How Much Does It Cost to Open a Coffee Shop? Part 3: How Much Will Your Shop Make?

Part 3 of a 4-part series on the real cost of opening a coffee shop.
Start at Part 1: menu, equipment, and the space decision, or catch up on Part 2: startup costs and funding.

If you've been following this series, you might be wondering why a post about revenue is showing up in a series about what it costs to open a coffee shop.

Here's why. You've been asking the right question — "how much does it cost to open a coffee shop?" — but that's only half of a much bigger question. The complete question is: how much does it cost to open and stay open through the startup season and the first slow season?

Let me show you what I mean with a number.

Say you do the revenue and expense work in this post and the next one, and it shows you're going to lose $30,000 across December, January, and February — three slow months in a row. If that $30,000 isn't sitting in your bank account the day you open, you will not make it to spring. It doesn't matter how beautiful your buildout was or how good your coffee is. You'll run out of cash and close.

You need to consider that $30,000 part of your opening costs. It just doesn't show up on a contractor's invoice.

Now, I already know what some of you are thinking — because I've thought it myself. We're going to ramp up fast. We won't have those slow months. That won't be us. It's the entrepreneur's instinct: you don't get into this business without a healthy dose of optimism. But that same optimism is exactly what talks people out of keeping cash on hand, and I've watched it get owners into real trouble. Myself included.

So hear me on this one. Plan for the worst and hope your instincts are right. Maybe you never touch the reserve. Maybe the slow months never come. Great — then you've got a pile of cash to put toward your second location. What you don't want is to lose everything you built because you bet the whole thing on a good feeling.

Ultimately, you can't answer "how much does it cost to open" until you know two things: what you're going to bring in, and what your potential losses will be in the lean months. The gap between them is cash you need to have on day one — before you've sold a single cup.

That's what this post is about. Before you can know what to hold in reserve, you have to know what you'll bring in. So let's figure out what your shop can actually make.

Before You Estimate a Single Number, Go Stand in the Neighborhood

Most people make the same mistake here. They find demographics on paper — population, income, age — see that the numbers look good, and start dreaming about revenue.

But demographics on paper don't tell you the thing that actually matters: who is physically present during your business hours, and what are they doing there? Are they staying, or passing through? Is there a reason to stop, or is your corner just something people drive past on the way to somewhere else?

The same demographic profile can tell two completely opposite revenue stories depending on when those people are actually around and whether your spot is a destination or a drive-by. I know this from experience. I've lived both scenarios. Same city. Same brand. Same owner — me. But two locations, two completely different outcomes.

Location One: The Industrial Neighborhood That Worked

The original Stay Golden sat in an industrial, commercial part of town. On paper it didn't look like a coffee shop goldmine. But here's what the paper didn't show:

There were roughly 10,000 people working within a five-block radius during business hours. We were open 7am to 4pm — their entire workday. There weren't many food options nearby. We had a big space, a parking lot, a central location, easy highway access.

The result: 175 to 200 transactions a day. That's close to 2% of the daytime population walking through our doors — and most of them getting both a drink and food.

The homework was easy on that one. A big captive daytime crowd, not much competition, and an obvious reason to be a destination. Those numbers were always going to work. I just didn't fully appreciate why they worked until I opened a second location where none of those things were true.

Location Two: The Bedroom Community That Fooled Me

When we moved, the current Stay Golden went into a bedroom community. And on paper? It looked better than the old spot. Densely populated. Young families. Good incomes. An up-and-coming neighborhood everyone was excited about.

What I didn't account for was this: most of those people work somewhere else. By the time my doors open at 7am, the neighborhood has already emptied out for the day. All that great density gets in a car and drives to another part of the city.

So the numbers split hard. Weekends — when everyone's home — we do 120 to 150 transactions, high food attachment, strong average ticket. Beautiful. But Monday through Thursday? Sometimes 40 transactions. That is genuinely unsustainable for any shop.

I didn't go stand in that neighborhood during my own business hours before I signed. I looked at the density and the demographics and the energy of an up-and-coming area, and I filled in the rest with optimism — the exact thing I just warned you about. If I'd sat on that block at 8am on a Tuesday, I'd have seen it with my own eyes: everybody's gone. Nobody's stopping. It's a pass-through, not a destination, five days a week.

Same city. Same concept. Same operator. Two completely different revenue stories — because they were two completely different markets.

