Pop-Up Shop Business Plan: Stock, Rent and Staffing in ChatGPT

Put your dates, rent quote, stock costs and rota hours into one block, and ask ChatGPT for a cost table, a break-even in trading days and three sales scenarios, with every sum shown. Then recompute each total by hand: it drafts the structure well, but its arithmetic and its sales guesses are yours to check.

A pop-up is mostly fixed costs for a fixed window, so the plan's real question is how many trading days it takes to cover them. Stock is the flexible part, because unsold stock goes back into the business instead of vanishing. ChatGPT cannot know your footfall, so ask it for scenarios, and set the go or no-go rules before you sign anything. The general method for plans is in the tutorial on writing a business plan with AI and what to check yourself; this one is the short-run, fixed-window version.

Follow me on Instagram@sagnikteaches

Put the facts in one block before asking for any plan

A vague request gets a vague plan. Gather the facts first, in a single block you can paste at the top of every chat. Keep it in a ChatGPT Project so each new chat starts from it; Projects are available on every plan, though file limits differ, so check yours. Here is an illustrative block for a four-week pre-holiday pop-up for a small homeware brand:

Connect on LinkedInSagnik Bhattacharya

POP-UP FACTS (use only these; label anything you add as ASSUMPTION)
Concept: four-week pop-up selling candles, mugs, planters, gift sets and cards
Trading: Tuesday to Saturday 10:00 to 18:00, Sunday 11:00 to 17:00 = 24 trading days
Rent quote: $1,000 a week for 4 weeks, plus $150 a day for 2 set-up days; breakdown day included
Deposit: $1,000, refundable at the end
Service charge and utilities: $400 flat. Insurance for the term: $200
Fixtures and lighting hire: $600. Signage and printing: $250. Card reader and till kit: $100. Marketing: $450
Card fees: 2.6% plus $0.15 per sale. Packaging: $0.35 per sale. Average basket: $42
Products (price / landed cost / base-case units over 24 days):
  Large candle $26 / $9.00 / 310; mug $28 / $9.80 / 240; planter $34 / $12.00 / 110;
  gift set $52 / $18.50 / 90; cards and small items $4.50 / $1.40 / 280
Staff: owner works 27 hours a week unpaid; paid staff $17.50 an hour;
  45 paid hours a week plus 12 hours for set-up and breakdown
Prices exclude any tax on sales.

The card rate in the block is not invented: as of October 2026, Square's pricing page publishes 2.6% plus 15 cents for in-person payments on its Free plan. Your processor's rates will differ, so put in your own. Nothing in the block is a guess about sales except the unit counts, and those are flagged as base-case. Before ChatGPT builds anything, make it ask what is missing. A good first prompt is: "Before you build anything, list the questions you need answered to make this plan reliable. Do not assume." The questions it returns should include the ones below, and each one changes a specific number in the plan, which is a good test of whether the question is worth asking.

Subscribe on YouTube@sagnikteaches

Rent: turn the quote into a cost per trading day

The quote was $1,000 a week for four weeks plus two set-up days at $150, which makes $4,300. Over 24 trading days that is about $179 a day for rent alone, or $204 with the service charge and insurance. Treat the $1,000 deposit as cash that leaves your account but comes back, not as a cost. That one distinction is an easy error to make in a first draft, as the check section shows.

Short-let rent comes in three common shapes: a fixed sum, a share of sales, or a mix. They behave very differently when sales miss the plan. This table prices each against the three scenarios built later (low is half the base-case units, high is 130%). Profit is after all other costs:

The share of sales is better for you in the low and base cases and worse only in the high one, which tells you what the landlord is giving up and why they will often refuse or ask for a minimum. The hybrid is the likely compromise. You would not know any of this from a single "estimated rent" line, which is why the prompt asks for rent in three shapes.

Short-term space is often let on a licence rather than a lease or tenancy, and the wording decides who pays for repairs, what cover you need and how easily either side can end it. Have a solicitor read the agreement. ChatGPT can turn it into a plain-English list of obligations and questions, which makes the solicitor's time cheaper, but it should not be the last reader.

Stock: buy in two orders, not one

The expensive mistake is buying the high case up front. Stock is cash tied up for four weeks, and a pop-up has no room to hold what does not sell. A calmer rule: buy the first two weeks of base-case sales plus 25% (that is 62.5% of the base units), and plan a top-up for week three. Here is the base case and the split:

The base case sells $24,460 for $8,519 of stock at cost, a gross profit of $15,941, or 65.2%. The opening buy costs $5,343.50 and the top-up $3,175.50. If sales come in at half the plan, you never place the top-up and about $1,084 of the opening stock is left over at cost. That is stock to return to the main business or sell later, which is why the low case is less frightening than its loss line suggests.

