Skip to main content
RADA Richness Co., Ltd. logo
AI WorkshopNEW About us
Services
Courses
Work
Knowledge Hub
Careers ContactChat with us on LINE
Free marketing consult
Starting a business · Startup with AI

Lean Canvas: Plan a New Business Model on One Page, Starting from the Customer's Problem, with a Prompt Guideline for AI

Have a new business idea but not sure customers really want it? Lean Canvas helps you put the whole idea on a single sheet, starting from the customer's problem, then shows which assumptions to test first. This article explains what data to prepare for AI and gives you prompts to copy straight into ChatGPT, Gemini or Claude.

RADA Richness brand character Creative team · RADA Richness21 November 2026 · 11 min read

What is a Lean Canvas

Lean Canvas is a 9-box business model sheet that Ash Maurya adapted from the Business Model Canvas in 2010 for startups and new businesses. The key difference is that it starts from the customer's problem instead of what we want to build. It is quick to draft, and every box is treated as an assumption to prove with real customers. Ash Maurya explains how to use it in his book Running Lean.

What a Lean Canvas looks like

The layout matches the BMC, but the 4 boxes in orange are different. They force you to think about the problem, the solution, the numbers to track and your advantage. Tap any box to jump to its explanation and prompt.

Click Copy prompt then paste it into ChatGPT, Gemini or Claude and replace the parts in [square brackets] with your business details. Prompts are written in English so the AI follows instructions precisely, but they tell it to answer in Thai, and you can type your data in Thai

Prompt Guideline: Data to Prepare for AI to Build Your Lean Canvas

If you only tell AI "make a Lean Canvas for my dessert shop", you get a sheet that looks complete but is pure guesswork. Prepare these 8 kinds of data first. The numbers in the last column show which boxes each item feeds (in the recommended order).

Data to prepareExamples of what to includeWhere to find itUsed in box
Your idea and where it came fromWhat the idea is and where it came from, for example a problem you faced, a customer request or a gap you spottedWhere to find itYourself, your idea notes34
Who you think your first customers areJob, behaviour, location, and who pays versus who usesWhere to find itCustomer chats, page follower data, people you know1
The problem in customers' own wordsSentences customers said or typed about the problem, not what we assume for themWhere to find itInterviews with 5 to 10 customers, chats, reviews23
Alternatives customers use today, and their pricesHow customers solve the problem now, what they pay and what frustrates them, with the month collectedWhere to find itInterviews, other shops' pages and apps236
What you can do with current resourcesSkills, hours per week, equipment, helpersWhere to find itYou and your team459
Channels your customers useWhere your first customers are and how they buy this kind of productWhere to find itInterviews, follower data, sales by channel5
Real costs and available moneyUnit cost, running costs, money you can useWhere to find itReceipts, quotations, accounts67
Numbers you can track todayOrders, new customers, repeat customers, ad results, with the time periodWhere to find itLINE OA Manager, Meta Ads Manager, sales log8

6 rules before you paste data into AI

  1. 01Start from customers' own wordsPaste what customers actually said or typed about the problem, not what we guess they think, so AI does not analyse our beliefs.
  2. 02Tell AI every box is an assumptionA first Lean Canvas is a set of guesses. Ask AI to separate clearly what has evidence and what still needs testing.
  3. 03Do not let AI guess numbersPrices, costs and market size must come from real data. If you do not have it, ask AI what data to go and find.
  4. 04Add a time period to your dataCompetitor prices, sales and ad results need the month they were collected so AI does not compare different periods.
  5. 05Remove personal dataRemove customer names, phone numbers and addresses before pasting chats or interview notes into AI.
  6. 06Work in versionsName it Plan A with a date, then ask AI to update it to a new version every time you learn something from real customers.

No data yet? Let AI interview you first

AI asks one question at a time, starting with where your idea came from, then customers and problems, and tells you how to find any data you are missing.

AI PROMPT
Act as a lean startup coach using Ash Maurya's Lean Canvas (Customer Segments, Problem, Unique Value Proposition, Solution, Channels, Revenue Streams, Cost Structure, Key Metrics, Unfair Advantage).
Interview me ONE question at a time, in Thai, starting with the story of where my idea came from, then customers and problems. If I do not know an answer, tell me how to find it cheaply (for example by interviewing 5 customers) and move on. When done, summarise my answers under the 9 boxes, mark each line as evidence or assumption, and list what is missing. Do not give recommendations yet.

