Don't Kill the Quirky One Yet: A Five-Point Stress Test for Odd Ideas
Every brand workshop has a moment when someone floats an idea that makes the room go quiet. Not because it's brilliant. Because it's odd. The instinct is to kill it quickly and get back to the sensible options. Sometimes that's right. Often it's just fear wearing a lanyard.
The useful question isn't 'does this feel like us?' It's 'is this the right kind of wrong?' That's where the blue lemon comes in. A blue lemon isn't a defective lemon. It's a lemon that makes you look twice.
Why 'wrong' is not a diagnosis
There are two types of wrong idea. The first is bad wrong: confusing, off-strategy, expensive for no reason, legally shaky, or just a bit embarrassing. The second is productive wrong: surprising, memorable and somehow more true than the polite version.
Most teams kill both with the same phrase: 'It doesn't feel on-brand.' That's not a diagnosis. It's a shrug. You need a test that separates the two before you bin the weird one.
The Blue Lemon Audit: five questions
Score each question from 0 to 5. Be honest. The total is less important than where the disagreement sits.
1. Is it strategically true?
Does the idea express a real brand truth, audience tension or product difference? Or is it weirdness for its own sake? A blue lemon still has to be a lemon. If you removed the odd bit, would there be a strategy left?
2. Is the weirdness doing work?
Would a normal version be forgettable? Does the strange element carry the message, or is it decoration? Productive weirdness makes the point harder to ignore. Bad weirdness makes the point harder to find.
3. Can we say it in one breath?
Can someone explain the idea to a friend without a diagram? 'We pay customers not to buy things' is one breath. 'A multi-platform ecosystem of reciprocal value exchange' is not. If it needs a white paper to survive, it may be the wrong kind of complicated.
4. Does it survive contact with the real world?
Can it be made properly? Does it work in the channels you actually have? Can legal and finance live with it? This isn't about playing safe. It's about whether the idea can be executed with enough craft to be judged fairly.
5. Would a sceptical customer remember it for the right reason?
Imagine your least enthusiastic customer seeing it. Would they remember the brand, or just the confusion? The aim isn't to shock. It's to be recalled for something true.
How to score it without turning it into a committee
Add up the five scores for a total out of 25.
- 0–10: bin it, or reframe it.
- 11–17: interesting but undercooked. Develop the weakest question.
- 18–25: worth prototyping. Champion it, but keep the sceptics in the room.
The scoring is a pause, not a veto. The most useful part is the spread. If one person scores strategic truth at 5 and another at 1, you've found the real argument. Have that argument before you kill the idea.
Worked example: the anti-loyalty scheme
Imagine a Bristol refill shop proposes a loyalty scheme with a twist: customers earn credit for repairing something or returning packaging, not for spending more. The line is 'We pay you not to buy'.
Score it:
- Strategic truth: 5. The brand exists to reduce waste. The idea makes that real.
- Weirdness doing work: 4. The counterintuitive mechanic is the message.
- One breath: 4. Easy to explain, slightly surprising.
- Real world: 3. Fraud and admin need solving, but it's doable.
- Sceptical customer: 4. Memorable and not confusing.
Total: 20. That's a prototype, not a final campaign. The audit doesn't prove it will work. It proves the team should test it rather than kill it because it feels odd.
When to ignore the score
If an idea scores low on strategic truth, no amount of weirdness saves it. If it scores high but fails the real-world test, fix the execution or walk away. The audit isn't there to make weirdness safe. It's there to make sure you're killing ideas for the right reasons.
A blue lemon is not a licence to be random. It's a prompt to look twice. Run the five questions. Then decide.
