Case Study: From Stock Chaos to a Measured Routine
Theory guides give frames; a case guide shows the frame working in the field. This is the story of a three-channel business moving from stock chaos to a measured routine in Turkey — in four acts, with the mistakes avoided between them. For a foreign reader it doubles as something else: a preview of what the same journey looks like when the owner watches it from another country.
A confidentiality note upfront: the business name and identifying details are changed; the numbers are real in direction and proportion. The aim is not a trophy cabinet but a road story with extractable lessons; the mistakes are written as openly as the wins.
The stage opens like this: a home-textile seller turning on three channels — its own store and two Turkish marketplaces. Revenue growing, the team tiring, the store score falling. The apparent culprit was “busyness”; the real culprit, as will be seen, was the setup.
The story’s principle is the same one running through this whole cluster: what goes wrong is not the tools but the pairings and the order. The case is that sentence, staged.
Act One: The Photograph of the Chaos
BU BÖLÜMÜN ÖZETİ
- The difference between complaint and diagnosis
- What the numbers said
- The root cause: an islands setup
- Between acts: the first mistake avoided
The table arriving at ours was a familiar one; the diagnosis came through a three-week photograph.
The difference between complaint and diagnosis
The business arrived saying “we need an integrator — which one should we buy”: a tool question. Our first job was rewinding the question: photograph first, tool after. Three weeks of data gathering surfaced the real pattern beneath the complaint — a rewind that works identically for a remote owner, since the photograph is made of numbers, not visits.
What the numbers said
The photograph was sharp: one full working day a week of manual stock-and-order carrying, double-sale cancellations repeating monthly, the store score falling on both marketplaces, and channel profitability never once separated. That the two best-selling products were losing money after commission was seen, for the first time, in this photograph — while revenue broke records.
The root cause: an islands setup
The store on one system, the marketplaces in their own panels, bookkeeping in a separate program; the only bridge between them, an employee copy-pasting. The pattern has a name in ecosystem language: the islands map. Not a missing tool but missing connections; the integrator question was right, its place in the order wrong.
Between acts: the first mistake avoided
The “buy the integrator now” pressure was strong in week one; it was resisted. Integration built on dirty data would have automated the chaos: barcodeless products, duplicate records, uncounted stock. The three-week photograph doubled as the cleaning list.
Act Two: Building the Order
BU BÖLÜMÜN ÖZETİ
- The unit-economics table first
- Then data cleaning and code discipline
- Tool selection: by the needs list
- Between acts: the second mistake avoided
After diagnosis the work was sequencing, not shopping.
The unit-economics table first
A decision table was set for the two loss-making star products: a price correction for one, a packaging-and-cargo cost revision for the other. This table was set before the integration — because accelerating loss-making sales is automation’s bitterest talent.
Then data cleaning and code discipline
A two-week cleaning period: duplicate records merged, missing barcodes completed, the variant structure unified, and one truth established with a full stock count. A boring fortnight; the confession came later: “the real work, it turns out, was these two weeks.”
Tool selection: by the needs list
Only at this point was the tool table set: three channel links, an accounting bridge and order automation wrote the needs list; candidates were eliminated against it, and the short list was tried on the business’s hardest scenario — a variant-heavy product plus a campaign rehearsal. The choice was made on a hard day, not in a brochure.
Between acts: the second mistake avoided
One short-list candidate charmed with a long feature list; the needs list held no match for it. “Maybe we’ll need it later” was spoken at the table and weighed on the scales: a certain cost today against an uncertain benefit tomorrow. The candidate fell; feature inflation was given no passage.
Act Three: The Build and the First Season
BU BÖLÜMÜN ÖZETİ
- The pilot and the edge cases
- Staged rollout and the daily check
- The season exam
- Between acts: the third mistake avoided
The build ran the standard five-stage road; the calendar sat on the season countdown — finishing a quarter before Turkey’s November peak.
The pilot and the edge cases
A pilot of the twenty most representative products: publishing, stock deduction, end-to-end orders and the edge cases — return stock write-back, cancellation-invoice flow, stock-zero — tried one by one. A rule error caught in the return write-back was fixed in the pilot, not in production; ten minutes of testing erased a probable season accident.