Have the data? Draft the whole Lean Canvas

Get a full Plan A across all 9 boxes in the recommended order, with evidence separated from assumptions.

AI PROMPT
Act as a lean startup coach. Draft my Lean Canvas (Plan A) using ONLY the data below.

Idea and where it came from: [PASTE]
Early adopters I have in mind: [PASTE]
Problems in customers' own words: [PASTE]
Existing alternatives and what customers pay (with months): [PASTE]
My skills, time and team: [PASTE]
Channels my customers use: [PASTE]
Real costs and money available: [PASTE]
Numbers I can track today: [PASTE]

Fill the 9 boxes in this order: Customer Segments, Problem (with existing alternatives), Unique Value Proposition, Solution, Channels, Revenue Streams, Cost Structure, Key Metrics, Unfair Advantage. Keep each box to 1 to 3 short lines. Leave Unfair Advantage empty if nothing real exists yet.
Answer in Thai, in short bullet points a beginner can act on. For every point, say which of my data it comes from, mark anything that is a guess or an assumption I still need to test, and never invent numbers or prices.

Which block first

Ash Maurya says there is no single right order. Start with the box you understand best. If you are not sure where to start, use the order he recommends, which always begins with customers and problems.

  1. Customer Segments (Customer Segments) Who has this problem, and who will try first (Early Adopters)
  2. Problem (Problem) The top 1 to 3 problems, and how customers solve them today
  3. Unique Value Proposition (Unique Value Proposition) One sentence on how you are different and worth noticing
  4. Solution (Solution) The simplest fix for each problem
  5. Channels (Channels) How customers will find you
  6. Revenue Streams (Revenue Streams) How much customers pay, and how
  7. Cost Structure (Cost Structure) How much money it takes to test and start selling
  8. Key Metrics (Key Metrics) 3 to 5 numbers that show the business is on track
  9. Unfair Advantage (Unfair Advantage) Something competitors cannot easily buy or copy

The 9 boxes of the Lean Canvas

Each box has guiding questions, an example and a prompt for AI to help draft it. The example is a homemade Thai dessert shop selling online through LINE OA, the same fictional shop used in the BMC, SWOT and 3Cs articles. This time it is exploring a new idea: meeting snack sets for companies.

1. Customer SegmentsCustomer Segments

Like the BMC, start with customers, but Lean Canvas separates "the whole market" from "the first customers who will try you" (Early Adopters). These people feel the problem strongly, are already looking for a fix and give honest feedback. If the person who pays and the person who uses are different, list them separately.

  • Who pays, and who uses?
  • Which group feels this problem most and is already looking for a fix?
  • If you had to find your first 10 customers tomorrow, where would you look?
Example: Whole market: office workers who buy Thai desserts for themselves or as gifts · Early Adopters: admin teams at companies in office districts who must arrange meeting snacks every week
AI PROMPT
Help me fill the CUSTOMER SEGMENTS box of my Lean Canvas.
Business idea: [YOUR BUSINESS IDEA]
Target customer: [WHO]
Location / market: [CITY, COUNTRY]
What I know about customers so far (chats, interviews, sales): [PASTE]
Separate the whole market from early adopters. Describe 2 possible early-adopter groups: who they are, why the problem is urgent for them, and where I can meet 10 of them this month. Recommend one group to start with.
Answer in Thai, in short bullet points a beginner can act on. For every point, say which of my data it comes from, mark anything that is a guess or an assumption I still need to test, and never invent numbers or prices.

2. ProblemProblem

The most important box on the Lean Canvas. Write only the customer's top 1 to 3 problems, and under them list the "existing alternatives" customers use today, because almost every problem already has some fix, even if it is not a direct competitor. If you cannot say how customers solve it today, you do not really know the problem yet.