The Lesson: Go See It Yourself

So before you estimate a single revenue number, you go to the neighborhood during the exact hours you plan to be open. Not on a lively Saturday afternoon. On a Tuesday at 8am, if that's when you'll have the lights on.

You're looking for the answers demographics can't give you: Are people stopping, or passing through? Is there a reason to linger, or is this a corridor? Is there parking — or does the hassle of street parking quietly send the car-dependent crowd somewhere easier? Those questions decide your revenue far more than the income data on a Census page.

Your Neighbors Tell You What to Sell

One more thing the neighborhood tells you: what food to put on your menu.

The current Stay Golden sits right next to a butcher shop with an outstanding lunch counter — beef, burgers, Cubans, roast beef. What they don't do is breakfast. No chicken, no salads, no lighter options. So our lunch menu leans into exactly what they don't offer. We're not competing with the neighbor — we're completing the block. Somebody wants a burger, they go next door. Somebody wants a lighter lunch or a salad, they come to us. Everybody wins.

You can only make that kind of call if you know what's already around you.

Let the Neighborhood Set Your Hours, Too

The same logic runs through your hours of operation. A working neighborhood empties out after 3:30pm — so a happy hour program is dead on arrival, because your guests are already gone. A residential neighborhood does the opposite: it fills back up after 5pm, and a happy hour could be the busiest, most profitable stretch of your day.

Your hours should follow the rhythm of your market — not some generic assumption about when coffee shops are "supposed" to be open. The neighborhood is already telling you when your people are around. Your job is to listen and match it.

The Three-Step Research Process

Now Turn What You Saw Into a Number

Standing in the neighborhood tells you whether a location has a pulse. But "it felt busy" isn't a revenue projection you can take to a bank. So here's how you turn what you saw into an actual transaction estimate you can build on. There are three steps in this process, and each one covers for the others' blind spots.

Step 1: Walk It Yourself

This is the one we just talked about, so I'll keep it short — but don't skip it, because it's the foundation for everything after.

Go to the neighborhood during the exact hours you plan to be open. Sit in a nearby business. Watch, and actually count. How many people walk by? Are they stopping or passing through? Where does the foot traffic concentrate — is there a natural gathering point, a corner or a block where people linger?

Then talk to people who live or work there. Where do you get coffee? Where do you eat? What does this neighborhood need? You'll learn more in three of those conversations than in an hour of staring at demographic charts.

This is your gut check. It's irreplaceable — and it's also not enough on its own, because your gut can be fooled by a lively afternoon the same way mine was fooled by good demographics. That's what Step 2 is for.

Step 2: Run the AI Location Analysis

There's a prompt you can run in any AI tool that has web browsing — I'll give you the whole thing in a second. It pulls public data (Census figures, municipal reports, transit ridership, Google Maps competitor info) and turns it into a data-backed transaction estimate for a specific address. It's the counterweight to your gut: cold numbers to check your instinct against.

I tested it on Stay Golden's actual address — 904 51st Ave N, in Nashville's Nations neighborhood — and the result was honestly a little humbling. The first-pass moderate estimate came back at about 50 transactions per day on weekdays — uncomfortably close to my real Monday-through-Thursday numbers. If I'd run this prompt and believed that number before I signed, it might have saved me from a hard couple of years.

So here's the prompt. Copy it, swap in your address, and run it:

Act as a commercial real estate location intelligence analyst specializing in food and beverage retail. I am a first-time cafe owner trying to build a foot-traffic and customer estimation model for a potential location at [INSERT ADDRESS/NEIGHBORHOOD]. I do not have access to paid platforms. Please perform a spatial interaction analysis for this location by executing the following steps: (1) SOURCE DATA ESTIMATION: Use your web browsing capability to search for public data. Look for residential population density within a 5-to-10-minute walk, daytime population influx via municipal or economic development reports, and public transit ridership volumes. (2) ASSUMPTIONS AND FILTERING: Based on local demographics, estimate the Coffee Affinity Group percentage and apply standard spatial decay rules. (3) ESTIMATION CALCULATIONS: Provide three scenarios — Conservative, Moderate, Optimistic — using: Estimated Daily Customers = (Total Daytime Commuters × Capture Rate A) + (Local Residents × Capture Rate B) + (Pedestrian Transit Passersby × Capture Rate C). Use realistic industry-standard cafe capture rates of 0.5% to 2.5% of total foot traffic. (4) PEER COMPETITION ADJUSTMENT: Identify the nearest major coffee competitors within a 3-block radius using maps. Adjust capture estimates down based on market dilution. Present the final daily and weekly transaction estimates for each of the three scenarios.

