Break-Even Analysis With AI: A Worked Example for a New Product

Break-even units equal the product's fixed costs divided by contribution per unit (price minus variable cost per unit). For the illustrative cold brew can below: $9,600 of first-year fixed costs ÷ $1.57 average contribution per can = about 6,110 cans. Then convert that into weeks at realistic sales rates and check it against minimum runs and shelf life.

The formula is simple; the inputs are where launches go wrong. Forgotten costs, a single average price across channels that sell at very different prices, and a sales rate that assumes full speed from week one all make break-even look closer than it is. AI is useful for two jobs here: finding costs you left out, and running scenarios quickly. It is less reliable at the arithmetic of mixed channels, as the example shows, so every number it produces should be rebuilt in a spreadsheet you can see.

Follow me on Instagram@sagnikteaches

The roaster, the product and the question

An illustrative specialty coffee roaster with five staff sells roasted beans online, at a weekend market and wholesale to about 30 cafés and delis. Several cafés have asked for ready-to-drink cold brew. The plan: a 250 ml can, made from the roaster's own coffee by a contract canner (a co-packer), sold chilled through the same three channels.

Connect on LinkedInSagnik Bhattacharya

The owner's question was blunt: "How many cans do we need to sell before this stops costing us money, and how long will that take?" A second question came later, and it turned out to matter more: "Can we sell the first production run before it goes out of date?"

Subscribe on YouTube@codingliquids

Step 1: List every cost, with AI hunting for gaps

Start with the costs you know, then ask an assistant to find the ones you do not. Ask for categories, not figures; you want it to jog your memory, not invent prices.

I run a small specialty coffee roaster. I'm launching a 250 ml canned cold brew,
produced by a contract canner using our coffee, sold wholesale to cafés, online in
12-packs and at our market stall. Chilled product.

Costs I have so far: coffee per can, co-packer fee per can, cans and labels,
artwork design, launch photography, an extra fridge.

List the cost categories I've probably missed, grouped as one-off launch costs,
ongoing fixed costs and per-can costs. For each, one line on why it applies to a
chilled canned drink. Do not suggest prices.

An illustrative reply, trimmed to the useful items:

One-off launch costs: recipe and extraction trials; initial shelf-life and safety
testing; co-packer set-up or first-run fees; barcode registration if not held.
Ongoing fixed: extra product liability cover; event or trade stand fees; refrigerated
storage running costs; ongoing batch testing if required per production run.
Per can: trays or cartons; freight from the co-packer; delivery to cafés; online
packaging and courier cost net of any shipping charged; payment processing fees;
waste from out-of-date stock; samples given to cafés.

Most of that is right and several items were genuinely missing: freight from the co-packer, samples, waste and the insurance change. Two things to fix. Batch testing was listed as both ongoing fixed and per can; the owner checked with the co-packer and it is charged per production run, so it became a per-can figure averaged over a typical run. And "barcode registration" was irrelevant because the roaster already had barcodes for its bags. Whether a chilled drink needs particular safety testing depends on the product and where it is sold; the co-packer and a food-safety adviser, not the chat assistant, set that requirement.

Step 2: Sort each cost into fixed or variable

A variable cost rises with every can sold. A fixed cost does not, at least within the first year. Launch costs are fixed too, but it helps to mark them separately because they do not come back in year two.

All figures are illustrative. Your co-packer's quote and your own roasting costs replace them.

Step 3: Contribution per can in each sales channel

Contribution is what each can leaves over after its variable costs, available to pay the fixed costs and then to make profit. It differs by channel, because both the price and the channel costs differ.

The weighted contribution uses the mix: (0.6 × $1.25) + (0.3 × $1.82) + (0.1 × $2.76) = $0.750 + $0.546 + $0.276 = $1.572 a can. The weighted average price is $2.89, so the weighted variable cost is $1.318.

Step 4: Break-even in cans, dollars and weeks

Break-even cans     = fixed costs / weighted contribution per can
                    = 9,600 / 1.572        = 6,107 cans (round up)
Break-even revenue  = 6,107 x 2.89         = about $17,650
Year-two break-even = 2,700 / 1.572        = about 1,720 cans
                      (ongoing fixed costs only: insurance, event fees and
                       a $1,000 marketing budget)

Cans become weeks through a sales plan. The owner expected a slow start: about 131 cans a week for the first eight weeks while cafés tried it, then about 262 a week (13 cafés taking a case of 12 each week, around 80 cans online and 26 at the market).