For a retailer reselling other people's goods, the same table works with a lower margin. The point is the arithmetic: units times price is sales, units times cost is stock. The tutorial on planning how much to pack each week for a farmers market covers the weekly top-up logic for stalls, and the sell-through approach carries over directly.

A staffing rota built from opening hours

Build the rota from the opening hours and not the other way round. Each day needs one person from half an hour before opening to half an hour after closing. Add a second person at the busiest hours, and cover for breaks, because one person alone cannot have a lunch break without closing. Here is the week, with the owner on the floor three days:

Add it up. The owner works 27 hours, Staff A 25 and Staff B 20, which is 72 person-hours; 45 are paid. Over four weeks that is 180 paid hours, plus 12 for set-up and breakdown, so 192 hours at $17.50 is $3,360. Add the employer costs your adviser or payroll provider tells you about, and they will lift this. If you paid yourself the same rate for your 108 hours it would be another $1,890, and the base-case profit would fall from $5,354 to $3,464. Leave it out of the break-even, but look at the second number before you say the pop-up is "profitable".

The break cover is the line first drafts omit. Without the six hours a week of Staff B's lunchtime cover, the rota costs $420 less and looks tidier. It also leaves the shop locked at lunchtime four days a week, twice on the owner's days and twice on Staff A's. If you want help turning rotas into shifts and checking them against demand, the tutorial on AI shift scheduling versus a rota spreadsheet compares the two approaches.

Break-even in days, and what has to be true to reach it

The fixed costs of the run are the rent ($4,300), service charge ($400), insurance ($200), fixtures ($600), signage ($250), kit ($100), marketing ($450) and staff ($3,360): $9,660 in all. Everything that varies with sales comes off first, leaving a contribution. In the base case, gross profit is $15,941, card fees are $723 (2.6% of sales plus 15 cents on each of 582 sales at a $42 basket) and packaging is $204. The contribution is $15,014, which is $626 for each of the 24 trading days. Break-even days are fixed costs divided by that: $9,660 over $626 is 15.4 days, or 64% of the run.

Turn the break-even into something you can test. Sales needed to cover the fixed costs are $9,660 divided by the contribution rate of 61.4%, which is about $15,740, or $656 a trading day. At a $42 basket that is 16 customers a day. Ask the landlord or centre for daily footfall counts for the same weeks last year and divide: at 300 people a day you need about 5.2% of passers-by to buy, at 450 it is 3.5%, at 600 it is 2.6%. My own rule of thumb is to be wary of a plan that needs more than about 5% of passers-by to buy; the sums tell you the figure, and the landlord's counts tell you whether it is plausible.

Break-even days also hide the calendar. Weekends carry more sales than weekdays. Say Tuesday to Thursday sell about the same, Friday 1.4 times as much, Saturday 1.9 times and Sunday 1.5 times. Then the base case breaks even on the Saturday of week three, the 17th trading day, and the high case on the Sunday of week two. The low case does not break even at all. Whatever the case, the cash out before the first sale is $12,854 if the rent and the other non-wage costs are paid up front: $6,300 of those costs, the $1,000 deposit, $5,343.50 of opening stock and $210 for the 12 set-up and breakdown hours, all counted up front to be safe.

What moves break-even most

Change one thing at a time in the base case and the picture is clear:

A 10% swing in sales is worth about $1,500, more than the whole break cover and more than a 10% rent cut. The lesson is where to spend your worry: on the venue and its footfall first, then on rent, and only then on shaving shifts. Margin matters just as much. This plan assumes the brand makes its own goods at a 65% gross margin. A reseller with a 50% gross margin, the same sales and the same fixed costs would find break-even at 20.5 days, 85% of the run, with a profit of $1,643. The same plan goes from passing the two-thirds test set out below to failing it on the margin line alone, so put your real margin in before you believe any result.

For a worked break-even on a different business, the tutorial on break-even analysis with AI for a new product uses the same formula, and the 13-week cash flow forecast shows how to track the cash gap week by week once the pop-up is running.

Where ChatGPT's draft went wrong, and how to catch it

Give ChatGPT the fact block and ask for the model with the working visible:

Using only the POP-UP FACTS above, build:
1. A table of fixed costs and variable costs. The deposit is refundable: show it as cash out, not as a cost.
2. Contribution per trading day = gross profit minus card fees minus packaging, divided by 24.
3. Break-even trading days = fixed costs divided by contribution per trading day.
4. Three scenarios: low (50% of base units), base, high (130%).
5. Cash needed before the first sale.
Show every calculation on its own line in words and numbers so I can recompute it. Label every assumption you add.
Do not give a profit figure without the arithmetic behind it.