A few tips to get the most out of it:

• Tell it what's nearby. Name the major landmarks, big employers, and anchor businesses around your spot — it sharpens the estimate.
• Ask it to cite sources so you can check the logic instead of trusting a black box.
• Feed it local documents. If your city council or chamber of commerce has a free neighborhood profile PDF online, paste the link right into the chat.
• Trust the first-pass number most. This one matters: as you feed it more and more detail, the estimates tend to drift optimistic — it starts weighting your space's capacity and best-case potential over where people actually are on a Tuesday. The honest number usually shows up first, before you've "helped" it.

Step 3: Cross-Check and Build Conviction

Now you've got two pictures: what your gut told you from standing on the block, and what the data told you from the prompt. Put them side by side.

If they land in the same ballpark — good. You've got conviction. Use the moderate scenario as your planning number and move forward.

If they diverge hard — stop, and figure out why before you sign anything. Either your gut is missing something the data caught, or the data is missing something only the neighborhood itself can show you. In my case, the walk and the honest first-pass number were both telling the same story — I just wasn't listening to either one closely enough.

Neither the walk nor the data is enough by itself. Together, they're the clearest possible picture you can get before you commit years of your life and a pile of money to a lease.

Translating Transactions Into Revenue

From "How Many People" to "How Much Money"

Now you've got a transaction count you believe in. That's the hard part, and it's done. But a transaction count isn't revenue yet — a shop doing 100 transactions a day at a $6 average ticket is a very different business from one doing 100 at $14. To get from bodies-through-the-door to dollars, you need two more things: your product mix and your average prices.

The Industry-Standard Method (and How to Improve It)

There's a standard way the industry does this. It's five steps, and it works — I'll walk you through the whole thing. Then I'll show you the one change that makes it dramatically more accurate.

Step 1 — Set your categories and their prices. Break your menu into categories and assign an average price to each. You can slice this however you want, but let me tell you how I do it in my own model, because the logic matters. I use six:

• Filter coffee
• Espresso drinks
• Non-coffee grab-and-go
• Pastry
• Food entrée
• Retail

Why so granular? Because averaging hides money. On my P&L, "food" is one line — pastries, entrées, sides, all of it. But an entrée might average $14.50 and a pastry $4. Lump those together and you get a blended "food" price that's true for nothing on your menu — and it quietly drags your whole estimate off. Same with drinks: a filter coffee and an espresso drink aren't the same price, and grab-and-go is different again. Splitting them keeps each price honest. (There's a second reason I split them — filter coffee and a grab-and-go drink can cost the same on the menu but have wildly different costs to make. That's a Part 4 problem, but it's worth splitting them now so you're already thinking that way.)

So set your categories, and pull an average price for each. You can do this from the competitor menus you already looked at during your neighborhood research — or, if you've done the menu work from Part 1, set a market-competitive price for each of your own menu items and average the prices within each category together.

Step 2 — Estimate your P-Mix. Your P-Mix is the percentage of total items sold that fall into each category. Not dollars — items. If you sell 100 items in a day and 45 of them are espresso drinks, espresso is 45% of your mix. All your categories together add up to 100%.

A typical small-menu shop might look roughly like: espresso 45%, cold drinks 30%, food 20%, retail 5%. But here's the problem — that's a guess. You're pulling those percentages out of the air before you've sold a thing. And this guess is dangerous, because it's about to flow into every number downstream.

Step 3 — Blend your item price. Multiply each category's price by its P-Mix percentage, add them all up, and you get your blended item price — the average price of any single item crossing your counter. For example:

($5.50 × 0.45) + ($3.50 × 0.30) + ($11.00 × 0.20) + ($4.00 × 0.05) = $5.72

Step 4 — Apply IPT to get your average ticket. People rarely buy exactly one item. Items Per Transaction (IPT) captures that — the industry rule of thumb is 1.3 to 1.5. Multiply your blended item price by your IPT:

$5.72 × 1.4 = $8.01 average ticket

Step 5 — Get to daily revenue. Multiply your average ticket by your estimated daily transactions:

$8.01 × 100 transactions = $801 per day

That's the method. And it's fine. But notice what just happened: two of those five steps were guesses — your P-Mix in Step 2, and your IPT in Step 4. Everything else was real. Your prices came from actual menus. Your transaction count came from the research in the last section. But the P-Mix and the IPT? You made those up. And they're multiplied into everything. If your real P-Mix skews more toward cheap filter coffee than you assumed, or your real IPT is 1.2 instead of 1.4, your $801 day might actually be a $650 day — and you won't find out until you're already open and short.

