> ## Content Index > Fetch the complete content index at: https://www.mindsetmethodmomentum.com/llms.txt > Use this file to discover other available public pages before exploring further. # Is that a failure, or just a stall? - URL: https://www.mindsetmethodmomentum.com/is-that-a-failure-or-just-a-stall/ - Published: 2026-11-05T16:58:45.000Z - Updated: 2026-11-06T10:23:27.000Z - Description: 'Fail fast' was meant to make you learn quicker. It has become a polite way of giving up - Author: Paul Worthington - Tags: Technology, Momentum > **TLDR** — "Fail fast" began as an engineering idea. In Jim Gray's fault-tolerant computers, a faulty part stopped so that its twin could take over and the system kept running. Business borrowed the phrase as a licence to try small things and learn quickly. Somewhere since, it has become permission to drop anything that wobbles and call the abandonment discipline. A stall and a failure call for opposite responses, and the test between them is one question: what did you learn that changed the destination? If you cannot answer it, you did not fail fast. You stopped fast. --- ## A phrase with three lives "Fail fast" has had three lives, and only the first two were sensible. In the first, it was an engineering term. Jim Gray, writing about fault-tolerant computers in the 1980s, wanted every part of a system to be fail-fast: either work correctly, or detect the fault, say so, and stop. But the stop was never the point. Each part had a twin, and when one stopped the other took over in milliseconds, so the system kept running. The part stops so the whole can keep going. That is the idea in its original form, and it is the opposite of giving up. In the second, it became a business idea, somewhere around the turn of the century. Try small things quickly, learn what does not work before you have spent a fortune on it, adjust. That is still a good idea. It is the [Rule of 3s](https://www.mindsetmethodmomentum.com/the-rule-of-3s/) with a slogan attached, and it fits the model well: short horizons give you honest feedback, and honest feedback is what lets you correct. The third life is the one I keep running into. Something is tried, it does not work first time, and a decision is taken to stop. The stop is described as failing fast, and everyone nods, because it sounds disciplined. What has actually happened is that a stall has been reclassified as a failure so that nobody has to do the hard part, which is working out what went wrong and going again. ## A stall is not a verdict The distinction the phrase has lost is between a stall and a failure, and it matters because they call for opposite responses. A stall is what happens when you set off and lose momentum. The first attempt was clumsy, the plan was wrong in some detail, the person who was going to drive it got pulled onto something else. Nothing about the destination has changed. What the stall tells you is that your method needs adjusting, or your attention has wandered, and the answer is to correct and restart. A wrong first attempt is data. It is not a result. > **"The part stops so the whole can keep going."** A failure is different. A failure is when you have learned something that changes the destination. The customer did not want it. The technology cannot do it at that price. The assumption the whole thing rested on turned out to be false. That is worth stopping for, and stopping quickly is a virtue, because every week you carry on is money spent on a thing you already know is wrong. A stall looks like a failure from a distance, and from a distance is where most decisions to stop get taken. So the question I now ask of any "we failed fast" is a simple one. What did you learn that changed the destination? If there is an answer, good, you stopped for a reason. If the answer is some version of "it didn't work", it wasn't a failure at all. It was a stall, and stopping on a stall isn't failing fast. It is stopping fast, which is just a more efficient way of wasting the time and money you have already spent. ## A definition of dead Most teams already know half of this. Agile has a name for something like it, the definition of done: agree before you start what finished looks like, so nobody can declare victory early. Hardly anyone writes down the other half, which is what would tell you to stop. Call it a definition of dead. Not "if it doesn't work", which is a mood, but a specific thing that, if you saw it, would change the destination. With both written down, the first stall has something to be checked against. Did the thing that would prove you wrong happen? If it did, stop, and stop fast. If not, you have a method problem or a momentum problem, and the answer to both is a smaller next step, not a funeral. So the next time someone tells you they failed fast, ask them what they learned that changed the destination. If they can't say, they didn't fail. They stalled, and then they flinched. ---