  • Which problem do customers bring up themselves without being led?
  • How do customers solve it today, and how much money or time does it cost them?
  • How much does it hurt if the problem is not solved?
Example: 1) Must arrange meeting snacks every week and are tired of the same menu 2) Most shops are awkward to order from as a company and cannot issue receipts 3) Snacks arrive stale · Alternatives today: buying from the bakery near the office, or ordering shop by shop on delivery apps
AI PROMPT
Help me fill the PROBLEM box of my Lean Canvas.
Business idea: [YOUR BUSINESS IDEA]
Target customer: [WHO]
Location / market: [CITY, COUNTRY]
Problems customers told me in their own words (chats, interviews, reviews): [PASTE]
Pick the top 1 to 3 problems, rank them by how painful and frequent they seem, and list the existing alternatives customers use today for each. Tell me which problem has the weakest evidence and 5 interview questions to check it without leading the customer.
Answer in Thai, in short bullet points a beginner can act on. For every point, say which of my data it comes from, mark anything that is a guess or an assumption I still need to test, and never invent numbers or prices.

3. Unique Value PropositionUnique Value Proposition

The box in the centre: one sentence on how you differ from the alternatives and why customers should care. Ash Maurya suggests tying it to problem number 1, keeping it short enough for a website headline or ad headline, and talking about the outcome customers get, not features. You can add a short High-level Concept under it.

  • Does this sentence answer problem number 1 directly?
  • Does it say what you sell and who it is for?
  • Placed next to a competitor, would customers see the difference at once?
Example: "Fresh, less-sweet Thai dessert meeting snacks. Order on LINE by 4 pm, delivered to your meeting room the next morning, with a receipt"
AI PROMPT
Help me write the UNIQUE VALUE PROPOSITION for my Lean Canvas.
Business idea: [YOUR BUSINESS IDEA]
Target customer: [WHO]
Location / market: [CITY, COUNTRY]
My top problem: [PASTE]
Existing alternatives: [PASTE]
Write 5 headline versions (max 120 characters each) that focus on the customer's outcome, not features, and say what it is and who it is for. Add one sub-headline for the best version and a short high-level concept. Explain how each version differs from the alternatives.
Answer in Thai, in short bullet points a beginner can act on. For every point, say which of my data it comes from, mark anything that is a guess or an assumption I still need to test, and never invent numbers or prices.

4. SolutionSolution

Ash Maurya made this box small on purpose, because founders tend to fall in love with their solution too early. Write the simplest fix for each problem, one line each, without detail yet. Once you talk to real customers, problems often get reprioritised and the solution changes with them.

  • Which problem does each solution match?
  • What is the simplest, cheapest version you can test?
  • Which parts do not need to be built yet?
Example: 1) 3 meeting snack set menus that rotate weekly 2) Order through a LINE OA form that can issue receipts 3) Made fresh on the morning of delivery, delivered in a morning round
AI PROMPT
Help me fill the SOLUTION box of my Lean Canvas.
Business idea: [YOUR BUSINESS IDEA]
Target customer: [WHO]
Location / market: [CITY, COUNTRY]
My top 3 problems: [PASTE]
What I can do with my current skills and budget: [PASTE]
For each problem, suggest the simplest solution I could test within 2 weeks, in one line each. List features I should NOT build yet and why.
Answer in Thai, in short bullet points a beginner can act on. For every point, say which of my data it comes from, mark anything that is a guess or an assumption I still need to test, and never invent numbers or prices.

5. ChannelsChannels

Channels are the paths that lead you to customers. At first, any channel that gets you talking to real customers is fine, such as reaching out yourself or asking people you know for referrals. But think about channels that can scale from day one, because the channel that brings your first 10 customers may not get you to 1,000.

  • Which channel gets you to Early Adopters fastest this month?
  • Which channel can scale as the business grows?
  • How many channels can your current team handle?
Example: Testing phase: message admin contacts you know on LINE and LinkedIn · Channels that can scale: LINE OA, Facebook ads targeting admin staff, referrals between companies
AI PROMPT
Help me fill the CHANNELS box of my Lean Canvas.
Business idea: [YOUR BUSINESS IDEA]
Target customer: [WHO]
Location / market: [CITY, COUNTRY]
Where my early adopters spend time and how they buy today: [PASTE]
My team size and weekly hours: [PASTE]
Separate channels to reach the first 10 customers from channels that can scale later. For each, give effort and what to measure. Do not estimate costs or conversion rates: tell me how to find the real numbers.
Answer in Thai, in short bullet points a beginner can act on. For every point, say which of my data it comes from, mark anything that is a guess or an assumption I still need to test, and never invent numbers or prices.

6. Revenue StreamsRevenue Streams

Do not leave pricing for later. Price is part of the product, signals your brand position and decides which customers you attract. A customer willing to pay is the best first proof that the problem is real. State what customers pay for and how, and try a price from the testing phase.

