Snack Business Plan: Unit Costs, Margins and Co-Packer Quotes With AI

A snack business plan should show cost per pack split into ingredients, packaging, conversion and freight; the price at every step to the shelf; the margin left after retailer, distributor and promotion cuts; and the cash needed for the first minimum run. AI can build the sheet and compare co-packer quotes, but you must check every figure.

Most snack plans fail at one spot: they take a cost, add a markup and call the result a margin. The shelf price is not yours. Between a customer's $4.49 and your invoice sit a retailer who wants a large share, perhaps a distributor, a promotions budget and some waste. Start from the shelf price and subtract downwards, and the "40% margin" often becomes 20%.

Follow me on Instagram@sagnikteaches

A co-packer is a contract manufacturer who makes and packs your recipe in their own factory, and for a young brand most of the numbers that matter come from their quote. The running example is a 40 g bag of seasoned baked chickpeas. Every figure is illustrative; replace them with your own quotes.

Connect on LinkedInSagnik Bhattacharya

The plan's numbers, in the order a lender or retailer reads them

A snack plan does not need fifty pages. It needs a handful of tables that agree with each other. These are the ones worth building before you write any prose.

Subscribe on YouTube@sagnikteaches

1. Unit cost: what one finished, boxed pack costs you, delivered to your warehouse.
2. Price waterfall: shelf price down to what you actually receive.
3. Margin per pack and per channel: direct, via retailer, via distributor.
4. First-run cash: the money out before the first pack sells.
5. Break-even: how many packs recover fixed costs.
6. Quote comparison: at least two co-packer options on the same basis.

The order matters. The quote comparison feeds the unit cost, the unit cost feeds the margin, and the margin feeds break-even. Build the sheet so that changing one quote changes everything beneath it.

Step one: list every cost that lands on one pack

Founders usually remember ingredients and forget the rest. Walk through the pack's journey and write down each cost, with the unit it is quoted in, because mixing "per kilo" and "per pack" is the most common cause of wrong totals.

• Ingredients. Weight per pack times price per kilo, plus waste. Dry goods lose a few per cent in processing; add a yield factor.
• Conversion. The co-packer's fee for cooking, seasoning and filling. Often quoted per pack or per kilo.
• Primary packaging. The bag or pouch (film), printed with your design.
• Secondary packaging. The shipper carton holding, say, 12 or 24 packs, plus labels or pallet wrap.
• Set-up charges. Printing plates or cylinders for film, product trials, line cleaning, which are one-off but must be spread across the run.
• Freight in. Moving ingredients or packaging to the factory, and finished stock to you.
• Quality and testing. Lab tests, shelf-life checks, metal detection if charged separately.
• Over-run and loss. Runs can come in a few per cent over or under the order; rejected stock is still paid for.

You will not know all of these on day one. Put a figure in each row anyway, mark the ones that are guesses, and replace them as quotes arrive. The point of a sheet is that gaps are visible.

Ask AI to check the list for gaps

I am planning a packaged snack: a 40 g bag of seasoned baked chickpeas, sold in cartons of 12, made by a contract co-packer. Below is my list of cost lines per pack. List anything a first-time brand typically forgets, say which lines are usually quoted per kilo versus per pack, and tell me which lines are one-off and need spreading across the run. Do not give prices.

My lines: chickpeas, seasoning, conversion fee, film, carton, freight to my warehouse.

It will usually remind you of waste and yield, film plates, product testing, pallets and storage. It cannot know your prices, so do not ask it for any.

Step two: build the unit-cost sheet

Set the sheet up with one column per line, and put assumptions (run size, yield, units per carton) in their own cells at the top so formulas refer to them. Hard-coded numbers buried inside formulas are how errors survive for months.

For the chickpea bag, imagine two quotes arrive. Co-packer A offers a turnkey service: they buy the ingredients and make the product. Co-packer B charges for conversion only; you buy the ingredients and film yourself and send them in. Here are the figures, all illustrative.

