We Tried That Already, and No One Remembers How
"We already tried that" ends more good ideas than any budget does, and it almost always runs on a memory nobody can check. A team keeps some record of what it decided and almost none of what it attempted, so the only thing that survives an old experiment is the verdict that it did not work, with the conditions and the method stripped off. This looks at why attempt memory disappears while decision memory sometimes survives, the two opposite failures the gap produces, what a usable record of an attempt actually holds, and how to make the objection earn its authority instead of ending the conversation on its own.
The sentence that ends the conversation
Someone proposes an idea in a meeting. Before it has room to breathe, a person who has been around longer says we already tried that, and the room moves on. No one asks when. No one asks what happened. The idea is quietly filed under settled, and the meeting rolls forward as if a fact just got established.
The sentence carries a lot of weight for four words, and it earns that weight from a simple asymmetry. The person raising the idea has a hopeful guess. The person shutting it down has what sounds like experience. Experience wins that exchange almost every time, even when the experience being cited is thin, half remembered, or about a situation that no longer exists. The sentence works precisely because no one in the room can check it.
It is almost always a guess
Pull on we already tried that and it usually comes apart in your hands. What exactly was tried? When? Who ran it? What were the conditions at the time, the team size, the tools, the market, the budget? What did the team actually expect, and what actually happened, as opposed to the rounded off version that got repeated afterward? Most of the time the person saying it can answer maybe one of those, and only in broad strokes.
What survived is not the experiment. It is the verdict. Somewhere along the way a real attempt, with its own reasons and conditions and messy result, got compressed into two words: it failed. The evidence that would let anyone judge whether the verdict still holds got dropped on the way to the summary. So the team is left carrying a conclusion with no record of how it was reached, and treating it as if it were checked.
The same gap, two opposite failures
When there is no record of what was attempted, the damage runs in two directions at once. The first is a good idea dying for no good reason. The old attempt may have failed because of a constraint that is simply gone now, a tool that did not exist yet, a team a third of the current size, a market that has since turned. The idea in the room might work today for reasons that have nothing to do with why it failed before. But nobody can see that, because the conditions of the old attempt were never written down, so the failure gets treated as a property of the idea rather than of the moment.
The second failure is the mirror image. A team walks straight into an attempt it already made and forgot, spends a quarter rediscovering exactly why it does not work, and arrives back where a colleague who left last year already stood. Here the memory did not distort, it disappeared. The same missing record that kills a workable idea in one meeting lets a dead one get funded again in the next. One gap, two costs, pointing opposite ways.
Teams remember decisions, not attempts
Even teams that are disciplined about decisions rarely keep anything about attempts. There is a reason for the imbalance. A decision produces an artifact almost by nature. Something gets chosen, a plan gets written, a direction gets announced, and the choice leaves a trace because the whole point was to move forward on it. An attempt that did not pan out produces the opposite instinct. The natural move is to close the tab, drop the thread, and get the failure behind you.
So the attempt evaporates while the decision, sometimes, survives. This is why keeping a rationale for your decisions, worth doing on its own, does not solve this problem. A decision record tells you why you chose what you chose. It says nothing about the three things you tried on the way that did not work, or the one you ruled out after a real experiment rather than a hallway opinion. Those attempts are a different kind of memory, and almost no team keeps it.
A verdict is not a record
It did not work is not a record of an attempt any more than the score is a record of the game. It is the outcome with everything that produced it removed. To reuse an old attempt, whether to respect it or to challenge it, you need the parts the verdict throws away: what was actually done, how far it got, the conditions it ran under, and the gap between what the team hoped for and what it saw.
Without those parts, the verdict cannot be argued with, only obeyed or ignored. Obey it and you may be honoring a result that no longer applies. Ignore it and you may be repeating a genuine mistake. Neither is a decision, because a decision needs something to weigh, and a bare verdict gives you nothing on the scale. The two word summary feels like knowledge and behaves like a rumor.
What a usable attempt record holds
An attempt worth remembering does not need a report. It needs a handful of fields captured once. What we tried, in plain terms. When, so the age is visible. The conditions at the time, the few facts that would have to still be true for the result to still apply. What we expected going in. What actually happened. And why we stopped, which is often the most revealing line, because there is a real difference between we proved it does not work and we ran out of time before we could tell.
Notice how small that is, and how much it changes. It is roughly the shape of a decision record with two extra fields, the conditions and the honest reason it ended. Those two fields are the whole difference between a memory you can reason with and a slogan you can only repeat. They are also exactly the fields that never get written, because they are the ones you have to slow down and be honest about at the moment you least want to.
