Retail Sourcer Bag Buying: How One Knit Range Actually Moves From Plan to Floor in Twelve Weeks

The Brief: Five Styles, Three Colourways, One Floor Date

A retail sourcer buying knitted bags is not buying a product; the buyer is buying a date, and every other decision bends around it. In the case followed here the range plan called for five styles in three colourways with a floor date twelve weeks out.

That single sentence carries the whole programme, and most of its weight sits in the numbers rather than in the design intent. Five styles in three colourways is fifteen style-and-colourway pairs, and at MOQ 100 pcs per pair that is 1,500 pieces against a plan conceived as five products. Twelve weeks sounds generous until it is split: the bulk production block takes 30 days on its own, inspection and freight take most of another month, and accounting for the range of delays is prudent.

The sourcer's first move was not to ask for a lower unit figure. It was to check whether fifteen pairs could be edited down before they became fifteen commitments, because that is the only variable that changes both the exposure and the calendar. Everything that follows in this account descends from that one decision being made in week one rather than week six.

Week One: Cutting the Grid Before It Becomes Commitment

The grid was cut from fifteen pairs to eleven by dropping one colourway from each of four styles, which took 400 pieces out of the programme without removing a single line from the planogram.

This is the move that separates a retail sourcer from a product developer, and it is available precisely because MOQ attaches to the pair rather than to the style. Two of the five styles were clearly going to carry eighty percent of the floor; giving those two their full three colourways protected the range. The other three were supporting acts, and two colourways each was materially more than a fourteen-week display warranted. The piece count fell from 1,500 to 1,100 and the programme exposure fell with it.

The secondary benefit showed up later in the calendar. Fewer pairs means fewer machine set-ups threaded through the same 30-day bulk window, and fewer inspection lots to release. In practice the reduced grid gave the sourcer room to run a correction round inside the existing schedule instead of choosing between shipping late and shipping uncorrected.

Where a Retail Sourcer Actually Spends the Week

The time goes to specifications, documentation and packaging rather than to negotiation, because the retail end of a programme is governed by what has to be present in the carton rather than by what it cost.

A useful diagnostic for any sourcer is to log a week honestly. In this case the largest block went to writing the technical pack: GSM target, gauge, yarn composition, hardware finish, colour references and tolerances. The second went to compliance documentation, which for a multi-market retailer means pulling BSCI or Sedex SMETA audit evidence, REACH declarations for the European Union, California Prop 65 statements and, for anything aimed at children, CPSIA evidence. The third went to packaging decisions, which turn out to be surprisingly confining once barcode placement and bag-warning requirements are taken into account.

Negotiation came last and took the least time, which is not a moral point about diligence but an arithmetic one. When the specification is complete, the number quoted is the number to compare. When it is incomplete, the two parties are pricing different bags, and every hour spent sharpening a figure that describes the wrong product is wasted twice.

The Twelve-Week Critical Path, Station by Station

Twelve weeks splits into roughly one week of brief and specification, 7 working days of sampling, three weeks of approval and correction, 30 days of bulk, then inspection and freight across the remaining weeks.

The two fixed blocks are the ones worth memorising, because they do not move for anybody. Sampling takes 7 working days and bulk takes 30 days on this site regardless of whether eleven pairs or fifteen were ordered, and regardless of how important the retailer considers itself. Everything else in the twelve weeks is compressible: the brief can be sharpened, approvals can be batched, freight mode can be bought. A sourcer who internalises this stops applying pressure in the wrong places and starts protecting the two tracks that genuinely cannot be shortened.

The most under-managed station is the approval loop. In this case three weeks were allocated, and most of it went to internal alignment between the sourcer, the merchandiser and the quality team rather than to technical disagreement with the production side. Building a single sign-off meeting into the calendar at the start of that window, with all three functions in the room, is worth more than any amount of chasing afterwards.

Brief and technical pack

About 1 week; fixed by internal calendar rather than by the line

Sampling cycle

7 working days per round, regardless of piece count

Approval and correction

About 3 weeks; mostly internal alignment

Bulk production

30 days from approved sample and deposit

Inspection

AQL 2.5 under ISO 2859-1; lot 1,100 takes a level II sample of 80 pieces

Freight from Xiamen

Variable by mode and destination; quoted FOB Xiamen

Specifying for Store Conditions, Not for a Sample Room

A knitted bag heading for a store floor is specified against abrasion and light rather than against how it photographs, which for this range meant Martindale abrasion to ISO 12947 at 20,000 rubs for everyday styles and 40,000 for the loaded carry.

Store conditions are harder on a bag than any sample room. A bag hanging on a fixture rubs against its neighbour every time a shopper moves along the rail, sits under retail lighting for weeks, and gets handled by dozens of people who are not trying to be careful. Those three exposures map to three tests: abrasion to ISO 12947, colour fastness to light to ISO 105-B02 at grade 4 or better, and colour fastness to rubbing to AATCC 8 at grade 4 or better. Specifying all three at the brief stage costs nothing; discovering that the bag fails one after bulk arrives costs the entire programme.

