Real operating evidence, anonymised

Six client situations, shown from problem to outcome

Every figure below comes from our own weekly account reviews. Compare the starting point, work involved, result and remaining limits to decide which approach fits your order.

500+active client accounts
4,000+orders fulfilled daily
6case flows shown below

Three recurring pathways

The useful pattern is the decision flow

Different products create different details, but the operating shape repeats. These are the three paths represented across the six cases.

Growth
TestStabiliseScale

Add stock, branding and automation only after live orders prove the workflow.

Recovery
CrisisTriageRecover

Protect paid orders first, then rebuild the missing records and supplier flow.

Quality
SignalEvidenceCorrect

Repeated defects should change the product or supplier, not create endless replacements.

Six case flows

The work behind each number

Compare what each client needed, the work required and the result. The problems and subsequent changes show which risks to address before choosing a similar approach.

Mobile accessoriesMultiple marketsDropship at scale

From ordinary dropshipper to 400–500 orders a day

1Starting point

This account arrived as a routine dropship operation and grew into one of the largest on our books: two related brands, several hundred orders a day, sold into multiple markets. Growth at that speed stops being a fulfilment question and becomes a supply question. Factory capacity, packing consistency and defect rates that were invisible at fifty orders a day all surface at once.

2What we did
  • Ran both brands through one fulfilment workflow so order issues, invoicing and stock sat in a single place rather than three.
  • Traced a viral design back to the actual manufacturer when the client asked, instead of quoting from a marketplace listing.
  • Held packaging at plain while the product was still proving itself, then ran a 10,000-unit custom printed box order once repeat demand justified the commitment.
  • Moved the failing item onto a redesigned version and stopped shipping the old one.
3What it took

Sustained rather than one-off. A dedicated support agent handled replacement claims on the affected item close to daily for weeks while we worked the supplier side in parallel. Separate threads ran at the same time for weekly invoicing, packaging artwork and chasing factory production, including through a typhoon that pushed the production timeline out.

4What they got
  • 10,000 custom printed boxes in the brand’s own design
  • A redesigned product version to replace the failing one
  • One consolidated weekly invoice covering both brands
  • Replacement and order-issue claims handled on the client’s behalf
5Result

Weekly fulfilment on the account passed $20,000 while the brand moved off generic packing into its own printed boxes, with the problem product replaced rather than patched.

400–500orders per day
$20,000+fulfilled in one week
10,000custom boxes reordered
Reusable lessonBranding should follow proven demand. Quality escalation should follow repeat evidence, not an indefinite queue of individual replacements.
What went wrong

Two failures on our side, not one. Near-daily reports of an item peeling or breaking within about an hour of use were handled as individual replacements for far longer than they should have been before we treated the pattern as a supplier root cause. Separately, when a production backlog built up we prioritised European orders ahead of US ones and did not tell him — he had to ask us why his US customers were waiting.

What we changedThe item moved to the redesigned version and the old one stopped shipping. Any triage decision that changes who waits is now communicated at the time it is made, rather than discovered by the client afterwards.

Fashion accessoriesInternationalMulti-brand

Five sub-brands operating through one account

1Starting point

One operator running five separate storefronts in a single market, with invoice batches swinging between roughly $1,200 and $9,700 depending on the week. The request was to add a completely different product category in a new market without rebuilding the setup or re-onboarding from scratch.

2What we did
  • Consolidated the sub-brands into one account structure with a single point of contact and one invoicing rhythm.
  • Connected the new category through the existing ERP link so it inherited the workflow instead of starting a new one.
  • Corrected incomplete destination addresses before orders reached the dispatch queue rather than after a delivery failed.
  • Sourced material variants as the range matured, so the catalogue could develop without changing supplier.
3What it took

Address quality is the standing daily cost on this account. Destination addresses in that market arrive incomplete often enough that we were making dozens of corrections on busy days. When failed-delivery and redelivery notices outgrew what a chat thread could track, we moved the account onto shared tracking sheets so individual parcels stopped falling through.

4What they got
  • Five storefronts running on one fulfilment workflow
  • A new product line in a new market, live without re-onboarding
  • Same-day invoice-to-picking on the largest batch
  • A shared tracking sheet covering open delivery exceptions
5Result

The largest batch of roughly $9,700 was settled and released to picking the same day it was issued, with the next batch queued immediately behind it, while an entirely new category went live in a second market.