  • How much do customers pay for the alternatives they use today?
  • Will you charge once, weekly or as a package?
  • How soon can you ask customers to pay or leave a deposit?
Example: Price per set per person, minimum per order, and a monthly package for companies that order every week. Real prices are set from the real costs you calculated.
AI PROMPT
Help me fill the REVENUE STREAMS box of my Lean Canvas.
Business idea: [YOUR BUSINESS IDEA]
Target customer: [WHO]
Location / market: [CITY, COUNTRY]
What customers pay for existing alternatives today (with source and month): [PASTE]
My real unit costs: [PASTE]
Suggest 2 to 3 ways to charge (per order, package, subscription) and how to ask for payment or a deposit during the test. Use ONLY my numbers; if data is missing, tell me what to collect.
Answer in Thai, in short bullet points a beginner can act on. For every point, say which of my data it comes from, mark anything that is a guess or an assumption I still need to test, and never invent numbers or prices.

7. Cost StructureCost Structure

No need for a 3 to 5 year forecast. Think short term first: how much money it takes to test, build and launch your first version (MVP) in the next 3 to 6 months, and what your monthly running costs are. Use real numbers from quotations or inquiries.

  • How much money do you need to test with your first 10 customers?
  • What are your monthly running costs?
  • How many months will your money last?
Example: Testing phase: ingredients, boxes, delivery and a small ad budget · Monthly: kitchen rent, wages, LINE OA fees. Every number comes from real receipts.
AI PROMPT
Help me fill the COST STRUCTURE box of my Lean Canvas.
Business idea: [YOUR BUSINESS IDEA]
Target customer: [WHO]
Location / market: [CITY, COUNTRY]
My planned test for the next 3 months: [DESCRIBE]
List the one-time costs to define, build and launch my first version, and the monthly running costs. Do not invent amounts: give me a table with empty amount columns, where to get each real number, and how to calculate how many months my money will last.
Answer in Thai, in short bullet points a beginner can act on. For every point, say which of my data it comes from, mark anything that is a guess or an assumption I still need to test, and never invent numbers or prices.

8. Key MetricsKey Metrics

Pick only 3 to 5 numbers that show the business model is on track. Favour numbers that measure what customers do (outcomes) over how much work you did, and favour leading numbers such as customers trying you or repeat order rate, which show the trend before month-end sales.

  • Which number shows fastest that customers see the value?
  • Which numbers only look good, like likes, without telling you about sales?
  • Where will you collect these numbers, and how often?
Example: Companies ordering for the first time, companies reordering within 2 weeks, average order size, and companies that came through referrals
AI PROMPT
Help me choose KEY METRICS for my Lean Canvas.
Business idea: [YOUR BUSINESS IDEA]
Target customer: [WHO]
Location / market: [CITY, COUNTRY]
My Lean Canvas so far: [PASTE]
Data I can track today (LINE OA, sales records, ads): [PASTE]
Suggest 3 to 5 outcome-based, leading metrics. For each: why it matters, how to calculate it, where the data comes from, and how often to check. Flag any vanity metric I am currently watching.
Answer in Thai, in short bullet points a beginner can act on. For every point, say which of my data it comes from, mark anything that is a guess or an assumption I still need to test, and never invent numbers or prices.

9. Unfair AdvantageUnfair Advantage

An advantage competitors cannot easily buy or copy, such as special data or know-how, a customer community, unique relationships or a long-built reputation. If a competitor with enough money or time could copy it, it does not count. Most new businesses do not have one yet, so leaving it empty is better than writing something untrue.

  • If a better-funded competitor started doing the same thing tomorrow, what could they still not do?
  • What knowledge, relationships or data do you have that others do not?
  • What gets stronger as you gain more customers?
Example: A family recipe whose taste regular customers recognise, and a base of companies that refer each other in the same district (still weak at first and needs building)
AI PROMPT
Help me think about the UNFAIR ADVANTAGE box of my Lean Canvas.
Business idea: [YOUR BUSINESS IDEA]
Target customer: [WHO]
Location / market: [CITY, COUNTRY]
What I have that others may not (skills, relationships, data, community, recipes): [PASTE]
Test each item: could a competitor with more money copy it within a year? Keep only what passes, say honestly if nothing passes yet, and suggest what I could start building now to create an advantage over time.
Answer in Thai, in short bullet points a beginner can act on. For every point, say which of my data it comes from, mark anything that is a guess or an assumption I still need to test, and never invent numbers or prices.

