pouring more money into ads (customers still cannot find the product)
3 Messaging mismatch	Reorder the selling points, staff training, switch reach channels	Adding budget to target the same people
4 Channel-stuffed false sell-through	Stop shipments, clear inventory, tie rebates to sell-out	Promotion (the more you promote, the more you stuff)
5 Inventory mismatch	SKU-level replenishment / clearance, PSI rescheduling	Blanket promotion
6 The product itself	VOC attribution, iterate or exit the market	Using marketing to mask a product problem

The only situations in which promotion is allowed: the diagnosis confirms "price anchor" and you have decided to give up part of the margin, or a targeted clearance of slow-moving SKUs within "inventory mismatch". Promotion is a means, not the default answer.

6. Dashboard and Thresholds

Metric	Definition	Warning	Circuit breaker
Three-ledger variance rate	\|actual sell-out − implied sell-out\| ÷ implied sell-out	> 5% for 2 consecutive weeks	> 10%
Channel inventory turnover	Closing inventory ÷ average weekly sell-out	Rising for 3 weeks	> 8 weeks
Sell-out attainment rate	Stores meeting target ÷ total registered stores	< 70%	< 50%
Return rate	Return volume ÷ sell-out volume	+3pt month over month	+5pt month over month
False sell-through share	(stuffing + data + pull-forward + write-back) ÷ total sell-out	Unattributable share > 5%	—

Thresholds must be calibrated per category before they are written in; the defaults in this document must not be carried over directly as an industry benchmark.

7. Common Mistakes

Mistake	Consequence	Correction
Treating Sell-In as sell-out	False prosperity that explodes at delisting	The three ledgers
Diagnosing before the books reconcile	The diagnosis rests on fake data	Reconcile first
Promoting as soon as sell-out looks bad	Gross margin is destroyed and the remedy is off-target	Six-disease diagnosis
Mistaking false sell-through for weak demand	The more you promote, the more the channel stuffs	Identify the four forms
Looking at sell-out in aggregate without breaking out SKUs	Hides inventory mismatch	SKU-level breakdown
Tying rebates only to purchase volume	The channel will inevitably stuff inventory	Tie to sell-out / merchandising execution
Running promotion with no cooling-off observation period	Pull-forward is misjudged as failure	Observe for 2~4 weeks

Execution Checklist

[ ] The three ledgers are live: sell-in / channel inventory / sell-out, updated weekly
[ ] The identity is aligned: opening + sell-in − closing ≈ sell-out, with every variance attributed
[ ] All five variance categories are used: in transit / glut / slow-moving / distortion / return write-back
[ ] The four forms of false sell-through are ruled out: stuffing / data / pull-forward / write-back
[ ] The rebate linkage is checked: whether it hooks only onto purchase volume (the driver of stuffing)
[ ] The six diseases are worked through in order: no step skipped, no concurrent case misjudged
[ ] The SKU-level breakdown is complete: separating broad slow-moving from inventory mismatch
[ ] The remedy matches the disease: promotion is not used to solve non-price problems
[ ] Thresholds are calibrated per category: defaults are not carried over directly
[ ] A cooling-off observation period is set after promotion: 2~4 weeks, separating pull-forward from genuine failure

05.25 Sell-Out: Three-Ledger Reconciliation and Six-Disease Root-Cause Diagnosis ​

1. Why the Order Cannot Be Reversed ​

2. The Three Ledgers: Splitting "Sold" Into Three Layers ​

Five Categories of Variance Attribution ​

Three-Ledger Reconciliation Table (Template) ​

3. Four Forms of False Sell-Through ​

4. Six-Disease Root-Cause Diagnosis ​

Diagnosis Order (No Step May Be Skipped) ​

5. Targeted Remedy Mapping ​

6. Dashboard and Thresholds ​

7. Common Mistakes ​

Execution Checklist ​

Related Documents ​