5+sub-brands
$9,700largest invoice batch
Same dayinvoice to picking
Reusable lessonA reusable account structure makes the next storefront cheap to add. Checkout address quality is part of fulfilment design, not a customer-service afterthought.
What went wrong

Quality on two of the sub-brand lines slipped far enough that the client described it as a liability for his business, and that description was fair. We also let the address-failure workload stay a manual daily grind for longer than we should have, absorbing it as effort instead of fixing the process behind it.

What we changedShared tracking sheets replaced chasing individual notices through chat, which visibly reduced what sat unresolved. The quality complaints went back to the supplier as a pattern rather than as separate cases.

Consumer electronicsEuropeForward stocking

Moving a proven product into local European stock

1Starting point

Cross-border transit had become the limiting factor for a product with steady regional repeat demand. Every order was making the full journey individually, so delivery time was capped by the route rather than by anything we could tune in the warehouse.

2What we did
  • Set up a standing inventory deposit so stock could be bought ahead of demand rather than per order.
  • Sent roughly 100-unit batches into a European warehouse, making the final delivery leg domestic.
  • Held the agreed price through a component-cost spike, absorbing a small loss rather than reopening the quote mid-relationship.
  • Refunded excess stock cleanly when a line was scaled back, and sourced accessory products as the range grew.
  • Set up proper monthly invoicing ahead of incoming European e-invoicing requirements.
3What it took

A multi-week build rather than a single decision: agreeing the deposit, sizing the first batches, and then handling the documentation work around it. When one order went wrong badly enough to cost him money, that included producing the paperwork he needed to account for the loss internally.

4What they got
  • A local stock position with a domestic final delivery leg
  • Monthly invoicing compliant with incoming regional requirements
  • Accessory products sourced and added to the range
  • Documentation to support a disputed order
5Result

Delivery time stopped being governed by an international route. Inventory position became the lever instead, and the account moved from per-order dropshipping to a stocked regional model.

~100units per stock batch
€2,560standing deposit
Domesticfinal delivery leg
Reusable lessonOverseas stock is a commitment that suits proven demand. It is not a default starting position, and it only pays back when repeat orders are already there.
What went wrong

Three things went wrong on this account. His agreed logo sticker quietly stopped being applied after the supplying factory changed, and he discovered it himself rather than hearing it from us — his words were that it was simply not normal that he did not know. One order then sat in an early tracking status for over two weeks until his customer opened a payment dispute and was automatically refunded. And a unit shipped that did not match the advertised feature set, which he had to demonstrate to us with video evidence.

What we changedAny factory change now triggers an explicit specification check with the client before anything ships. On the disputed order we produced the documentation he needed rather than leaving him to absorb an unexplained loss, and the feature mismatch went back to the factory with his evidence attached.

FootwearEuropeDemand spike

Preparing for 3,000 orders inside 48 hours

1Starting point

A fixed-date collaboration with a large regional influencer was expected to generate somewhere around 3,000 orders within 24 to 48 hours. The store integration was not finished, the stock position was not settled, and the date could not move.

2What we did
  • Completed the ERP and store connection before the campaign rather than during it.
  • Arranged warehouse coverage across a public holiday so the launch window was staffed.
  • Told the client plainly that we could not reserve factory stock at that volume without a deposit, instead of agreeing to a guarantee we could not honour.
3What it took

Compressed and genuinely pressured. The integration work happened against a fixed external date, and warehouse staff agreed to work through a public holiday to cover the window. The harder work was the conversation: setting a realistic stock expectation while the client was understandably pushing for certainty.

4What they got
  • A working store and ERP integration, finished pre-launch
  • Warehouse coverage across the holiday launch window
  • An honest, written stock position going into the campaign
5Result

The store and the warehouse were ready before demand arrived, instead of the integration gap being discovered inside the surge.

~3,000orders expected
48 hourslaunch window
Pre-launchintegration completed
Reusable lessonA campaign forecast is only useful if fulfilment hears it early enough to change capacity and systems. Tell your partner when the date is set, not when the orders land.
What went wrong

He pushed hard for a guarantee that two to three thousand units would be held ready within 24 to 48 hours, and we could not give it. Factory stock cannot be reserved at that volume without a deposit. Confirm the reservation and capacity plan before setting a campaign delivery expectation.

What we changedWe now open the stock and capacity conversation when a campaign date is first mentioned, rather than when the orders are expected to arrive. The deposit route stays available for anyone who does want volume genuinely reserved.