A realistic first draft contains mistakes that look tidy. This is what one might contain (illustrative), set against the hand figures:

Notice that the card-fee slip changes break-even by less than a tenth of a day, so it never shows in the headline figure; only the line-by-line check catches it. The deposit error moves the answer by 1.6 days, which you might notice. The same logic covers other common slips: dividing contribution by 28 occupied days instead of 24 trading days understates the daily figure, dropping break cover understates staff cost, and using revenue instead of gross profit makes break-even look quicker than it is.

Run five spot checks, each against a source ChatGPT did not produce: recompute the total rent from the quote; recompute total fixed costs from the list; recompute card fees from the transaction count; recompute paid staff hours times the rate; and recompute break-even with a calculator. Put ChatGPT's figure, your figure and the gap in three columns and do not move on while any gap is unexplained. If you prefer to build the check in a spreadsheet, ask ChatGPT for the formulas rather than the answers, so every total recalculates when the landlord revises the quote.

One more prompt earns its place before you sign: "List eight ways this plan could fail. For each, say how I could test it before signing and which number in the plan would change." Expect obvious answers such as low footfall and supplier delay, and the useful ones, such as a hidden service charge increase, a landlord's right to end the arrangement early, or a sales peak that falls before the shop opens. Each one is a question for the landlord or supplier, not a thing for ChatGPT to settle.

Go or no-go: the rules to fix before you sign

Decide the rules while you are still calm, not after the agreement has been signed. Here are mine for a first pop-up, using the figures above:

1. The base case must break even within about two-thirds of the trading days. Here it does, at 64%.
2. The low case must lose an amount you can afford without borrowing. Here it is $2,153, plus about $1,084 of stock at cost that comes back to you.
3. The conversion you need, against the landlord's own footfall counts, must be believable. About 16 customers a day at 300 passers-by is 5.2%, which I would test hard.
4. The cash out before opening, $12,854 here, must be money you have, and not money you hope the first week will return.
5. There must be a way to limit the damage if the run goes badly: shorter hours, fewer paid shifts, no top-up order.

The fifth rule becomes a stage gate. Base-case sales for week one are about $6,115. If week one sells under roughly $4,280, which is 70% of plan, cut the paid second-person hours and skip the top-up until the numbers recover. Write that rule in the plan, so that the decision in week two is already made. The same pattern, a short run, a fixed pitch and a hard look at the week-one numbers, is in the tutorial on a street food stall business plan with pitch fees, margins and stock.

What ChatGPT contributed to all this was speed and structure: the fact block turned into questions, the rent in three shapes, the rota arithmetic, the scenarios and the list of ways to fail. What you contributed was every number that mattered and every check that stopped a tidy mistake. That division is the plan.

Questions pop-up owners ask before they sign

Is the free ChatGPT plan enough for a pop-up plan?

The prompts here are text, tables and arithmetic, which a free plan can handle, but usage limits and features change, so check what your plan includes before you rely on it. Whatever the plan, do the final totals in a spreadsheet, because the arithmetic is where drafts go wrong.

How much stock should a pop-up buy before opening?

A workable start is the first two weeks of base-case sales plus 25%, with a top-up order planned for the second half. If your supplier cannot deliver inside a week, buy more up front. Display space also caps the buy: stock you cannot show does not sell.

Should I pay a fixed rent or a share of sales?

Compare them with your own scenarios, as in the rent table here. A share of sales moves risk to the landlord, so expect a minimum or a higher percentage in return. A hybrid of a smaller fixed sum plus a lower percentage is often where the talk ends.

Do I need a permit or insurance to trade in a short-let?

Landlords usually ask for evidence of liability insurance, and some places require a permit for temporary trading. The rules differ between places, so ask the landlord and a qualified adviser. ChatGPT can list the questions to ask, but it cannot tell you what applies to you.

Further reads

• Coffee Cart Business Plan: Costs, Pitch and Menu Built With ChatGPT — The same plan method for a small cart with daily costs.
• Hair Salon Business Plan: Chairs, Prices and Cash Flow With AI — Cash flow for a business with fixed rent and wages.
• Boutique Inventory Management: AI for Size Runs and Reorders — Reordering stock once the pop-up proves a line sells.
• Local SEO for a Retail Shop: Profile, Stock Posts and Reviews With AI — Get found while the pop-up is open.
• Party Supply Store Marketing Ideas: Seasonal Bundles With AI — Seasonal bundles for a short selling window.
• Best AI for Shop Design: Floor Plan and Display Brief for Your Fitter — A shop owner's guide to AI for a refresh: measure, draft a layout, mock up windows and signs, then hand the fitter a brief they can price and check.
• AI Tools and AI Development: The Complete 2026 Guide — the AI hub, including every tutorial in the AI-for-business series.

Sources: Square's pricing page (published in-person card rate on its Free plan, 2.6% plus 15 cents). Every other figure in this tutorial is illustrative and worked by hand; use your own rent quote, supplier costs and wage rates.