Idea lab: judging and pitching ideas
Most ideas should die early and cheaply, and the good ones need a clear case. This lesson gives you a sequence that does both: a quick pain, frequency and reach screen, a bottom-up size, evidence that someone will pay, the riskiest assumption and its cheapest test, kill criteria written before you start, and a one-page pitch a decision-maker can say yes or no to.
By the end you can
- Screen a list of ideas by pain × frequency × reach with anchored scales, and know what the screen misses
- Size an idea bottom-up from buyers, uptake, price and frequency, and sanity-check the result
- Find willingness to pay from behaviour and commitments before surveys
- State the riskiest assumption, the cheapest test and kill criteria with a state and a date
- Write a one-page pitch with a clear ask
The everyday picture
You want to open a food stall at a busy bus stop. Before you sign a year's rent, you stand there for a morning and count how many people pass at breakfast time. You ask a few of them what they buy now, where, and what they pay. You check whether you can buy ingredients cheaper than the stall across the road. You work out how many plates a day you need to cover rent.
Then you do one more thing that most people skip. You write down, before you open, what would make you close: "If I'm selling fewer than 60 plates a day by the end of week six, I stop." On a bad Tuesday in week five you won't have to argue with your own hopes. That is the whole lesson: count, check who pays, size it from the bottom, and decide in advance what failure looks like.
The first screen: pain × frequency × reach
The last lesson ended with ten or more idea cards. You can't size them all, so start with a fast screen. For each idea, score three things from 1 to 5, using anchors so that two people give the same score:
- Pain, each time it happens. 1: under a minute or a few cents. 2: a few minutes. 3: ten minutes or more, or a visible complaint. 4: a lost sale, a refund or a cash shortfall. 5: a safety, legal or large money risk.
- Frequency, for each person who has the problem. 1: monthly or less. 2: a few times a month. 3: weekly. 4: several times a week. 5: daily or more.
- Reach, the number of people who have it. 1: under 50. 2: 50 to 500. 3: 500 to 2,000. 4: 2,000 to 100,000. 5: more than 100,000.
Multiply the three. Multiplying matters: an idea that scores 1 on any of them is weak however high the others are, because a rare problem or a tiny group can't carry a product. Say which side you scored. "Live substitutions" scored for customers is a different number from the same idea scored for pickers.
This is not RICE (pm-12). RICE ranks work that is ready to build, with effort and confidence. This screen comes earlier, on problems, before anyone has estimated a solution.
| Idea (prompt) | Side scored | Pain | Freq. | Reach | Score |
|---|---|---|---|---|---|
| Request desk (16) | Store and office staff | 3 | 5 | 4 | 60 |
| Live substitutions (6) | Customers | 4 | 2 | 5 | 40 |
| Shop credit (7) | Partner shop owners | 4 | 3 | 3 | 36 |
| Dock booking (10) | Supplier drivers, DC staff | 3 | 5 | 2 | 30 |
| Proof of delivery (11) | Shop owners, drivers | 4 | 2 | 3 | 24 |
| Ops copilot (18) | Store managers | 3 | 4 | 2 | 24 |
| Staff lessons (12) | New store staff | 2 | 5 | 2 | 20 |
| Scan-and-go (2) | Customers at festivals | 3 | 1 | 5 | 15 |
Know what the screen misses. It measures pain, so it undervalues ideas whose value is new revenue or better decisions. Shop credit's main value is that shops buy more, which pain doesn't capture. A feedback summariser (prompt 20) scores low because few people feel the pain, yet it could change what the whole team builds. Keep one "wild card" past the screen if you can write down why the screen misjudges it.
Quick check
An idea scores Pain 5, Frequency 1, Reach 4. Another scores 3, 3, 3. Which statement is right?
Multiplying gives 20 and 27, so the steady weekly problem beats the rare severe one on this screen. A 5 for pain deserves a second look, because severe rare problems such as safety risks can still matter. Effort belongs to RICE later, not to this screen, and adding the scores hides a weak dimension.