Why just write down your failures does not work
The obvious answer, keep a log of what you tried, is correct and almost never happens, and it is worth being clear about why. Writing up an attempt that did not work is the least appealing note anyone takes. It arrives at the exact moment the team wants to stop thinking about the thing, it reads like admitting a loss, and it competes with the next fire, which always feels more urgent than documenting the last one. A discipline that depends on people volunteering to record their own misses at the worst possible time is a discipline that will not hold.
So the fix cannot be a new rule that people try harder to file post mortems. It has to be capture that forms from the work as it happens, so the attempt, its conditions, and how it ended are already sitting in a place the team can find, without anyone stopping to write the eulogy. A memory of the work should keep what was tried the same way it keeps what was decided, as a byproduct of doing the work rather than a separate chore that only the most conscientious person on the team ever performs, and only until they get busy.
Remember the work, not who failed
There is a real fear underneath the missing attempt record, and it deserves a straight answer. A log of everything that did not work can read like a list of who to blame, and the fastest way to guarantee no one ever writes one is to let it become a scorecard on people. That is the wrong thing to remember and it poisons the right thing.
The line to hold is that this is a memory of the work, not a file on the person who ran it. The useful record is what was attempted, under what conditions, and what happened, kept because the next person deserves to inherit it, not so anyone can be marked down for an honest experiment that did not land. A team that treats failed attempts as information behaves very differently from one that treats them as evidence, and only the first kind ever builds the memory that makes the sentence checkable.
A test for the next time you hear it
You do not need a new system to see whether your team has this gap. Wait for the next time someone says we already tried that, which will not take long, and ask three plain questions. When did we try it? What were the conditions then, and are they still true? And what actually happened, as opposed to that it did not work? Then watch how many of the three anyone can answer.
If the answers come easily, your team has a memory of its attempts and the objection is doing honest work. If the room goes quiet, you have just found a place where good ideas are being decided by confidence rather than evidence. The goal is not to relitigate every settled question, which is its own waste. The goal is narrower and worth it: make we already tried that earn its authority from a record, so it stops being the cheapest way in the building to end a conversation.
FAQ
Is not we already tried that just useful institutional memory?
It can be, when there is a real record behind it. The problem is that the sentence usually runs on a memory nobody can check, a two word verdict that the evidence has been stripped out of. Useful institutional memory would tell you when the attempt was made, what conditions it ran under, and what actually happened, so you could judge whether the result still applies. Without those, the sentence is not institutional memory, it is an unverifiable claim wearing the costume of one, and it ends conversations it has not earned the right to end.
How is a record of an attempt different from a record of a decision?
A decision record tells you what a team chose and, at its best, why. An attempt record tells you what a team tried, under what conditions, and how it turned out, including the things it tried on the way that did not work. Most teams keep some version of the first because a decision naturally leaves a trace, and almost none keep the second because a failed attempt produces the urge to move on rather than to write anything down. You need both. The decision tells you where you landed. The attempt record tells you which roads you already walked and what you found on them.
What should actually go into an attempt record?
Keep it small enough that it happens. What was tried, in plain language. When, so its age is visible. The conditions at the time, meaning the few facts that would need to still be true for the result to still hold. What the team expected. What actually happened. And why it stopped, since running out of time is a very different result from proving something does not work. That is roughly a decision record plus the conditions and the honest reason it ended, and those two extra fields are what turn a slogan back into something you can reason with.
Will not a log of failures just turn into a way to blame people?
It will, if you let it become a scorecard, and then no one will ever write one. The discipline that keeps this useful is to record the work and not the person, what was attempted and what happened, kept so the next team inherits the lesson rather than so anyone can be marked down for an honest experiment. Teams that treat failed attempts as information build the memory. Teams that treat them as evidence against people make sure it never gets recorded, and then keep paying for the same failures twice.
How does a work memory help with this without adding another chore?
The reason attempt records do not get written is that filing a post mortem is the least appealing note anyone takes and it lands at the moment everyone wants to move on. Asking people to try harder does not fix that. The workable version is capture that forms from the work itself, so what was tried, the conditions, and how it ended are already retrievable without anyone stopping to write the eulogy. Driffle is built to keep a memory of the work as a byproduct of doing it, which is the only way the record of what you attempted survives long enough to check the next time someone says you already tried it.