Home › Field notes › Assortment Planning for Modular Ranges: How Many Bodies and Modules

Modular Bag Pro · 2026-10-11 · 2,633 words · 12 min read

Assortment planning for modular ranges answers two questions before any others: how many carrier bodies and how many module types enter the first catalogue, and every other assortment decision follows from multiplying those two figures against tooling cost, stock-keeping lines and tied-up cash. A defensible starting judgement — derived from programme arithmetic rather than from published market statistics — sets three or four carrier bodies against six to nine module kinds, wide enough to demonstrate the platform and narrow enough that each reference carries its own minimum. Programme terms bound that judgement: minimum quantity of 500 units per reference, development samples in 6-10 working days extending to 12-15 on complex builds, series production of 35-50 days, inspection at AQL 2.5 and shipment FOB Xiamen. The scope is civilian load carriage — commuting, trade tools, hiking, field work and first aid — with no ballistic or weapon-carriage claim made here; unit counts used below are illustrative planning figures to be replaced by your own sales history, not observed category data.

What an Assortment Matrix Is: Carriers Along the Top, Modules Down the Side

An assortment matrix for a modular range is not a photograph of a catalogue; it is a compatibility record. Carrier bodies run along the horizontal axis, module types down the vertical one, and every intersection carries one of three marks: standard fit as designed, fit with a noted restriction, or no fit. Everything downstream — how many references you can afford, what you can promise a buyer, and which line you reorder first — reads off that grid.

The matrix exists because the platform promise is interdependence. Buyers accept higher unit pricing on a carrier precisely because attaching something later is supposed to work, so any intersection left blank becomes a broken promise at the shop counter. Filling blanks later is expensive: adding compatibility retroactively means re-cutting an interface or issuing an adapter, whereas deciding the whole grid before the first pattern is cut costs nothing beyond discipline.

Two habits make the matrix survive contact with reality. First, write physical reason codes rather than ticks — a module may be marked "restricted" against a compact carrier because it overhangs the side panel by 30 mm, which is a fact a future planner can act on. Second, freeze the shared interface early: geometry documented in detail on the attachment grid reference means every module in the catalogue must respect those numbers or declare an adapter.

The platform's product families supply the horizontal axis in practice: everyday formats on the modular everyday carry page, larger bodies on the hiking-oriented page, and bespoke work under custom modular backpack programmes. Choosing which of these to open first is what the later sections quantify.

Selection rule: Complete the compatibility grid with reason codes before any reference is costed, because the cheapest place to resolve a blank intersection is the drawing, while the most expensive is a buyer standing in a shop with two units that were sold as one system.

How Many Carrier Bodies a First-Year Range Should Open With

Carriers carry the tooling burden, so their count is set by capital rather than by ambition. Each body needs its own pattern, its own sample round at USD 50-150 refunded against the order, and frequently its own set of screens or dies at USD 300-2,500, and none of that cost is shared with a module. A range of six bodies in year one typically spends its entire development budget before a single unit is discounted.

Three bodies cover the demand spread in most civilian channels: a compact everyday unit around 18-22 litres, a mid-size workhorse near 26-32 litres, and a large body around 40 litres for field or multi-day use. Below three, a channel either cannot show the platform or cannot meet extreme sizes; above four, colourway and size overlap starts cannibalising rather than expanding.

Read the grid by column before choosing anything. A column with many "restricted" or "no fit" marks is telling you that body is too far from the centre of gravity of the module family; either widen the body's envelope or accept it as a specialist line that sells on its own rather than as part of the system.

The same grid also settles colourway questions. Because the buyer's eye compares bodies and modules side by side, the colourway plan has to be shared across both axes; three colourways on bodies with two on modules reads as unfinished, and a season with mismatched lot reservations produces visible steps that no amount of photography hides.

Verdict: Open with three carrier bodies and add a fourth only where a specific channel demonstrably needs an extreme size, since the fourth body costs roughly another full development cycle while shifting units that three well-chosen references would have sold anyway.

