Reverse-Reasoning Innovation
Find an interesting effect first, then discover the problem it solves
- Difficulty
- Moderate
- Time to result
- ~months to results
- Steps
- 5
- Confidence
- 74%
Sutherland contrasts forward-reasoning science, which defines a disease and hunts a cure, with reverse-reasoning innovation, which finds an interesting compound or effect and then asks what it is good for. The largest breakthroughs, he argues, happen backwards: Viagra was an angina trial whose 'side effect' was the product; penicillin was a cure looking for a disease. The discipline is to refuse to discard a result just because it failed the original hypothesis, and instead ask what different question it answers. A drowsiness side effect that ruined a daytime flu remedy became the whole selling point of Night Nurse once reframed as a nighttime product. An immune-boosting drug that was wrong for autoimmune disease was exactly right for cancer immunotherapy. The bug becomes the feature the moment you change the context.
Origin
Rory Sutherland draws this from advertising's tolerance of post-rationalization and from pharmaceutical history, recounted on Modern Wisdom via Viagra, Night Nurse, and immunotherapy.
Core principles
- 01The biggest breakthroughs are often cures looking for a disease, not the reverse
- 02A failed result against your hypothesis may be the right answer to a different question
- 03A bug in one context is a feature in another
- 04You are allowed to post-rationalize; how you arrived at the answer doesn't invalidate it
How to run it
- 1
Watch for anomalies
Pay attention to unexpected effects and side effects in trials, usage data, or observation rather than treating them as failures.
Pro tip In the Viagra trial, patients declining to return unused pills was the signal.
- 2
Ask the different-question test
Before shelving a result that failed your hypothesis, ask: what question is this the right answer to?
Watch out A German firm reportedly noticed the same Viagra effect and rejected the drug for failing as an angina cure.
- 3
Reframe the bug as a feature
Identify the context in which the unwanted property becomes desirable and reposition around it.
Pro tip Drowsiness kills a daytime remedy but sells a nighttime one.
- 4
Post-rationalize and test
Reverse-engineer a plausible mechanism for why it works, then validate empirically; don't demand you predicted it in advance.
Pro tip If it's an answer, it's an answer; this isn't a maths exam where you must show your working.
- 5
Reposition and launch
Rebuild the product's story around the new use and take it to the audience for whom the effect is a benefit.
In the wild
A flu treatment had the drawback that it sent users to sleep, ruining it as a daytime remedy. A marketer suggested positioning it as a nighttime flu remedy, at which point the drowsiness became a bonus rather than a bug. Night Nurse was born by reframing rather than reformulating.
→ A commercial success created purely by changing the product's context.
Drugs tested for autoimmune diseases like rheumatoid arthritis were shelved because they boosted the immune system, the opposite of what those patients need. Later, someone recognised that for cancer immunotherapy a drug that strengthens the immune system is a feature, not a bug.
→ A discarded 'failure' repurposed into a valuable cancer therapy.
Common mistakes
Discarding off-hypothesis results
Judging a discovery only against its intended target throws away answers to different, sometimes larger, questions.
Demanding pre-rationalized justification
Insisting an idea be logically justified before acting blocks the abductive, post-rationalized route that produces the biggest breakthroughs.
Is it for you?
Best for
R&D teams, product marketers, and inventors sitting on 'failed' experiments or awkward side effects.
Not ideal for
Domains that legally or ethically require a pre-specified, single-target result before any action.
From the transcript
“the real breakthrough things happen backwards okay which is Viagra penicillin... it's almost a cure for which there is no known disease”
“if you position it as a nighttime flu remedy okay the fact that it sends you to sleep is a bonus it's not a problem”
“don't apologize for a post rationalization if it's an answer it's an answer”
From the episode
Hidden Psychology Of The World’s Best Advertising - Rory Sutherland - #735
Rory Sutherland