Every Post-it note on your desk started life as a failed experiment.
Not a happy accident. Not a lucky break. A failure. A chemist set out to make one thing, made the opposite, and then spent years carrying the wrong answer around his office looking for a question to fit it.
I think about this story a lot, because Gillian Jones-Williams and I had a friendly argument about failure. Her line was simple. There is no failure, only a lesson. I pushed back. I think failure is real, and I think pretending otherwise lets us skip the part where it hurts.
The Post-it story settles the argument for me. We were both right, and the reason we were both right is the useful bit.
The glue nobody wanted
In 1968, Spencer Silver was a senior chemist in 3M's central research lab. He was trying to build a strong adhesive for aircraft construction.
He failed.
What he made instead was a low-tack glue built from tiny acrylic spheres. The spheres only touched a surface at their tips, so the glue held paper in place but let go without tearing it. You stuck it, peeled it, and stuck it again.
For an aircraft, useless. For anything else, nobody knew.
Here is the part I love. Silver did not bin it. According to the Lemelson Center at MIT, he spent the next five years showing it to colleagues, in corridors and in seminar presentations. He pictured it as a spray, or as a coating for bulletin boards. None of it landed. 3M's own brand history later called it "a solution without a problem".
Five years. Picture being the guy at the seminar with the weak glue. Again.
The man with the falling bookmarks
In 1974, Art Fry sat in one of those seminars. Fry worked in 3M's tape division, and he sang in his church choir on Sundays.
His problem was small and annoying. He marked his hymns with scraps of paper, and when he stood to sing, the scraps slid out and fell to the floor.
Fry put the two things together. Silver's useless glue on the back of a bookmark. It stayed put, peeled off clean, and left no mark on the page.
Even then, the road stayed bumpy. The Lemelson write-up says management worried the product would look wasteful. 3M test marketed it as "Press 'n Peel" in four cities in 1977, and the results disappointed. A year later 3M flooded offices in Boise, Idaho with free samples, and more than 90 percent of the people who tried them said they would buy. Post-it notes launched across the United States in 1980.
From failed aircraft glue to national launch... twelve years.
Even the color came from a scrap. The yellow was whatever spare paper the lab next door had lying around.
Why Gillian and I were both right
Silver failed. Period. He set a goal, he missed it, and the thing in his jar did not do the job he wanted. Calling it "a lesson" in 1968 would have been a nice story told by someone who knew the ending.
He did not know the ending. Nobody did.
So I still say call a failure a failure. Name it. Feel the sting. Write down what you were aiming for and what you got instead.
Gillian's point lives in what happens next. A failure becomes a lesson when somebody keeps it, labels it, and shows it around. Silver's glue only turned into raw material because he refused to let it disappear into a drawer.
The failure is the event. The lesson is a choice you make afterwards, and you keep making it for years.
Most dead ends go in the bin
Here is my worry. Most organizations make Silver's choice impossible.
Amy Edmondson at Harvard Business School has spent decades studying how teams handle failure. In her HBR piece on learning from failure, she describes asking executives two questions. How many failures in your organization deserve blame? Their answers usually sit at 2% to 5%. How many get treated as blameworthy? After a pause or a laugh, they say 70% to 90%.
Her conclusion: people stop reporting failures, and the lessons vanish with them.
Think about what this means for a modern Spencer Silver. He makes the weak glue. His boss asks why the aircraft adhesive is late. He mumbles something, tips the jar down the sink, and never mentions it again. No seminars. No five years of corridor conversations. No Art Fry.
My own research on bad bosses found 99.5% of survey respondents had experienced one or more types of bad boss. I would bet a big share of those bosses never threw anything at anyone. They punished the wrong failures, and their people learned to hide the evidence.
I wrote recently about the difference between a mistake and a pattern of underperformance. Silver's glue was neither. It was an experiment, and Edmondson has a name for the good kind.
Not every failure deserves a shelf
I don't want to romanticize this. Plenty of failures are plain carelessness. Nobody needs to keep the report you forgot to proofread.
Edmondson's work on intelligent failure sets out four tests:
- New territory. You had no recipe to look up.
- A credible shot at something worth having. The goal mattered.
- A hypothesis. You had a reason to think it would work.
- As small as possible. You spent no more than you needed to learn the lesson.
Silver ticks all four. Strong aircraft glue was new ground, worth money, grounded in chemistry, and tested in a lab jar rather than a fuselage.
Those are the dead ends worth keeping. The rest you fix and forget.
What I keep on my own shelf
I have a spreadsheet of post ideas. It runs to more than 1,200 rows. Hundreds sit there unused, some for months. A few of them looked brilliant when I wrote them down and look silly now.
I don't delete them. Every so often one sits beside another and the pair makes something neither managed alone. This post came from row 896 of the sheet. It waited its turn.
I have the same habit with code. Old prototypes, half-built tools, scripts written for a problem I solved a different way. Most of them will never see daylight. One or two will meet their Art Fry.
How to keep your dead ends alive
If you lead a team, or only yourself, here is what the Post-it story asks of you.
Label it honestly
Write down what you aimed for and what you got. "Wanted strong glue. Got weak, reusable glue." Silver's description of his failure turned out to be the product spec.
Put it where other people trip over it
Silver ran seminars. Five years of them. The person who solved his problem worked in a different division and had a hobby Silver knew nothing about. Your Art Fry likely sits outside your team. Share the dead end widely, not in a private folder.
Separate the event from the person
If your people hide failures, you get no raw material at all. Treat the 2% to 5% blameworthy cases as blameworthy. Treat the rest as data. I lean on this when I work with leadership teams at Step It Up HR, because the hiding starts at the top.
Expect a long wait
Twelve years from lab to launch. If your dead end has not found its use in a quarter, the dead end is fine. The quarter is too short.
The failure is yours. The lesson is a choice.
I wrote a while back about how not trying is its own kind of failure. This is the other half. Once you try and miss, the miss belongs to you. What it becomes depends on whether you keep it.
So here is my question for you. What is sitting in your jar right now... the project, the idea, the prototype you wrote off... and who have you not shown it to yet?