How to Start a Cafe in Vietnam: Complete 2026 Guide
From first idea to day 90: pick a customer and concept with a weighted scorecard, choose an operating model, design the bar workflow and set review points.

Two people in Ho Chi Minh City each have savings and a dream of running a cafe. One goes straight to property listings. The other first answers a single question: who will buy my coffee, at what hour, and why would they skip the place next door? The second person usually moves faster in the end. This guide follows the second approach. You will choose a customer and concept, pick an operating model, design the workflow behind the bar, decide which numbers to watch, and then walk through the first 90 days with a decision at each checkpoint. Detailed budgeting and paperwork deserve their own treatment and are only touched on here.
Any figures below are invented to show how a calculation works. They are not market data. Substitute what you measure yourself.
Start with a one-sentence position
Complete this sentence before you design anything: "My cafe serves [type of customer] at [time of day] who want [need], and it differs from nearby options by [one thing]." If you cannot name the difference, the concept is not ready.
- Customer: office workers, students, apartment residents, tourists, freelancers.
- Time: the commute, the lunch break, the late afternoon, the evening, the weekend.
- Need: a fast takeaway cup, a table to work at, a place to meet, speciality coffee, a photogenic corner.
- Difference: bean quality and brewing, speed, atmosphere, price, a signature drink, service style.
That sentence later dictates the location, opening hours, menu length, number of seats and shift staffing. Whenever you face a decision, test it against the sentence.
Choose the concept with a weighted scorecard
Do not pick a concept because you like it. Shortlist three, then score each against criteria that matter in your situation. The table is an illustration with assumed scores; define your own criteria and weights.
| Criterion | Weight | Grab-and-go | Work-friendly lounge | Speciality coffee bar |
|---|---|---|---|---|
| Fit with customers nearby | 3 | 4 | 2 | 2 |
| Capital needed versus what you have | 2 | 4 | 2 | 3 |
| Skills already in your team | 2 | 3 | 3 | 5 |
| Ease of copying to a second site | 1 | 5 | 2 | 3 |
| Direct competition nearby (higher is better) | 2 | 2 | 3 | 4 |
To compute the total, multiply each score by its weight and add. Grab-and-go: 12 + 8 + 6 + 5 + 4 = 35. Lounge: 6 + 4 + 6 + 2 + 6 = 24. Speciality bar: 6 + 6 + 10 + 3 + 8 = 33. The two leaders are close, which tells you to gather field evidence rather than trust the arithmetic. The scorecard exists to force your assumptions into the open, not to produce an automatic answer.
Choose an operating model
The concept answers what you sell and to whom. The model answers how selling is organised, and each model carries its own cost structure and failure mode.
| Model | Strength | Main risk | Fits when |
|---|---|---|---|
| Kiosk or takeaway window | Small footprint, few staff, easy to repeat | Depends on passing trade and rush-hour flow | You have a busy spot and want to test the market |
| Sit-down cafe | Higher spend per visit through longer stays and food | Heavy fixed rent and running costs | People will stay and the space is distinctive |
| Cafe plus online ordering | Reaches customers beyond the neighbourhood | Adds pressure to the peak hour | The counter process already runs smoothly |
| Cafe combined with another business | Several reasons to visit | Blurry brand, two things to manage | You know the second trade well |
| Franchise or branded partnership | Ready-made procedures and recognition | Reliance on the franchisor's terms and costs | You want speed and accept less control |
If this is your first venture and you have no real demand data, a small model with low fixed costs and easy adjustment buys you learning time. If you consider a franchise, read the terms on fees, supply obligations and exit conditions, and ask for operating results from existing outlets.
Test demand before the big commitment
Cheap experiments can tell you whether people will pay for your idea before you sign a long lease.
- Sit near a comparable cafe at three times of day and count walk-ins, takeaway buyers and people who stay more than an hour.
- Sell your signature drink at a weekend market, event or cart to test price and appeal.
- Interview fifteen to twenty people in your target group: where they drink coffee, what they pay, what would make them switch.
- Run recipe tastings with a small panel and collect feedback drink by drink, not just "was it good".
Summarise the result in a short note: who the customer really is, a believable price, the lead product, and the biggest unanswered risk.
Design operations around the bar
A lovely room with a jammed bar still loses customers. Draw the path of one order from the door to the cup and find where it bottlenecks.
- The guest orders and pays.
- The order reaches the person making drinks.
- The drink is made and handed over or served.
- The area is cleared and reset.
Assume a drink takes 90 seconds for one person. In an hour, one barista can make at most 3,600 / 90 = 40 drinks. If you expect 70 drinks in your peak hour, you need two people on the bar or a simpler menu that shortens preparation. Replace 90 seconds with a time you measure with a stopwatch, and you can size the peak shift from evidence instead of instinct.
