Follow the recipe once, then keep the parts that help.
Use the situation as your starting point, follow the steps, and save the parts that become a habit.
Try saying
“I am in this situation. Walk me through the recipe and tell me what good looks like.”
Name the situation
Start with the real moment you are in.
Follow the recipe
Do one step, then check the result.
Make it yours
Save the workflow when it works.
Run a quarterly assumption workshop
Situation
You're planning the next few months. The roadmap looks reasonable, but it rests on things nobody has checked. This 90-minute session makes those beliefs visible, picks the few that could sink the plan, and ends with a test and an owner for each.
What you need
- 3 to 6 people: product, design, engineering, and someone who talks to customers
- A saved product brief in Settings → Product, so Specky can read your canvas in context
- The bets you're considering for the next few months
- 90 minutes, then 15 minutes a week afterwards
Steps
1. Name your bets (10 minutes)
Write down each bet. A bet is a choice you could lose on, like "sell to teams instead of individuals". Keep it to three or four. If everything is a bet, nothing is. You'll add them to the board in step 6, once it has notes on it.
2. Fill in the Lean Canvas together (15 minutes)
Open Assumptions, go to the Lean Canvas tab and write down how the product is meant to work. Rough is fine. Disagreement about what goes in a box is useful, so write down both versions.
3. Imagine it failed (10 minutes)
Ask the room: it's a year from now and this failed, so what went wrong? Click Run Pre-mortem in Strategy canvas (Growth plan), or do it out loud. People will say things in this framing that they won't say in a planning meeting.
4. Pull out the assumptions (10 minutes)
Press Find the assumptions on the Lean Canvas tab, keep the ones worth tracking, and add them to the board. Assumption cards you saved from the pre-mortem are already there. Add anything else the room raised with New assumption. Check that each one could actually be proven wrong. "Customers want this" can't. "Teams of five to fifty will pay €50 a month" can.
5. Break down the big ones (10 minutes)
If one feels too large to test, open it and press Break down. A big assumption is false if any one of its parts is false, and the parts are much cheaper to check.
6. Place the notes together (15 minutes)
Add your bets under Your bets. Drag each note to where the group thinks it belongs: up for impact if it's wrong, right for how little you know. Then click each note and pick the bet that depends on it under Which bet depends on this?. The arguments here are the point, because they show where the team is guessing.
7. Choose what to test (20 minutes)
Work on the numbered notes in the top right, and stop at three to five. For each one, press Pick a test and choose the cheapest method that could change your mind. Specky fills in the plan. Agree on the pass signal now and give the test one owner.
If your team plans in Miro, click Send to Miro and choose This board to copy the board and canvas to a new Miro board. Connect Miro in Connect tools first.
What you should get
- Three to five tests with an owner and a pass signal
- A board showing which beliefs matter most and how little you know about them
- A view of each bet, and which ones rest on the most guesswork
Every week after
Spend 15 minutes on the board. Record what finished in Experiments. The note moves and the bet above it updates. Then decide:
- Supported: move on to the next riskiest belief.
- Unclear result: run a sharper test, or accept the risk and say so out loud.
- Contradicted: stop and decide whether to change the bet, aim it at someone else, or drop it.
Where this usually goes wrong
- "We don't have time." A week of small tests costs less than a quarter spent building the wrong thing.
- Discussing instead of testing. If the room has gone around twice, the answer is a test.
- Testing to feel good. Set the pass signal first, so you can't reinterpret the result.
- Running one test. New ideas need several, moving from conversations to real behaviour.
- Handing testing to someone else. The people making the decision should see the evidence first-hand.
- Trusting weak evidence. What people said in an interview is a starting point, not proof.
How to adapt it
- Solo founder: do it alone in 45 minutes, then show your top three to one customer and one person who'll disagree with you.
- Several products: run one per product, then compare the bets side by side.
- No customers yet: expect most notes on the far right. That's the honest starting point, so begin with conversations.
Ready to apply this?
Start with your own product and keep the first read grounded. You'll create an account before the guided result so it can be saved to your workspace.