MigrationEUOnboarding

Taking over 200+ paid but unshipped orders

1Starting point

A seller arrived mid-crisis with more than 200 customer orders already paid for and still unshipped by his previous agent. He had left that agent over pricing that changed from day to day, which meant the commercial relationship had broken down before the operational one was untangled.

2What we did
  • Connected the store through Shopify collaborator access on the same day rather than scheduling an onboarding.
  • Separated genuinely outstanding orders from records that could not be verified.
  • Re-sourced products where usable supplier or tracking evidence did not exist.
  • Put stable, consistent quoting in place so the original reason for leaving did not repeat.
3What it took

The work here was triage rather than sourcing. Establishing what was actually outstanding meant reconciling a partial order export against whatever tracking evidence survived, then deciding order by order whether to chase the original shipment or start again.

4What they got
  • Store access connected and verified on day one
  • A reconciled view of what was genuinely outstanding
  • Re-sourced replacements where records were unrecoverable
  • Consistent quoting the client could plan against
5Result

The backlog moved into a controlled recovery workflow while the customer-refund window was still open, rather than after it had closed.

200+orders inherited
Same daystore connected
Nominimum to migrate
Reusable lessonStart a migration while records and communication still exist. Every silent week with a failing agent removes recovery options that cannot be bought back later.
What went wrong

A new agent inherits the quality of the previous agent’s records, and here that was thin. Missing tracking and supplier data cannot be recreated on demand, so some orders had to be sourced again from scratch — time we could not give back to customers who were already waiting.

What we changedWe now ask for the previous agent’s order export and tracking list on day one of any migration, so the gaps are identified before we commit to a recovery timeline rather than discovered midway through it.

Fashion accessoriesEuropeQuality control

Consolidating after a cluster of returns

1Starting point

We were not winning this account. The seller was actively comparing us against a competing agent on price and on packaging detail, and had told us plainly that he was keeping part of his catalogue with them. That part stayed with the competitor for months.

2What we did
  • Kept quoting and delivering the rest of the catalogue consistently rather than discounting to win the disputed line back.
  • Ran pre-dispatch checks on every order, which caught a supplier mis-pick before it reached his customer.
  • Compared his return pattern against physical product evidence when the other supplier started failing.
3What it took

Unglamorous and continuous. There was no pitch and no campaign to recover the line. The only thing that changed the outcome was checking every parcel before it left, over a long enough period that the difference became visible when the comparison arrived on its own.

4What they got
  • A wrong item intercepted before it reached the end customer
  • Return-pattern evidence he could act on
  • The remaining catalogue moved onto one accountable workflow
5Result

His other supplier began sending items that arrived damaged or unusable, producing a low-tens cluster of returns in one week. He moved the line to us and began consolidating the whole catalogue.

Low tensreturns in one week
A mis-pickstopped before dispatch
Wholecatalogue consolidating
Reusable lessonThe cost of checking should be compared against returns, support load and store reputation, not only against a supplier unit price.
What went wrong

The client initially chose a lower-priced supplier for this line, then returned after repeated quality problems. Compare the inspection scope and the cost of returns alongside the quoted unit price.

What we changedThe existing checks caught a supplier mis-pick before dispatch. For a similar order, agree what will be checked and how a recurring defect will be investigated before expanding the catalogue.

One improvement shared by every account

A client challenged one route price, so we changed the route

This was not a one-account discount. It was an operating improvement applied across the relevant lane.

A competing quote exposed a specific gap. We tested alternative customs-clearance channels and moved the lane onto one with a lower per-code pass-through, recording an early fast-channel delivery in nine days. The duty itself is set by the EU, not by us; what changed is the channel the parcel clears on, and it costs roughly two to three extra working days in transit. The useful response to a precise price challenge is to investigate the route, not defend the old number. We are still working on it: further clearance routes are in testing to bring the per-code cost down again, and on some lanes we are close.

How to use this page

Keep the evidence useful without exposing the client

What we deliberately anonymise

No client, brand, store or contact names appear here. Exact products are reduced to broad categories because the operating decision is useful and the winning product is private.

  • Ranges stay as ranges
  • Rounded figures are rounded down
  • The friction stays in the case

See the product-confidentiality answer →

What would your first controlled test look like?

Send a complete product-and-destination brief and your current order stage. We will tell you the next useful step and return a real quote within one working day.

Operating since 2018 Warehouse in Guangzhou, China Ships to 150+ countries Shopify Partner Quoted in your currency — we carry the FX move