Recipes timed 3–30 August 2026 · Published: 28 October 2026 · Last updated: 28 October 2026 · Reading time: ~9 minutes · The timing log, the photographs and the recipe screenshots in this post are my own.
Recipe times are the most-read number in cooking and the least-checked. Nobody stands there with a stopwatch. You read "30 minutes," you believe it, you start cooking at half seven, and you eat at twenty past eight and assume you were slow.
So for a month I used a stopwatch. Eighteen recipes, every one timed from the moment I opened the fridge to the moment the plate hit the table. Not from the moment the pan went on. From the beginning, because that is when the meal starts.
The total was 828 minutes against a stated 655. That is 26% over, and the interesting part is not the average. It is that the error was not spread evenly. The recipes that claimed to be quickest were the ones lying hardest.
What I timed, and what I did not
The clock started when I opened the fridge and stopped when the plate went down. In between: getting things out, chopping, cooking, and the washing up I did as I went. The pans left for after dinner were not counted, because the meal was over by then and I am measuring how long dinner takes, not how long the kitchen takes to recover.
I did not count shopping. If a recipe needed something I did not have, it did not go in the log. That introduces a small bias, because the recipes I had everything for are the ones I make most often, and I am faster at those. I am telling you because it probably makes the numbers slightly generous to the recipes rather than the other way round. A bigger sample would probably show a bigger gap.
One detail about the window: the log ended on the same Sunday I counted my freezer for a different experiment, which is why the final entry is a short one. Two pieces of data collected on the same afternoon is not a problem, but it means the last recipe is a smaller sample than the other seventeen.
The eighteen, in three groups
I have sorted them by what the recipe claimed, because that is the variable I was testing. Four claimed 15 minutes, eight claimed around 30, and six claimed 45 or more.
| Group | Recipes | Stated | Actual | Over by |
|---|---|---|---|---|
| Claimed 15 minutes | 4 | 60 min | 124 min | +107% |
| Claimed about 30 | 8 | 240 min | 335 min | +40% |
| Claimed 45 or more | 6 | 355 min | 369 min | +4% |
| Total | 18 | 655 min | 828 min | +26% |
The four quick recipes took 30, 32, 28 and 34 minutes. The fastest nearly doubled its stated time and the slowest more than doubled it. The eight medium ones averaged 42 against 30. The six long ones were almost exactly right: worst three minutes over, best one minute over, none came in under.
The overhead is fixed. The stated time is not.
The overshoot per recipe fell as the recipes got longer: 16 minutes for the quick ones, 12 for the medium ones, and 2 for the long ones. On the face of it that is a curve, not a constant, and it looks like it contradicts the idea of a fixed cost. I nearly wrote it up that way. So I went back through the log.
The overhead was there in all three groups. About sixteen minutes of getting things out, heating pans, reading the next step, and finding the thing that goes in the thing. What changed was how much of it the recipe had already counted.
The 15-minute recipes counted none of it. A recipe that said 15 minutes and included the prep would say 30, and nobody clicks on 30 when 15 is sitting next to it. The 30-minute recipes counted about a quarter of it. The 60-minute recipes counted most of it, because a recipe that already says an hour has room to include the getting-out and no reason not to.
That is why the long ones came in within three minutes and the short ones came in at double. It is the same sixteen minutes every time. The long recipes just told me about it in advance.
Which means the overshoot per recipe and the overhead per recipe are the same number. That is not a coincidence. The quick recipes overshot by sixteen because sixteen minutes of work was not in the number at the top, and the long recipes overshot by two because fourteen of those sixteen were already in there. The variable is not the overhead. The variable is whether the recipe admits to it.
What the sixteen minutes actually was
I broke it down across the four quick recipes, because that is where it showed up most clearly. Roughly sixteen minutes of overhead per recipe, and it divided into two halves that are not morally equivalent.
About seven minutes was unavoidable. The pan coming up to heat. Water coming to the boil. The oven if it was involved. None of that is me being disorganised. It is physics, and a recipe that says 15 minutes and involves a hot pan has to include it or the number is fiction.
About nine minutes was avoidable. Getting things out of the fridge one at a time. Reading the recipe again mid-cook because I had forgotten the next step. Washing a board because I needed it for something else. Finding the thing that goes in the thing.
That second half is the one I can act on, and it is also the one that explains why I am slower than the recipe. The recipe was written by someone who had already got everything out. I had not, because I decided to cook at 7:20pm and started at 7:21pm.
The thing I got wrong
I assumed recipe times were estimates that drifted. They are not estimates. They are durations of a part, presented as if they were durations of the whole.