The Fix: Stop Guessing, Start Counting

Here's the change that turns this from an educated guess into something you can actually trust. And you're already halfway to doing it.

Remember how you're visiting local coffee shops during your neighborhood research? While you're sitting there, count items. Not just "did that person buy food, yes or no" — every single item, by category, for every customer you watch. Do this for a minimum of 25 transactions at each of four shops — 100 transactions total.

That tally hands you both of your guessed numbers as observed numbers:

• Your real IPT: total items you counted ÷ 100 transactions. If you counted 140 items across 100 transactions, your IPT is 1.4 — measured, not assumed.
• Your real P-Mix: each category's item count ÷ total items counted. If 50 of those 140 items were espresso drinks, espresso is 36% of your mix — measured, not assumed.

Then you run the exact same five-step method — but now Steps 2 and 4 aren't guesses anymore. They're real data from your actual market. Same formula, honest inputs.

Here's what that looks like fully worked. Say you sit in four shops in your target neighborhood and tally 100 transactions. You count 138 total items, broken down like this:

• Espresso drinks: 52 items
• Filter coffee: 30 items
• Food entrée: 24 items
• Pastry: 18 items
• Non-coffee grab-and-go: 10 items
• Retail: 4 items

Your observed IPT: 138 ÷ 100 = 1.38.

Your observed P-Mix: espresso 38%, filter 22%, entrée 17%, pastry 13%, grab-and-go 7%, retail 3%.

Now bring in the prices — pulled from those same shops' menus, or from your own menu if you've set it — say espresso $5.50, filter $3.50, entrée $12, pastry $4.25, grab-and-go $5, retail $16. Blend them:

($5.50 × .38) + ($3.50 × .22) + ($12 × .17) + ($4.25 × .13) + ($5 × .07) + ($16 × .03) = $6.16 blended item price.

Average ticket: $6.16 × 1.38 = $8.50.

And if your research pointed to 90 transactions on an average day: $8.50 × 90 = $765 a day.

Every number in that chain traces back to something you watched with your own eyes or read off a real menu. Nothing was invented. That's a revenue estimate you can build a plan on — and defend to a banker.

For Context: A Real Cafe's Product Mix

Here's what the mix looks like at an established specialty cafe, so you have a benchmark to sanity-check your own tallies against — not a target to hit:

• Filter coffee: 31–37% of transactions
• Espresso drinks: 42–63%
• Non-coffee grab-and-go: 12–21%
• Pastry: 1–14%
• Food entrée: 58–100% (peaks on weekends)
• Retail: 1–12%

Notice how wide some of those ranges are — that's the difference between a slow Tuesday and a packed Saturday. Your mix isn't one fixed number; it's a range that moves with your week.

Where the Model Comes In

Now, that's the industry standard. It's real math, and you can absolutely do it by hand — I just did it above, and you can too. A notebook, an afternoon in four coffee shops, and a calculator will get you there.

But it's fiddly. Six categories, a hundred tally marks, a blended average, a couple of multipliers — and if you fat-finger one number, the error rides all the way through to your daily total without announcing itself. When the output is the number you're using to justify a loan or a lease, "I think I did the arithmetic right" is a shaky place to stand.

That's exactly why I built an automated, complete financial model. You drop in your observed mix, your prices, and your transaction estimate, and it runs all five steps instantly — blended price, average ticket, daily revenue — then carries it out to monthly and annual, and layers in your seasonal swings automatically. No arithmetic to fumble. It's the difference between hoping your model is right and knowing the math is clean so you can argue about the assumptions, which is where your attention actually belongs.

The model is part of the How to Open a Coffee Shop Masterclass I'm building. If you want it when it's ready — along with the full system for pressure-testing a location before you sign — get on the waitlist here.

From a Daily Number to a Yearly Picture

One Day Isn't Your Week

Here's where your projections can go really sideways if you aren't careful. Often folks take one daily number and multiply it across the whole week, as if every day looks the same. It doesn't. Not even close.