There is also a structural version of the same point. A bag that looks crisp in a photograph has often been photographed empty. Load it with a laptop and a water bottle and the body slumps, which is why GSM should be set from the intended carry weight rather than from appearance. For this range the everyday tote settled at about 320 g/m squared at E12, and the loaded carry at about 400 g/m squared at E7 in 600D.

Documentation: What Has to Sit in the File Before Shipment

For a multi-market retailer the file usually has to carry social audit evidence, chemical declarations, restricted-substance statements and child-safety documentation where the product could be used by children, and none of it can be assembled after inspection.

The tendency is to treat documents as a closing task, something to chase in the final week. That is the wrong shape, because the slow parts are the ones belonging to third parties. A BSCI or Sedex SMETA audit report has to cover the production site and be inside its validity window; a REACH declaration has to trace back through the yarn supply chain; a Prop 65 assessment depends on the materials finally chosen. Each of these has a dependency outside the sourcer's control, so each has to start early and be tracked as a line item rather than assumed. Reference points worth keeping include the [restricted substance list maintained under REACH](https://echa.europa.eu/) and the [Prop 65 programme administered by California OEHHA](https://oehha.ca.gov/proposition-65).

Sedex SMETA and BSCI are not interchangeable, and a retailer that accepts one often rejects the other. Establishing which one the buyer's own compliance team accepts, before the technical pack is finalised, avoids arriving at inspection with a file that is thorough and unusable.

Packaging Is Where Retail Programmes Quietly Slip

Packaging slips because it is decided last and is physically in the critical path: carton count, polybag specification, barcode placement, hangtag and any bag-warning print all have to be resolved before bulk can be packed.

In this case the unresolved item was the barcode. The retailer required a GS1-issued number printed to a specified size on the hangtag, with a scannable verification pass at the distribution centre, and nobody had asked for the artwork template until week nine. Packaging artwork is a small job with a long shadow; it takes days to produce and weeks to get wrong. Setting the carton count, the polybag gauge and the hangtag template alongside the technical pack in week one removes an entire class of late-stage surprise.

The interaction between packaging and freight is the second quiet cost. A partly compressed packed carton costs the same to ship as a properly filled one, so a carton count that leaves air in the box is being paid for twice. Where retail-ready packaging is mandated by the buyer this is controlled for the sourcer; where it is a free choice it is one of the few remaining levers on landed cost that requires no quality compromise.

Inspection: What a Failed Lot Costs a Store Group

Release runs to AQL 2.5 under ISO 2859-1, and on a lot of 1,100 pieces the general level II sample is 80 pieces with acceptance at 5 defects and rejection at 6, so the gate is real rather than ceremonial.

The cost of failing is asymmetric, and that is why the number deserves attention at the brief rather than at the port. A rejected lot means a remake or a sort, either way pushing past the floor date, and the consequence of missing the floor date is not a late invoice but an empty fixture during the weeks the range was planned to earn. Retailers plan markdown cadence, staffing and window space around a range date; a late arrival disrupts all three, and the charge that gets remembered is the one nobody planned for.

The correct posture is to treat inspection as part of the specification rather than as a gate at the end. Writing the defect classification into the technical pack, along with the acceptance number, means both parties are measuring the same thing when the samples come off the line. Disputes at inspection are almost always definition disputes, and they cost far less to resolve on paper.

What the Sourcer Would Do Differently at Week Zero

Given the range again, the sourcer would fix three things before the first sketch: the pair count, the test list and the packaging template, because those three produced every hour of avoidable work in the twelve weeks.

Nothing in that list is about production. The two fixed blocks of 7 working days of sampling and 30 days of bulk were never the problem, and neither was the quality of the sewing because there was none. The friction was in decisions that sat inside the sourcer's own organisation and only surfaced once the calendar was tight. That is the characteristic failure mode of retail buying generally: the external supply chain runs to a published clock, and the internal one runs to whatever attention is available.

One structural change is worth adding. Assigning a single internal owner for the technical file, rather than distributing it across merchandising, quality and compliance, removes the queueing delay that consumed most of the approval window. The file does not get lighter; it just stops waiting for someone else's week to finish.

Reading This Against Your Own Range Plan

Take any open range plan, multiply styles by colourways to get the pair count, and compare it against the floor date using 7 working days of sampling plus 30 days of bulk as fixed: if fewer than four weeks remain unaccounted, the plan has no room for a correction round.

That calculation takes about two minutes and repeatedly finds problems that weeks of status meetings do not. The reason is that the plan and the calendar are the same document, and reading them together makes the slack visible. Where the answer is negative the remedy is nearly always editing pairs rather than accelerating production, and editing pairs is a decision the sourcer already owns.

Frequently Asked Questions

People Also Ask

The Plan and the Calendar Are One Document

Twelve weeks is enough time for a knitted bag range, provided nobody mistakes it for eleven weeks of production time and one week of paperwork. The two fixed blocks of 7 working days and 30 days account for barely half of it; the remainder belongs to decisions inside the buying organisation, and those are the ones that slip.

For the wider market picture behind these buying seats, see knitted bag market trends. For the adjacent profiles, see promotional bag buyer and seasonal bag demand planning.