Staged rollout and the daily check
The catalogue opened in three waves; a daily five-minute exception check ran through the first month. Mapping errors caught in the early weeks were fixed while small; that is staged rollout’s only job, and it did it.
The season exam
The build had finished a quarter before the campaign season; the season was the first real stress test. The result: zero double sales, manual carrying down to minutes a day, the order-invoice-cargo chain turning without manual touch. The team said it had passed a season, for the first time, “watching instead of fearing” — and the owner’s report of it needed no translation: the card said the same thing in every language.
Between acts: the third mistake avoided
Mid-season, an appetite arose for adding a new channel; it was postponed. A structural change in peak weeks would have gambled the working routine; the channel opened after the season, on a calm calendar. Engines are not dismantled in season; this case did not break the rule.
Act Four: The Report Card’s Numbers
BU BÖLÜMÜN ÖZETİ
- The unexpected side gain
- The honesty row: what did not go smoothly
- Three lessons from this story
The switch’s return was read by the card, not by feel; the table summarises the first quarter.
| Gauge | Before | End of first quarter |
|---|---|---|
| Manual carrying hours | One full day a week | Minutes a day |
| Double sales / stock accidents | Repeating monthly | Zero |
| Store score direction | Falling on both channels | Recovering on both |
| Channel profitability visibility | None | Per channel on the monthly card |
| Loss-making star products | Two, invisible | Zero; price and cost corrected |
The unexpected side gain
With the accounting bridge built, month-end invoice labour melted too; a gain outside the plan’s targets, the connection principle’s gift. One line built right in a system improves its neighbours.
The honesty row: what did not go smoothly
Not everything shone: one channel’s category mapping took two revision rounds, one team member’s panel habits settled slower than hoped, and once in the first month a notification setting delayed the exception queue’s discovery. This row, missing from most case studies, is the row of reality.
Three lessons from this story
One: the tool question gets rewound; the photograph comes first. Two: cleaning and sequencing are more than half the build. Three: the card is the switch’s insurance. Each was told separately across this cluster’s guides; the case is proof of all three working in one story.
The Story in One Visual
The visual gathers the four acts and the direction.
The acts’ durations teach too
The photograph three weeks, sequencing and cleaning two, the build and pilot a few more: the calendar’s long item was decision and cleaning labour, not technique. The weeks a rush would have saved are a rounding error beside the season a rush would have lost.
If you want to start with the photograph
This case’s first act opens through the same door for you — from any country: a digital audit is the three-week photograph, accelerated and run remotely; our table’s order sits on the e-commerce consultancy page. In the first meeting, your numbers are discussed — not tool names.
Frequently Asked Questions
Sık Sorulan Sorular
To the team, the hours row; to the owner, the profit row; to us, the card row itself — because it made every other row visible, and for an owner reading from abroad it was the operation’s only honest window. Improvement that cannot be measured cannot be repeated; the card turned a one-time win into a standing routine.
If manual hours grow, accidents repeat and profit stays invisible, the photograph stage is waiting for you; the full signal list sits in the threshold guide, the order’s logic in the sequence guide, the build’s road in the setup guide.
A three-channel home-textile business with growing manual hours, repeating double sales, falling store scores and invisible channel profitability; the arriving question was “which integrator should we buy.”
Because integration built on dirty data automates the chaos; first came a three-week photograph, surfacing the root cause (an islands setup) and the cleaning list.
The unit-economics table (two loss-making star products corrected), two weeks of data cleaning, code discipline and a full stock count; the tool was chosen only after, by the needs list.
Elimination by the needs list, then a trial on the hardest scenario (a variant-heavy product plus a campaign rehearsal). A charming long-feature candidate fell for lacking a needs match.
The standard five-stage road: a twenty-product pilot, edge-case tests (returns, cancellations, stock-zero), a three-wave staged rollout and a daily exception check through month one.
In the first quarter double sales hit zero, manual hours fell from a day a week to minutes a day, store scores recovered and channel profitability appeared on the card for the first time.
No: one channel’s category mapping took two revision rounds, panel habits settled slowly for one person, and a notification setting once delayed the exception queue. Honesty is part of the case.
Rewinding the tool question and putting the photograph first; second, correcting the loss-making products before automating. The order was worth more than the tool.
Your own photograph: manual hours, accident frequency, channel profitability. If the threshold signals burn, the photograph stage waits; a remote audit opens that door.