How Many Module Types Belong in the First Catalogue

Modules carry a different cost profile: low tooling, high stock-keeping. Adding a module rarely requires new tooling — often only a pattern tweak — but each one adds a reference that must be minimum-produced, stored, forecast and eventually discounted. Six to nine kinds covers the reasons people buy into a platform without turning the catalogue into a drawer of near-duplicates.

Attach rate decides the rest. The planning arithmetic below uses illustrative numbers because every programme differs: if eighty units for every hundred carriers sold is entirely unrealistic in your channel, replace it with your own. Assume a first-year carrier run of 3,000 units; assume, for illustration, that one module follows each carrier and that a third of those buyers take a second module. That yields roughly 4,000 module units to be spread across the catalogue, and dividing by the 500-unit floor leaves room for eight references at modest depth rather than twenty at none.

Breadth has a ceiling set by the buyer's attention, not by finance. A wall display with fifteen module types looks like a parts bin; the same space with eight, grouped by job rather than by price, reads as a system. Grouping matters more than count in practice: "diagnostic tools", "daily documents", "hydration and heat" outperform "small, medium, large".

Two module categories deserve special treatment. A first aid module carries labelling and sometimes registration requirements in the buyer's market, which lengthens its introduction; a large carry module effectively turns a mid body into a large one, which means it is competing with your own fourth body.

Bottom line: Launch six to nine module kinds chosen by job rather than by size band, because additional references beyond that point add forecasting work and obsolescence risk faster than they add basket value.

Which Role Each Reference Plays: Traffic, Margin, Credibility or Reserve

Every reference earns a role before the range is finalised, and the role determines how it is priced, how deep it is bought and how fast it is reordered. Confusing the roles is the commonest assortment error: a high-margin module bought deep because someone expected it to drive visits will sit in a warehouse while the traffic body runs dry.

Notice what the table says about stockouts, because it is the point that assortment arguments usually miss. A missing margin module defers an attachment — annoying, usually forgiven. A missing traffic body sends the buyer to another product entirely, taking every future module order with it, which is why "we sold out of the popular body" is a planning failure rather than a success story.

Credibility modules exist for a different reason: they make the platform claim believable. A labelled first aid module or a proper instrument insert may sell in small numbers but converts sceptics, and its thin inventory should be bought with that promotional cost accepted in advance.

Takeaway: Assign one explicit role to every reference and set its depth from that role, because depth chosen per reference rather than per role is what produces a warehouse full of margin modules next to an empty hook where the traffic body hung.

Depth per Reference: How Minimum Quantity Shapes the Count

The 500-unit minimum per reference is the arithmetic that turns a wish list into a plan. Multiplying your candidate reference count by that floor gives the smallest possible opening inventory, and multiplying again by unit cost gives the cash position before a single unit sells. Many first ranges are designed by counting backwards from desire rather than forwards from that floor, which is why they arrive with ten references and no depth in any of them.

Depth and breadth trade directly against each other at constant cash. The same spend that opens twelve references at roughly the minimum delivers a thin, fragile assortment that cannot survive one strong month; six references at double the floor gives enough runway to reorder before the hook empties. For a first year without reliable history, depth in the references you believe in beats breadth across ones you are testing.

Colourway multiplies silently. Three references in three colours is nine purchasing decisions at 500 units each, not three — a common surprise in the first buying meeting. A useful discipline is to hold colourways to two in year one and let the second year add the third where it has demonstrably earned a place.

Reorderability is the other half of the decision. A reference that will be reordered quarterly can be bought thinner than one that will never repeat, because the repeat cycle is itself safety stock. Anything bought once and never again must be bought deep enough to cover its whole life, including the tail.

Judgement: Convert your candidate count into cash before refining it, by multiplying references by 500 units and by colourway count, then cut references rather than depth until the total fits, since thin is worse than narrow in a modular range where every reference has to be available together.

