Eight things to pin down before a supplier can quote you anything

Stage 1 of eight. This article is one step in the production line buyer’s roadmap — the full sequence from defining the product to the first year of operation.

The order nobody uses

Most first-time projects begin with a machine: a video, a trade fair stand, a price from a marketplace. The buyer then works backwards, trying to fit a product and a business to equipment that has already fixed most of the important decisions.

The forward order is the product first, then the process, then the layout and building, then the equipment. We have written about the process half and the site half separately. This article is about the first step of all — the eight things you have to decide before anyone can design anything, and which you, not a supplier, are best placed to decide.

1 — The pack format and sizes

Not “bottled water” but 500 ml and 1.5 litre PET, screw cap, wrap-around label, shrink-wrapped in twelves. Not “maize meal” but 10 kg and 25 kg woven polypropylene, stitched, printed. The pack decides the filler, the sealer, the labeller, the end-of-line and half the conveyor design, and changing it after the order is a different line rather than a modification.

Two sub-decisions get missed. Who supplies the packaging, and is it consistently available in your market at the specification the machine expects — film thickness, bottle neck finish, cap dimension, bag gusset? And what does the pack have to survive: your distribution is not a European supermarket chain, and a carton that works in a container may not work on a bakkie on a gravel road.

2 — The recipe and its process conditions

Every formulation carries conditions: temperatures, times, shear, pH, moisture targets, cooling curves. Those conditions, not the product name, determine the equipment. Two juice products with different acidities need different materials and possibly different pasteurisation. Two biscuits with different fat contents behave differently on the same forming line.

If the recipe is still developing, say so and size for the range rather than a point. What does not work is fixing a machine against a recipe that changes afterwards.

3 — The raw material specification, and how much it actually varies

This is the one most likely to go wrong in African projects and the one least likely to be on a specification sheet. The point is made well from the tooling world, where the same lesson is learned expensively: “give someone the finished metal part. Give them the mould too. They still might not make it reliably. What's missing is the recipe for the material going in.”

The same is true across every sector. Moisture in grain, foreign matter in a delivery, oil content in a seed, viscosity in a batch, resin grade in a bag of pellets — each determines what cleaning, sizing and conditioning has to happen before the main process, and how tolerant the main process must be.

Where this bites locally is variability. A line specified against an ideal material and fed a variable one does not fail; it runs at reduced rate with elevated scrap, permanently, and the cause is never obvious. Industry reporting on local manufacturing returns to this repeatedly — processors describing difficulty securing consistent supply of quality domestic inputs, season by season.

What to do: test your actual material, from your actual suppliers, across a season if the timeline allows. Then specify the range the line must handle rather than the average, and make that range part of the acceptance test — the line should be proved on your worst realistic material, not your best sample.

4 — The target output, sized to demand

Output should come from what you can sell, not from what the brochure can make. A line sized to an ambition and running two days a week is a loss-maker with excellent equipment. This is covered at length in your first production line, and it remains the most common and most expensive first-line error.

5 — The shift pattern — which halves or doubles the machine

Output is a rate multiplied by hours, and the hours are your decision. Ten tonnes a day on one shift needs roughly twice the machine of ten tonnes a day on two shifts, and roughly twice the capital.

So decide the shift pattern before the capacity, honestly: can you staff, supervise and maintain a second shift where your plant is? Is there a security, transport or overtime cost that makes night work impractical? A smaller line running longer is usually cheaper to buy, cheaper to connect, easier to keep fed and less exposed to a single breakdown — but only if the second shift is real.

6 — Shelf life, and what it forces upstream

A shelf-life requirement is not a marketing decision; it is an engineering constraint that reaches back through the whole line. It determines thermal processing or the lack of it, the barrier properties of the pack, the sealing integrity you must achieve and verify, whether you need a controlled environment at filling, and what testing you must do to substantiate the claim.

Decide the number before the equipment. Extending it later frequently means new equipment rather than a new process setting.

7 — The standard the product must meet

Who are you selling to, and what will they require? A national standard or mark, a sector regulator's registration, a retailer's supplier audit, an export customer's certification, a food safety system. Each carries implications for hall layout, material flow, surface finishes, separation of wet and dry areas, sampling points and record-keeping.

Every one of those is inexpensive on a drawing and expensive after machines are grouted down. Find out what will be inspected, and design for it from the start — the country guides for Ghana, Nigeria, Kenya and Tanzania set out the regulatory landscape for each market.

8 — The SKU mix, including what you will not make

Every additional pack size or formulation the line must carry adds change parts, changeover time and often capital — and changeover time comes straight off your output.

The discipline is to write three lists: the SKUs the line must run at launch, the SKUs it must be able to run later (which becomes a design requirement for tooling and interfaces), and the SKUs it will not run at all. The third list is what keeps the specification affordable, and it is the one nobody writes down — which is how a line intended for two products arrives able to make six of them slowly.

The eight, on one page

Use the list as a filter in both directions

Send these eight to a prospective supplier and watch what comes back. A supplier who quotes without asking for most of them is offering you a standard machine and leaving the gap for you to find at commissioning. A supplier who asks for all eight, and asks awkward follow-up questions about the third one, is designing a line.

The list also works on you. If you cannot answer four of the eight, the honest conclusion is not that you need a better supplier — it is that the project is not yet at the equipment stage, and a few weeks spent on the product will save months later. That is precisely what a feasibility stage exists to do, and a feasibility that occasionally concludes “build smaller” or “not yet” is the only kind worth paying for.

How we work on this part

We start here, before any equipment discussion. The eight items become a written product brief; the brief drives the process route and mass balance; those drive the machine list and the building requirements. Where an item is genuinely unknown — a recipe still in development, a raw material not yet tested — we specify a range and make it an acceptance condition rather than an assumption. Feasibility and line sizing describes what that stage produces.

Frequently asked questions

Quotes are reproduced verbatim from public industry discussion collected during 2026 and anonymised. Regulatory requirements differ by country and sector and change; treat the standards question as one to confirm with the relevant authority in your market.