How is Lean Canvas different from the Business Model Canvas

Lean Canvas keeps the BMC layout but changes 4 boxes to suit businesses that are not yet sure customers really want what they offer. The other 5 boxes keep their names and positions.

KPKAKRVPCRCHCSC$R$
Business Model Canvas
ProblemSolutionMetricsUVPUnfair Adv.CHCSC$R$
Lean Canvas
Box in the BMCChanged toWhy
Key PartnersProblemMost new businesses fail because they solve a problem customers do not really feel, so the problem must be clear first. Partners can wait until the model is clear.
Key ActivitiesSolutionAt first you do not yet know what you will do routinely, so write the simplest solution instead and tie it to each problem.
Key ResourcesKey MetricsA new business needs to know whether it is heading the right way, so this box becomes the numbers to track.
Customer RelationshipsUnfair AdvantageIt asks plainly what stops competitors from copying you, which helps filter out ideas with no long-term footing.

Which one to use: If it is still an idea, or you are launching a new product and do not know if customers really want it, use Lean Canvas. If the business already has customers and revenue and you want to see the whole operation, including partners and resources, use the Business Model Canvas Many businesses use both: test new ideas with Lean Canvas, then move what is proven into the BMC.

Use AI to find risks and update your plan

Lean Canvas is for finding your riskiest assumption, then testing it with real customers before you invest serious money. Once you have results, come back and update the sheet to a new version.

Find your 3 riskiest assumptions

Let AI play a tough investor and design tests you can run within 2 weeks.

AI PROMPT
Here is my Lean Canvas:
[PASTE YOUR 9 BOXES]
Act as a skeptical investor. Rank the 3 riskiest assumptions (the ones that would sink the business if wrong). For each, design one cheap experiment I can run within 2 weeks: what to do, who to test with, what to measure, and the pass / fail line I should set BEFORE starting (leave the number for me to decide).
Answer in Thai, in short bullet points a beginner can act on. For every point, say which of my data it comes from, mark anything that is a guess or an assumption I still need to test, and never invent numbers or prices.

Update from Plan A to Plan B

Paste what you learned from customers, and let AI show which boxes must change and whether to keep going or change direction.

AI PROMPT
Here is my Lean Canvas Plan A:
[PASTE]
What I learned from customers since then (interviews, orders, experiments, with dates): [PASTE]
Update the canvas to Plan B. Show what changed in each box and why, which assumptions are now validated or rejected, and whether the evidence suggests I should keep going or change direction.
Answer in Thai, in short bullet points a beginner can act on. For every point, say which of my data it comes from, mark anything that is a guess or an assumption I still need to test, and never invent numbers or prices.

For a step-by-step way to test assumptions, read Build-Measure-Learn and MVP

Mistakes beginners often make

  • Starting from the solution box because you love your idea, instead of starting from customers and problems
  • Writing problems you imagine for customers without ever asking a real one
  • Forgetting the alternatives customers use today, and underestimating the competition
  • Using likes or follower counts as key metrics, even though they do not show customers will pay
  • Writing an advantage that is easy to copy, such as "cheaper" or "better service"
  • Making one sheet and filing it away, never updating to Plan B even after learning from customers

Checklist before you use it

  • Choose one group of Early Adopters you can actually reach this month
  • 1 to 3 problems come from customers' own words, with existing alternatives listed
  • The unique value proposition is one sentence that answers problem number 1
  • Prices and costs come from real data, not numbers AI made up
  • Key metrics measure what customers do, not likes
  • List your 3 riskiest assumptions and set a date to update to Plan B

References: Ash Maurya: What is a Lean Canvas? (recommended fill order), Ash Maurya: What is the Right Fill Order for a Lean Canvas?, Wikipedia: Lean startup (Lean canvas section), Strategyzer: The Business Model CanvasAsh Maurya, Running Lean · Lean Canvas is adapted from The Business Model Canvas and licensed under CC BY-SA 3.0 · The business in the examples is fictional. AI analysis is a suggestion to consider, so check it against real data before making investment decisions.

Was this guide useful?
Share it with a friend who is starting a business.
Share to Facebook Share on LINE