Sequencing the First Twelve Months of a Modular Range

A twelve-month plan recognises that nothing in the range matures at the same time. Bodies need the longest runway because tooling and fitting come first; modules can follow later and still meet the same selling window, provided their interface was frozen with the bodies.

A workable sequence puts months one to three on body development: pattern, first samples returned within 6-10 working days, then refinement rounds where a complex body needs 12-15. Months three to five cover module development in parallel, which works only because the grid geometry was locked in month one. Months five to seven see the first bulk order placed with series production of 35-50 days, then 25-35 days of sea freight or 5-8 by air depending on how much of the season is at risk.

Transit packing deserves attention inside that plan because cube, not weight, tends to set the freight cost for a bag range. Planning figures of roughly 28 CBM for a 20GP against 68 CBM for a 40HQ let you convert unit volumes into container counts early enough to choose shipping mode deliberately; and parcel-shipped replacements should be qualified to the ISTA 3A procedure so single-unit dispatch does not generate damage-based deductions.

Months eight to twelve are the review window. Real sales data replaces the illustrative planning figures used at the start, the compatibility grid is revisited with actual returns reasons, and year-two decisions — a fourth body, a retrenched module family — are made against evidence rather than enthusiasm.

Spec rule: Freeze the interface geometry in month one and hold it unchanged for the entire platform life, because once compatibility becomes a selling claim, altering it mid-stream silently splits your installed base into two halves that can no longer share accessories.

Reviewing and Pruning the Assortment at Season End

Pruning is harder than launching because every reference has an advocate. A defensible rule separates references into three lists using two measures only: units sold against expectation, and whether the reference sells attached or alone. A module that only ever sells attached to one body is doing its job; a module that never sells at all is not, regardless of how much anyone likes it.

The second measure often surprises. Some modules turn out to be the reason a body sells — buyers choose the mid-size carrier precisely because it accepts the labelled module — which means its low standalone unit count understates its value. Recording attach pairs at point of sale, or simply reviewing bundle orders monthly, prevents pruning a complement rather than a drain.

Pruning decisions also interact with promise. Retiring a module is fine; retiring the only module that gave a body a distinctive function changes what that body is sold as, and the listing photography should change with it. Retiring a module is routine; retiring the only one that gave a body its distinctive function changes what the body is marketed as, and the listing photography should change with it. Any revision of that kind ripples into instruction sheets, carton marks and listing copy rather than stopping at the product itself.

Vetted partner facilities provide 4,950 m² of SGS-verified floor staffed by 137 people working 149 machines on 7 production lines, with monthly output across running programmes near 200,000 units; the founder's bag production work began in 2004 and the company has traded since 2014. Because those lines also build the modules, a pruning decision taken early enough can be converted into volume on the references that earned it within the same 35-50 day production window.

Turning Assortment Decisions Into a Buying Calendar

An assortment only exists once it is expressed as purchase orders with dates. Working backwards from each season's selling window gives the sequence: development rounds to pre-production approval, then 35-50 days of series build, inspection at AQL 2.5 with defect classes fixed in advance under ISO 2859-1, then transit. Nothing about that chain compresses, so a decision made two weeks late is a season missed rather than a season trimmed.

Administrative detail belongs in the same calendar. Quotations come back within 24-48 hours on an indicative FOB Xiamen basis, payment terms run 30% deposit against 70% before shipment, and any labelling change intended for a specific market should be raised while the first sample is in hand rather than after the bulk is packed.

Two buffers protect the plan. A reference-level buffer of 5-10% above forecast, and a calendar buffer of two weeks ahead of the computed date, absorb the two events that most often derail a first year: a late decision about the gift of the range, and a documentation query raised after goods are already on the water.

Finally, write the plan down with the assumptions visible. Any plan justified with illustrative unit counts should state plainly which numbers came from history and which were planning placeholders, so the person revising it next year can update the placeholders without rebuilding the whole argument.