Use the same path to decide the bar layout (preparation, cold storage and washing in the order of the movements), how many ordering points you need and where takeaway orders wait. Decide early how you will record sales and stock; the cafe software comparison lists evaluation criteria you can borrow before trialling tools with your own menu.
The numbers worth tracking from day one
A small cafe does not need a complicated dashboard. These five measures reveal most problems.
| Measure | Formula | What it tells you |
|---|---|---|
| Tickets per day | Number of receipts | Traffic by weekday and hour |
| Average ticket | Revenue divided by receipts | Effect of add-ons and pricing |
| Ingredient cost ratio | Ingredient cost divided by revenue | Waste or supplier price creep |
| Labour ratio | Labour cost divided by revenue | Whether shifts are sized correctly |
| Repeat rate | Customers with two or more purchases divided by identifiable customers | True health once the novelty fades |
Repeat rate needs a way to recognise customers, for example a phone number or membership card. If you plan a loyalty scheme, count its cost inside your margin instead of treating it as free.
The first 90 days: three phases, three decisions
Rather than staring at every day, split the period into three phases, each with an aim and one deciding question.
Days 1 to 30: stabilise
The aim is to do things correctly before doing more of them. Keep the menu short, time every order, log drink errors and wrong tickets. Spend ten minutes at close reviewing traffic, best sellers and waste. The question at the end of the month: does the peak-hour routine hold together, and which items should be dropped or re-specified?
Days 31 to 60: adjust
Use the first month's data to tune prices, menu and rosters. If one hour is always empty, narrow the shift or aim a promotion there. If another is overloaded, add a person or simplify drinks. Start recognising regulars and ask why they return. The question at the end of the month: are average ticket and ingredient ratio tracking the plan?
Days 61 to 90: judge and choose a direction
Compare reality to your original scenario: traffic, revenue, fixed costs and net result. Then pick among three paths: hold and refine, adjust the concept or trading hours, or prepare to expand with a new channel or site. The question: if current trends continue, how long does the reserve last, and what must change now?
| Checkpoint | Deciding question | Data you need |
|---|---|---|
| Day 30 | Is the peak-hour routine stable? | Service times, wrong orders, waste |
| Day 60 | Do price and menu deliver the planned margin? | Average ticket, ingredient ratio, sales by item |
| Day 90 | Hold, adjust or expand? | Net result, repeat rate, remaining reserve |
Build the team around the concept
A skilled barista does not guarantee a good cafe; they must understand your concept. For speciality coffee, they should explain flavour. For fast takeaway, they must stay calm under pressure. Define the behaviours you expect for each role and test them during a probation period with real situations.
- Write a short role description for each position: what to do, and what not to do.
- Train recipes, hygiene and complaint handling before opening.
- Reward quality and speed, not only sales volume.
- Keep a trained backup for peak shifts so the cafe does not hinge on one person.
Common mistakes at the start
- Starting from the room instead of the customer: a beautiful space nobody needs.
- A concept for everyone: a long menu and a blurry identity.
- Never timing the drinks: understaffed at the first rush.
- Watching revenue only: ignoring cost ratios and repeat rate.
- Reacting to each complaint: constant change with no data behind it.
- Expanding too early: copying an unstable routine only copies its faults.
Frequently asked questions
Should I choose the concept or the location first?
Start with the one-sentence position and the target customer, because they give you criteria for judging locations. If you already hold a good location, shape the concept around the people who pass it.
Should a first cafe be small or large?
Small and low in fixed cost gives you more months to learn before your reserve runs out. Go large only with evidence of steady traffic.
How many staff do I need per shift?
Measure preparation time, compute each person's hourly capacity, and compare it with expected peak-hour demand as in the example above. Adjust after the first few weeks of real data.
Should I offer delivery from the start?
Only when the counter process is stable, since online orders add load to the busiest hour. If you plan to, begin with a small delivery radius.
How should I weigh software costs?
Check the current pricing page and add up the yearly total, including fixed fees, any per-outlet or per-user charges and setup. Do not compare on monthly price alone.
Is there a software option that grows with several outlets?
Look at the page on multi-branch cafe management software to see which apps currently exist, then test any candidate against your own menu and workflow.
Next steps
- Write your one-sentence position and check it with at least ten people from the target group.
- Score three concepts with a weighted table of your own.
- Pick the operating model and time your drinks to size each shift.
- Fix the five tracking measures and choose the tool that records them from day one.
- Book three review sessions at days 30, 60 and 90, each answering a single deciding question.
A reminder: this is a general planning guide and does not replace professional legal, tax or financial advice. Registration, food safety and tax rules evolve, so verify them with the authorities or an accountant first.