A recipe's stated time is how long the cooking takes, written by someone for whom the ingredients are already on the counter and the oven is already on. That is a legitimate thing to state. It is not a legitimate answer to "how long until dinner," and those are two different questions that happen to be printed in the same place.
Once I saw that, the numbers stopped looking like errors and started looking like a category mistake I had been making for years. I was not being slow. I was timing a different thing from the one the recipe timed. Three of the four quick recipes were also new to me, which is worth a few minutes: reading a step twice, checking what "until translucent" means. But not thirteen, which was the smallest gap in that group. Unfamiliarity was a factor and it was not the explanation.
And it explains something from an earlier experiment that I had put down to attention. The four recipes I actually repeat, from the six weeks I logged weeknight cooking, were the ones with the fewest ingredients to buy, and I concluded at the time that it was about attention rather than effort. I think it is both, and this month shows the mechanism. Fewer bought ingredients means less getting-out, which means a smaller overhead, which means the stated time is closer to the real time. The recipes I keep are not just the ones that let me be absent. They are the ones where the clock is telling the truth.
So the rule I came out with: before you cook a new recipe on a weeknight, add fifteen minutes to the stated time. If that number is too big, the recipe is not a weeknight recipe, whatever it says at the top.
And the corollary, which is the one that changed my behaviour: if you want the stated time to be true, get everything out before you start. Nine of the sixteen minutes were me, not the recipe. A recipe that says 15 and takes 16 because you were organised is a recipe telling the truth.
What I would do differently
I would have timed the washing up separately. I counted only the washing up done during cooking, and I suspect the amount of it varied a lot between recipes in a way that is hidden in the average. A one-pan recipe and a three-pan recipe both claimed 30 minutes. One of them had a much better chance of being right.
And I would have noted whether I had made the recipe before. Three of the four quick ones were new to me, and being new is probably worth a few minutes on its own. That is a different kind of overhead from the pan heating, and I have lumped them together in the sixteen.
Should you do this?
The month of stopwatch timing, no. It is more effort than the answer is worth, and I could have got most of the way there by timing three recipes.
The fifteen-minute adjustment, yes, and it costs nothing. Next time a recipe says 15 minutes, plan for 30. If it says 30, plan for 45. If it says 60, believe it. The long ones were the only honest ones in my month.
And if you want to know which of your own recipes are lying, time one. Pick the one you make most often and clock it from fridge to plate. You will find out whether your overhead is closer to my nine minutes of disorganisation or my seven minutes of physics, and those need different fixes.
Frequently asked questions
Did you include the time it took to decide what to cook?
No, and that is a real omission. The clock started when I opened the fridge, which means the decision had already been made. On the nights I had not decided, I was not cooking from a recipe at all, so those nights are not in the log. If you count the deciding, every number here gets worse.
Is this just because you are a slow cook?
Partly, and I cannot separate that out. The nine minutes of avoidable overhead is me being disorganised, and someone better at this would have a smaller number. But the seven minutes of physics does not go away with skill, and the long recipes came in within three minutes, which suggests the cooking part of my cooking is fine.
Do recipe writers know they are doing this?
I think most of them are stating the cooking time honestly and the format does the rest. An honest card would have to say something like "45 minutes, plus 15 if you haven't got everything out yet," which is accurate, and which nobody would click. The number at the top of a recipe card is in a box, next to a clock icon, and there is nothing in that box that says "plus prep." I have started reading it as "cooking time" and it is a different number from "dinner time," and the card does not distinguish them.
Did this change what you cook?
Yes, in one specific way. I stopped trying new 15-minute recipes on weeknights, because a 15-minute recipe is a 30-minute recipe and 30 minutes is the point at which I start thinking about ordering food instead. The quick recipes are for weekends now, when a 30-minute dinner is fine. Which is a strange outcome for a category that exists to solve weeknights.
About this post
Written by Maha Alkumaish. I write about the routines and systems that survive a real working week, tested on myself before they get written about. The timing log and the photographs in this post are my own.
All photographs in this post are my own. I was not paid or sponsored by any brand mentioned, and there are no affiliate links in this post. More about me and how these experiments are run.
Related reading on this blog:
- The weeknight dinners I actually repeat, where the recipes I kept were the ones with the fewest ingredients to buy, and this post is the mechanism behind that
- I walked after dinner every night for two weeks, where a fixed seven-minute overhead took 41% of a ten-minute walk, which is the same structure
- I ate only from my freezer for a week, where a week of cooking from what was already there removed the getting-out time entirely
- More Food & Kitchen experiments
0 Comments