Evidence-controlled workflow
Validation Map
Three validation checkpoints. Each block shows evidence strength, what to test, and the next move.
☰ View details
Recommended next move: describe the idea and run Problem Fit.
Stage 1 · Problem validation
Startup test journey
Start with a saved project, then move one step at a time.
1
Idea
Describe what to test
2
Problem
AI analysis
3
Customer
Empathy map
4
Validate
Real notes or bot
5
Solution
Fit + value prop
6
Market Fit
Best segment
Know what to test
before you build.
before you build.
Define the customer and problem. LeanNav finds evidence, surfaces the demand assumption, and helps you decide what to test next.
1
Name the customer
Role and company type.
2
Describe the pain
No feature language.
3
Find the risk
Test demand first.
1
Define
Role + company type. The analysis scores every piece of evidence against this person.
Describe their pain, not your product. If you catch yourself naming features, describe the moment the customer struggles instead.
This reads like a product description. Try describing the customer's struggle instead.
Add solution idea as context
Optional. Problem analysis uses the customer and problem first; solution details are parked for later.
Most useful: complaints and workaround attempts.
LeanNav uses AI and synthetic customer simulations. Results are directional and should be validated with real customers before making build decisions.
Searching for real customer evidence...
reddit.comtrustpilot.comg2.comproducthunt.com
Analysing signal patterns...
Extracting what could kill this idea...
Calculating evidence resonance score...
Generating problem statement...
Problem
Problem in plain language
What is happening
Who is affected
Why it matters
What success looks like
2
Prioritize
Run an analysis first
Root assumption
—
How often it comes up
—
Tried to fix themselves
—
How frustrated they are
—
Would switch for a fix
—
Evidence you collect for any assumption below also counts toward this statement.
Recommended next test
Targeted evidence
Finds public signal showing whether this problem is real and urgent for your target customer.
Not run
Narrow by country:
→
Change direction or keep going?
Signal strong · 1 synthetic customer explored · validate with real customers
→ Keep going
Test your solution with real customers.
↻ Change direction
Update idea and run fresh analysis.
3
Test
Select an assumption to unlockTest your assumption — choose your method
🤖 Rehearse
Practice with a synthetic customer grounded in your evidence. Does not produce evidence.
Fills Sophie's empathy map as you go ·
💬 Send the bot
Send a link — a bot guides each customer through your assumption questions. Responses update your evidence score automatically.
Waiting for responses...
Personalise consent message
📄 Run it yourself
Download a printable interview guide and run your own in-person or video interviews. Import findings when done.
Evidence-grounded synthetic customer
SC
Empathy map
Synthetic · evidence-backed
0%
Hear empty
Chat to fill
Think & feel empty
Chat to fill
See empty
Chat to fill
Say & do empty
Chat to fill
SC
Customer
Synthetic profile
Chat to fill
Pains empty
Chat to fill
Gains empty
Chat to fill
Chat to fill the empathy map
SC
Customer
Building empathy map
ThinkHearSeeSayPainsGains
🎯 Ask open questions to understand how this customer thinks, feels, and behaves.
Fill the empathy map — ask what they think, feel, hear, see, say, and do
Generate a customer profile first
6 more evidence-grounded synthetic customers
Generating 6 more customers...
—
Customer
Synthetic profile · coached interview
Ask about their experience with this problem.
Pattern across 7 empathy maps
Agrees with kill condition
Disagrees — different trigger
Both factors play a role
Prepared interview guide
🔗 Send a link to real customers
Bot sends guided discovery questions from the prepared interview guide. Score updates when 3 respond.
📤 Already ran interviews — upload results
Paste transcript. LeanNav assesses quality and updates score.
⚠ Skip validation for now
Proceed on synthetic signal only — real evidence can be added later.
Lower confidence. Validate before committing to build.
—
Synthetic only
→
—
After interviews
✓ Interview responses received
0 responses received
Updated analysis — grounded in real customer voice
New pains surfaced
Updated problem statement
Before
→
After real interviews
Stage 2 · Solution Fit
1
Define Solution
Solution
2
Prioritize
Analyse solution first
Solution fit
—
Solves the pain
—
Would switch
—
Trust handled
—
Pushback risk
—
Collect solution evidence below to upgrade this from a model estimate to measured signal.
Targeted evidence
Searches for adoption signals, competitor reviews, and early-adopter feedback for this type of solution.
Not run
3
Assumptions
Viability and execution risks to validate before building
Analyse solution first
Responsible AI Risk Register
Not scored · not on the assumption board
4
Test
Analyse solution first
Test your solution — choose your method
🤖 Rehearse
Test your solution against synthetic customer profiles grounded in your evidence. Does not produce measured signal.
💬 Send the bot
Send a link — a bot tests customer reactions to your solution idea. Responses update your evidence automatically.
Waiting for responses...
📄 Your findings
Paste in customer feedback, experiment results, or observations from your own testing.
Stage 3 · Market Fit and decision
1
Identify Market
You'll see which customer segments have the strongest signal and which to pursue first — so you know where to focus before you scale.
Top segment
2
Market Signals
Run market analysis first
Market signal combines evidence quality, persona fit, solution fit, and validation status.
3
Assumptions
Market and go-to-market risks to validate before scaling
4
Decide
Run market analysis first
Commitment evidence
Add behavior that proves somebody is willing to act: an LOI, paid pilot, preorder, qualified signup, waitlist, or explicit buyer commitment.
Ready for a final recommendation?
See your scores, Build/Pivot/Stop verdict, and export the discovery report.
Final Report
What do you do next?
Based on your signal score, empathy maps, and real interview results.
🚀
Build
Signal validated. Start with the smallest testable version.
↻
Pivot
Evidence points to a different problem. Update and re-run.
✕
Stop
Signal too weak. Save time and resources.
Run Problem Fit analysis in Stage 1 first — your recommendation appears here once the evidence is scored.
Discovery report
Pilot feedback for Julia