First 8 weeks:  8 x 131      = 1,048 cans
Still needed:   6,107 - 1,048 = 5,059 cans
At 262 a week:  5,059 / 262   = 19.3 weeks
Break-even at about week 28, roughly six and a half months after launch.

First-year sales at this plan: 1,048 + (44 x 262) = 12,576 cans
Margin of safety: (12,576 - 6,107) / 12,576 = 51%
First-year profit: (12,576 - 6,107) x 1.572 = about $10,170

The margin of safety says sales could fall about half short of plan before the product loses money in year one. That is a comfortable cushion on paper. The scenarios test whether it survives contact with reality.

Which channel carries the product

Running break-even as if each channel sold alone shows where the product really earns its keep: wholesale only, 9,600 ÷ $1.25 = 7,680 cans; online only, 9,600 ÷ $1.82 = about 5,280; market only, about 3,480. Every hundred cans that move from wholesale to online bring break-even closer by a noticeable amount. That does not mean dropping wholesale, which provides the volume and the steady weekly orders. It does mean the online listing, the bundle with a bag of beans and the market stall display deserve as much attention as the café pitch.

Step 5: The scenarios, and two errors the AI made

Asking an assistant to run scenarios is quick. The owner uploaded the tables above and asked for break-even under five changes. The corrected results:

The first draft had two errors worth knowing about, because they are the usual ones.

• A simple average instead of a weighted one. Asked for a single break-even figure, the assistant averaged the three channel contributions ((1.25 + 1.82 + 2.76) ÷ 3 = $1.94) and reported about 4,940 cans. That ignores the fact that 60 per cent of sales go through the lowest-margin channel. It understated break-even by nearly 1,200 cans.
• No ramp-up. Asked how many months until break-even, it divided fixed costs by a full-speed monthly contribution and answered 5.4 months. The slow first eight weeks push the real answer to about six and a half.

Both came from reasonable-sounding shortcuts. Neither would be visible if you only read the answer, which is why the next step exists. Why AI slips on maths explains the pattern, and the fix is always the same: make the model show its working, then rebuild it where you can see every cell.

Step 6: Rebuild it in a spreadsheet and Goal Seek it

Put the inputs in labelled cells so every scenario is one change away: prices, variable costs and mix per channel, and total fixed costs. Then:

Weighted contribution (B20):  =SUMPRODUCT(Price_range - Variable_range, Mix_range)
Break-even cans, exact (B21): =FixedCosts / B20
Break-even cans, rounded:     =ROUNDUP(B21, 0)
Profit at planned volume:     =PlannedCans * B20 - FixedCosts

To answer "what wholesale price would we need to break even at 5,000 cans?", use Goal Seek (Data, What-If Analysis, Goal Seek): set cell B21 to 5,000 by changing the wholesale price cell. Point it at the exact figure rather than the rounded one, because Goal Seek works poorly on a formula that jumps in whole steps. In this model the answer is about $2.98, which tells the owner that selling fewer cans would need a price cafés had already said was too high. An assistant can explain a Goal Seek result or suggest what to test next, but the spreadsheet should hold the numbers. If you build the sheet with AI help, the job costing sheet walkthrough uses the same approach of named inputs and checkable formulas.

The check that changed the launch: minimum run and shelf life

Break-even assumes you only pay for cans as you sell them. A co-packer does not work like that. This one had a minimum run of 3,000 cans, and the chilled product had an illustrative shelf life of 90 days from canning. So the owner asked a different question: how many of the first 3,000 would sell inside 90 days?

Weeks 1-8 at 131:   1,048 cans
Weeks 9-13 at 262:  1,310 cans
Sold within ~13 weeks: 2,358 of 3,000
Left at expiry: about 640 cans (21% of the first run)

At $1.07 of variable cost each, 640 wasted cans cost about $690 and push break-even out by roughly 440 cans. Worse, the cash for all 3,000 cans goes out before most of them sell. The break-even model said yes; the first run said "not like this".

Profit break-even is not cash break-even

The model says the product covers its costs around week 28. Cash tells a different story, because the launch costs are paid before any cans sell and cafés pay on 30-day terms. An illustrative monthly view, with contribution of about $890 a month in the first two months and about $1,780 a month after that:

So the money spent on launch comes back in cash around month seven, not month six and a half, and the roaster needs roughly $9,000 of its own cash tied up for the first few months. Add the first production run, paid for up front, and the peak is higher. That is a question for a cash forecast rather than the break-even model, but it is worth asking before canning, not after.