Almost everywhere, weekends run heavier than weekdays — and it's not a small gap. Let me use my own numbers. On a slow Tuesday, the current Stay Golden might do 45 to 55 transactions. On a Saturday, we'll do 150. And it's not just more people — weekend guests buy more per visit, so the average ticket climbs too. A Tuesday might bring in around $650. A Saturday can do $4,500. Same shop, same week.

If I'd taken that $650 Tuesday and multiplied it across seven days and fifty-two weeks, I'd have built my entire plan on a number that only represents a slice of my actual year. It would've been wrong by a mile.

So you don't scale a day. You build one accurate week, then scale that.

Build One Real Week

You've already done half this work. In the last section, you sat in four local shops on a weekday and tallied transactions and items. Now go back — same shops, same time window — on a Saturday. And if you'll be open Sundays, do a Sunday too.

You're measuring two things:

• How much busier is the weekend? If a shop did 50 transactions when you sat there Tuesday and 150 on Saturday, that's roughly 3× the traffic. That's your weekend transaction multiplier.
• Does the average ticket climb? Run the same item tally you did before. Weekend guests often buy more per visit — a higher IPT, a richer mix — which pushes the average ticket up on top of the higher traffic. Capture that too.

Now build your week with real shape to it:

• Weekday revenue × your number of weekdays (Monday through Friday tend to run close enough to treat as one figure for planning — though many shops find Friday creeps up a bit).
• Plus Saturday at its higher transaction count and higher average ticket.
• Plus Sunday, if you're open.

Add those together and you have one honest week — not a fantasy built from your best day, not a doom-spiral built from your worst. Multiply that week out to get an average month.

Then Layer On Your Seasons

An average month is still just an average. Some months will run well above it, some well below — and you already know something about which is which, because your neighborhood research told you your market's rhythm.

• A working neighborhood might empty out over the holidays, when offices close for two weeks.
• A bedroom community — like my spot — might go quiet in the dead of winter, when nobody's wandering the neighborhood, then surge in spring and summer.

Take your average month and adjust it up for your strong stretches and down for your slow ones. What you'll end up with isn't a single number — it's twelve of them, a month-by-month revenue picture with the peaks and the valleys drawn in honestly.

Sit With Your Slow Months — You'll Need Them Soon

Look hard at the low points. The thin months. The ones where the number makes you a little uncomfortable.

Don't do anything with them yet — because right now they're only half the story. A low revenue month isn't a loss until you subtract what it costs to operate that month: your coffee and food costs, your labor, your rent, everything it takes to keep the lights on. You don't have those numbers yet. That's the next post.

But hold onto these figures. Because when we lay your expenses over this revenue picture in Part 4 and build a real month-by-month profit and loss statement, those slow months are exactly where the answer to this whole series lives — how much cash you need in the bank on opening day to survive the gap before the good months arrive.

You've now got the first half of that equation: what you'll bring in, month by month, all year. Time to figure out what you get to keep.

What Comes Next: Turning Revenue Into an Answer

Here's what happens in the next post. You take the revenue numbers you just built, and you subtract all of your operating expenses:

• Cost of goods — for a well-run specialty cafe, typically 25 to 30% of revenue
• Labor and payroll
• Rent and space costs
• Everything else it takes to keep the lights on — insurance, utilities, supplies, loan payments

When you do this for a single month, you've got a monthly profit and loss statement. But you can also run it across the whole year, month by month.

And that's the version that matters — because seeing your annual P&L broken out month by month exposes the stretches where you might be operating at a deficit. Those are the months that decide whether you survive your first year, and you cannot see them on an annual average.

Knowing those numbers is what finally answers the two questions this series has been building toward. The one you started with — how much does it cost to open a coffee shop. And the one that matters even more:

How much of this revenue do you get to keep — and will this investment actually pay off?

Let's Make This a Whole Lot Easier

You've now built the first half of the answer — a real, month-by-month sense of what your shop can bring in. The second half is knowing what you keep. That's the whole game, and it's what the How to Open a Coffee Shop Masterclass is built to walk you through — including the financial model that runs every calculation in this post for you, from product mix to seasonal projections.

Get on the waitlist for instant access to Café Confidential — my weekly playbook for building and running a coffee shop that actually works — and you'll be first in line when the masterclass launches.