Check one of these sums yourself. Quote A: $1.18 + $0.14 + $0.04 + $0.06 = $1.42 per pack, times 12,000 = $17,040, plus $1,800 set-up = $18,840, which is $1.57 per pack. Quote B: $0.70 + $0.46 + $0.16 + $0.05 + $0.03 = $1.40, times 5,000 = $7,000, plus $600 = $7,600, which is $1.52 per pack. These small checks matter because AI models sometimes slip on multi-line arithmetic; why AI is bad at maths explains where and why.

The headline: the two quotes land within five cents per pack, but A needs $11,240 more cash and a run more than twice the size. Which is better depends on whether you can sell 12,000 packs before the product goes stale and whether you can afford to wait for the money back. Those two questions, not the unit price, should drive the choice.

One more catch hides in the table. Quote B's $1.52 still leaves out delivery from the factory to your warehouse, because B quoted it separately. If that delivery costs the same $0.06 a pack that A charges, B lands at $1.58, a cent above A. A gap that small means the decision rests on cash, run size and terms, not on the unit price.

Step three: work backwards from the shelf price

Now the step that changes plans. Start from what the customer pays in the shop and subtract each party's share. The names of the parties vary, but the shape is the same.

Take a shelf price of $4.49. Say the retailer expects a 40% margin on the shelf price, so they pay $2.69 for the pack. If a distributor sits between you and the retailer and keeps 25% of the price the retailer pays, you receive about $2.02. Retailers and distributors often add promotion allowances, damage allowances or listing fees on top; assume 5% of your invoice for now, which is $0.10.

Retailer margin percentages differ between shops, categories and buyers, and I cannot give you a reliable figure here. Ask each buyer directly, and ask whether the percentage is taken from the shelf price or the cost price, which changes the sum. The same goes for distributor shares and allowances: treat all three as placeholders until you have them in writing.

The lesson is the gap between the two columns. The same pack earns 41% direct and 21% through a distributor, so the route to market decides whether the business works. Your plan should show both, and say which you will use first. If you sell through your own online shop, the sum differs again: the customer pays $4.49 or more, but you carry postage, payment fees and packaging for shipping.

Step four: first-run cash and break-even

A lender, an investor or a sceptical partner will ask how much cash you need before the first sale. For the chickpea bag it is the run cost plus the items nobody quotes: design fees, a barcode, insurance, product-liability cover, initial testing and storage.

Say you add $2,500 of one-off costs (design, label proofs, barcodes and tests) to Quote B's $7,600 run. Your cash need is $10,100 before a single pack leaves the warehouse, and probably more if the co-packer asks for a deposit and the balance on delivery, while retailers may pay 30, 60 or more days after invoice (ask each buyer for their terms). A plan that ignores payment terms shows a profit while the bank balance is empty.

Break-even on that set-up is simple: fixed costs divided by margin per pack. At the distributor margin of $0.40, the $2,500 of one-off costs needs 6,250 packs, which is more than the whole 5,000-pack first run. Selling direct to retailers at $1.04 a pack needs about 2,400. This is why plans that rely on distributors need either larger volumes or a lower cost.

The sums are simple enough that a spreadsheet will do them reliably. If you are not comfortable building one, the guide to using ChatGPT with Excel shows how to have AI write formulas you can inspect, rather than numbers you must trust.

Step five: get quotes that can be compared

Quotes arrive in different shapes: one per kilo, one per pack, one with film included, one without. If you ask each co-packer to quote from the same specification sheet, comparison becomes mechanical. Send the same one-page brief to every candidate.

• Product description, pack weight, count per carton and per pallet
• Recipe or a statement of whether you supply it, and the ingredients' allergens
• Run sizes to quote: your intended first run, and two larger steps (for example 5,000, 10,000 and 20,000)
• Who supplies ingredients, film and cartons
• Shelf-life requirement and storage conditions
• Certifications you need, such as a recognised food safety standard
• Delivery address and the date you would like the first run

Ask for the price at each run size, because the steps tell you how much a larger commitment saves. Ask also about the things that rarely appear in the headline price: payment terms, lead times, how an over- or under-run is charged, what happens to a batch that fails testing, who owns leftover ingredients and packaging, and whether your recipe stays confidential. For a first-time brand these terms often matter more than a cent per pack.