What the roaster decided

1. A smaller first run. The co-packer agreed to 2,000 cans at a $0.10 higher fee per can for the first run only. That costs $200 more but avoids around $690 of likely waste.
2. Pre-sell wholesale. Eight cafés committed to a case a week for the first month before canning, which lifts the early sales rate and shortens the slow start.
3. Hold the wholesale price at $2.40. The scenarios showed that dropping to $2.20 costs about 500 extra cans of break-even, more than the extra volume a lower price was likely to bring.
4. Review at 1,000 cans. Re-run the model with real prices, mix and waste after the first 1,000 cans, and decide on the second run then.
5. Count the owner's time. Break-even with the owner's time included (about 9,100 cans) is still inside the first year's plan. That was the figure the owner used for the go decision, because it answers the honest question.

Five ways break-even flatters a new product

• Leaving out your own time. Shown above: it moved break-even by nearly 3,000 cans.
• Ignoring what the new product takes from old ones. If one in five online cold brew orders replaces a bag of beans that would have earned $7.50 in contribution, each online 12-pack loses $1.50 of that, which cuts online contribution per can by about $0.13.
• Forgetting samples and giveaways. Two free cans to every café you pitch is a marketing cost. Put it in.
• Assuming the planned mix. The cheapest channel often grows fastest, because wholesale customers order in cases. Re-run with the mix you actually get.
• Treating break-even as the finish line. A product that just breaks even ties up fridge space, cash and attention. Decide in advance what profit would make it worth keeping.

Running the same analysis on other products

The steps do not change, but the costs that matter do. Three quick illustrations:

• A handmade jewellery seller launching a collection has fixed costs in moulds, samples and photography, and variable costs dominated by metal. Because metal prices move, run the break-even at today's price and at 15 per cent higher before setting retail prices.
• A craft brewery adding a seasonal beer has a hidden fixed cost: the tank time it takes from a core beer. If the seasonal occupies a fermenter for three weeks, the lost contribution from the core beer that tank would have produced belongs in its fixed costs.
• An online clothing shop launching a capsule range faces minimum order quantities per size and colour, and returns. Contribution per unit should use the price after the expected return rate, and the size curve matters: a minimum of 50 per size in five sizes is 250 units whether or not size XS sells.

Break-even tells you whether the product can work; pricing decides how comfortably. What AI can and cannot tell you about pricing covers the research side, and once the product is selling, checking margins product by product keeps the numbers honest. If the model shows you need a higher price on existing lines to fund the launch, planning and announcing a price increase is the next read, and forecasting next quarter's sales from your history helps set the sales rate that turns break-even cans into a date.

Break-even questions for a new product

Is break-even the same as payback?

Not quite. Break-even is the sales volume at which the product's contribution covers its fixed costs, so profit is zero. Payback is about cash: how long until the money you spent up front has come back. They differ when customers pay on credit, when you buy stock in large runs before selling it, or when launch costs are paid months before sales start. Work out both for anything that needs a big first order.

Should I include the cost of my own time?

Include it at least as a scenario. If the new product takes a few hours of your week and that time would otherwise go on work that earns money, it is a real cost. Leaving it out makes a product look viable when it only works because you are not paying yourself for it. Some owners show break-even with and without their time so the decision is explicit.

What if the product comes in several sizes or flavours?

Work out the contribution for each variant, then weight them by the sales mix you expect, just as you would for sales channels. Fixed costs shared across variants, such as artwork or a launch campaign, stay in one pot. Re-run the numbers once you have two or three months of real sales, because the actual mix is rarely the one you planned.

Further reads

• How to Build a 13-Week Cash Flow Forecast With AI Help — Check you can fund the first production run before the sales arrive.
• How to Calculate the ROI of an AI Automation Before You Build It — The same fixed-versus-ongoing logic applied to an automation project.
• How to Compare Supplier Quotes Side by Side With AI — Compare co-packer or supplier quotes line by line before you commit.
• How to Prepare a Business Loan Application With AI Help — Use the break-even model in a funding application.
• How to Find Dead Stock and Slow Movers With AI — Spot early if the new product is turning into slow stock.
• What Is Cash Flow Forecasting? A Plain-English Guide With AI Examples — A plain-English primer on the cash side of a launch.
• How to Use AI to Understand Your Profit and Loss Statement — Five steps to get a plain-English reading of your P&L from AI, with prompts, sample answers, a tour operator's year and the sums you must check yourself.
• AI Tools and AI Development: The Complete 2026 Guide — the AI hub, including every tutorial in the AI-for-business series.

Sources: Microsoft Support article on using Goal Seek in Excel (Data, What-If Analysis). All business figures in this tutorial are illustrative.