Turn two quotes into one table with AI

Quotes often arrive as PDFs, emails or spreadsheets. AI is good at lifting the numbers into a standard layout, so long as you check the result against the originals. Remove names and any commercially sensitive details you do not need to share, and use a plan that does not train on your content if the recipe is confidential.

Below are two co-packer quotes for the same product (40 g seasoned chickpea bag). Extract every cost line into one table with these columns: quote, line item, unit (per pack / per kg / one-off / per pallet), amount, run size it applies to, included or excluded, notes.

Then list:
1. Anything that is missing from one quote but present in the other (for example freight, set-up, testing).
2. Any line where the unit is unclear.
3. Terms that affect cash: deposit, payment terms, minimum run, over/under-run tolerance, lead time.
Do not calculate totals yet and do not recommend a supplier.

QUOTE A: [paste]
QUOTE B: [paste]

An illustrative extract of what you might get back:

Missing from Quote B but in Quote A: freight to customer's warehouse (B says "ex works").
Unclear unit: Quote A "film surcharge $0.02" - per pack or per 1,000?
Terms: A requires 50% deposit, balance before dispatch, lead time 8 weeks, over/under-run +/-10%.
       B requires 30% deposit, balance 30 days after delivery, lead time 6 weeks, +/-5%.

That is the most valuable output: the questions to send back. "Ex works" names the factory as the delivery point, so ask the co-packer to confirm who arranges and pays for transport from there, and add that cost. The unclear surcharge is two cents a pack or almost nothing depending on the unit, a gap of about $240 on a 12,000 run. Ask the co-packer, not the AI.

Five situations worth rehearsing before you sign

The cheap quote with expensive terms

One quote is $0.08 per pack cheaper but needs full payment before dispatch and quotes "ex works". Add the missing freight ($0.05), the cost of paying 60 days earlier than the alternative (the interest is small, but the cash gap may force an overdraft), and the lead-time difference. The cheap quote often is not.

A bigger run that saves five cents

The co-packer will cut the price by $0.05 per pack if you run 20,000 instead of 10,000. That saves $1,000 in cost but adds $14,000 of cash and stock to sell. If your sales forecast is 1,500 packs a month, 20,000 is more than a year's supply, and the product's shelf life may be shorter than that. Work out months of stock before you take the saving.

Over-run and under-run

The quote allows 10% either way. You order 10,000 and receive 9,100. If the price is per pack delivered, the cost per pack is unchanged but your plan is short 900 packs. If you ordered ingredients and film for 10,000, you hold surplus materials. Put the tolerance in the sheet as a low case and a high case.

A film price that depends on the run

Printed film is often cheaper per pack at larger volumes, and has its own minimum order. A co-packer may quote film at $0.14 for 12,000 and $0.19 for 5,000. If you commit to the smaller run, use the higher price. Mixing a large-run film price with a small-run conversion price understates cost, and it is an easy mistake when quotes arrive on different days.

A failed batch

A batch fails the co-packer's moisture test. Who pays for it? The answer should be in writing before the run, with a clause for rework, replacement and who owns the ingredients. If the quote is silent, assume you pay, and put a contingency in the plan. A realistic contingency for a first run is a few per cent of its cost, though only your quote history can say how much.

The comparison sheet, from top to bottom

Build the comparison on one tab so you can see it at once. A layout that works:

1. Assumptions block (top): run size, shelf price, retailer margin, distributor share, allowance percentage, units per carton.
2. Quote columns: one per co-packer, one row per cost line, identical row labels.
3. Calculated rows: set-up per pack, landed cost, cash for first run, months of stock at forecast sales.
4. Margin block: per channel, per quote, in dollars and percent.
5. Terms block: deposit, payment terms, lead time, tolerance, shelf-life support, confidentiality, with a plain "good, fair, poor" mark.
6. Notes: open questions and who you asked.

If you want AI to help, ask it for formulas rather than results: "Write the Excel formula for landed cost per pack in cell D12 using the values in D4 to D10, and explain each term." You get something you can read, test and audit. Checking your margins product by product with AI applies the same discipline to an existing range, and the cost-per-person sheet for a caterer shows a similar structure for food costing.

Then stress-test. Ask what happens to margin if ingredient costs rise 10%, if the retailer asks for 45%, or if the first run sells half as fast as planned. A short "what if" table of three or four scenarios is more persuasive to a lender or buyer than a single hopeful number.

What to put in the plan itself

The prose should be short, because the tables do most of the work. For each section, one paragraph of reasoning and the table:

• Product and unit economics. The landed cost, how it was built, which quote you chose and why.
• Route to market and pricing. Shelf price, channel mix, and the margin in each channel.
• First run and funding. Run size, cash needed, how it is funded, and when the money comes back.
• Risks. Ingredient price rises, a failed batch, late payment, slow sell-through and shelf life. Each with a one-line plan.
• Break-even and next steps. The volume that recovers set-up costs, and the trigger for a second run.

AI is useful here for drafting and for tightening, but it must work from your numbers. Paste the final tables and ask it to write each section using only the figures provided, and to say "not supplied" where something is missing. Then read it against your sheet, because a fluent paragraph can contain a wrong number.

Using ONLY the figures in the tables below, write the "Unit economics" section of a snack brand plan in 150 words, plain English. If a figure you need is missing, write [MISSING: description] instead of estimating. Do not add market statistics or claims about competitors.

TABLES: [paste landed cost table and margin table]

Finally, keep the quotes, the sheet and the plan in step. Quotes expire, ingredient prices move, and a plan that quotes a three-month-old price loses credibility with the first retailer who asks. Record the date of each quote in the sheet and a "valid until" date next to it. Once the first run is selling, setting reorder points and forecasting the second run will tell you whether the first-run assumptions held.

A final caution about food: ingredient declarations, allergen labelling, nutrition panels and health wording are regulated, and the rules differ by market. Before any pack goes on sale, confirm labelling with your co-packer and a qualified adviser, and do not let AI write allergen or nutrition information from memory.

Questions from first-time snack founders

What margin should a packaged snack brand aim for?

Aim for a gross margin you can still live with after the retailer, any distributor, promotions and spoilage have taken their share. Work backwards from the shelf price rather than forwards from cost. The figure left after those deductions is usually far lower than the margin on your own price list, which is why the waterfall matters more than the markup.

Is the cheapest co-packer quote the best one?

Not necessarily. Compare landed cost per pack, minimum run size, who buys ingredients and packaging, payment terms and what happens if the batch fails. A quote that is 5 cents cheaper per pack but needs three times the cash up front can sink a young brand.

Can AI read a co-packer's quote for me?

Yes, for extraction. Paste or upload the quote and ask it to list every cost line, assumption and unit, and to flag anything ambiguous. Then check its numbers against the original by hand, because a missed surcharge or a swapped unit can change your margin.

How many quotes should I get?

At least three for the first run, ideally from co-packers who already make similar products. Two quotes give you a comparison; three show you what is normal and which one is an outlier. Ask all of them to quote the same specification and the same volumes.

Should the first run be the minimum run?

Usually yes, if you can sell it. A larger run lowers cost per pack but ties up cash and risks stock reaching the end of its shelf life unsold. Match the first run to what you can realistically sell well inside the product's shelf life, and ask each retailer how much remaining life they expect on delivery.

Further reads

• How to Automate Purchase Orders and Supplier Emails With AI — Tidy up purchase orders and supplier emails once you are ordering.
• How Small Shops Can Use AI to Forecast Stock and Reorder — Forecast demand so the second run is the right size.
• How Bakeries Can Use AI to Predict Demand and Cut Unsold Stock — Cut unsold stock on a short-life food product.
• How to Write Shopify Product Descriptions With AI That Convert — Write the product pages for your direct sales.
• AI KPIs for Small Businesses: 12 Metrics Worth Tracking — Pick the figures to watch once you are selling.
• How to Forecast Covers and Cut Food Waste With AI — A waste-focused look at forecasting food demand.
• AI Tools and AI Development: The Complete 2026 Guide — the AI hub, including every tutorial in the AI-for-business series.

Sources: general supplier-quotation practice and the published list of Incoterms rules (EXW, Ex Works); the numbers in the examples are illustrative, not market data. Spreadsheet tips draw on the Microsoft Excel and Google Sheets help pages and the ChatGPT and Claude file-upload documentation.