The cost of interrupted work (2023) (opens in new tab)

I've found one thing that minimizes interruption cost: pair programming. At one startup we pair programmed all day, every day. Resuming from an interruption was almost seamless. Can't explain it, only experienced it.

My feelings exactly. Whenever I get one of those days I’ve found I get more done away from my desk. Taking a walk in the forest or doing some household chores can really help you get back on track.

Of course at times it’s just better to admit altogether this isn’t a day meant for work and spend it relaxing instead. Usually, the benefit of that is a really productive day at work the day after too, everybody wins.

For me, the difference is whether I had pre-planned my tasks or not.

Knowing what I wanted to focus on and achieve from 10-11 am makes it much easier to get back on track.

In contrast, when I simply begin working on something, I end up elsewhere easily, even without external interruptions.

For me, it is more about the nature of the interruption. An easy question that only requires pulling something from memory doesn't cost much. A question that requires some thinking has a huge cost, even more so when you have to check the code or documentation. Sometimes even reading an email/teams notification can throw me off. It's not about the interaction (someone coming to your desk, a call,...), it's about the topic.

However I think that in both cases, if the interruption happens while coding, the risk of bug is about the same.

I noticed it positively correlates with the amount of sleep I got the night before, and negatively with the amount of coffee I drank the week before.

It could be that some tasks require you to hold a lot of context in your head. Interruptions clear that context and it must be rebuilt.

Or you could be tired. It’s crazy how ineffective one gets with a little less sleep.

For me it's a combination of the nature of task, where I'm at and the nature and duration of the interruption. But usually an interruption causes a large amount of penalty points.

I find a bit of meditation before work and good coffee reliably put me in the latter state, maybe 19 times out of 20.

So the average is 23 min 15 sec right?

Do you think it's related to the level of emotion tied with the interruption? For example, finding out a friend or relative was in a bad accident is surely gonna be more distracting than someone asking if they can borrow a pen for a few minutes.

As I get older, I feel I take more time to get back on track.

I don't feel less intelligent, maybe more experience compensates for it. I probably make less wrong turns. But I have to be more rigid to prevent interruptions.

Right there with you :)

When I work places where interruptions are expected I change how I do my work. So an observer may see them costing me less time but that’s amortized across the rest of my work as a consequence of leaning into interruption. It’s still there, it’s just spread out more.

> This is a really common problem with science reporting in general. Its often the case that the news will say things about the paper that aren't in the paper, often they say something that is completely the opposite of what the paper actually represents with its data.

I wonder if perhaps a part of LLM hallucinations can be explained by them being provided such reporting and having it (mistakenly) tagged as high-quality training data.

  > This is a really common problem with science reporting in general.

There's a lot that actually becomes really sinister with this. I'm sure the people writing those news articles pass them off as little white lies, "close enough", or whatever. And in some sense, there's a lot of truth to that. But the problems stem from what grows out of these "small errors" and into far larger ones.

1) it certainly contributes to a significant part of the g̶r̶o̶w̶i̶n̶g̶ distrust in science. People are getting their science from the news, not the horse's mouth. If you report something silly while saying "scientists say" then people will point the finger at the scientists more than they will point to the reporter. The scientists aren't writing in plain English after all[0]. Things like chocolate being healthy for you, red meat causing cancer, or machine learning quantum blackholes. There's always elements of truth to these things but truth is not sensational. Truth and complexity go hand in hand. As complexity decreases you will have to sacrifice accuracy. This is a tough balance to play[1].

2) The public isn't very scientifically literate and it is easy to misconstrue meaning. Hell, even scientists routinely struggle with this stuff. Let's take the red meat issue as an example. It's true, but a lot of those studies were looking at daily intake of 50g or 100g and results like "25% higher risk of rectal cancer" at the higher end of the estimates. For context, a Costco hotdog is 110g while a Nathan's Hotdog is 48. We're talking about 1-2 hotdogs per day. We're also talking about a percent increase in risk, not a percent risk. The CDC site says approximately 3.9% of men and women will be diagnosed with colorectal cancer. If that's our baseline then a 25% increase is a 4.9% risk. That's a disgusting amount of hotdogs to move from a 4% cancer risk to a 5%. Concerning on a national level but not on a personal. This feeds back to #1 as people are interpreting the news as saying "eating hotdogs makes you likely to get cancer" while people observe heavy hotdog eaters around them not getting cancer. Their observation wouldn't run counter to what the research says but it will against the narrative on the news. The incongruence between observation and understanding does justify mistrust. But there's just more ways to misinterpret something than there are to interpret correctly (relates to [2]). This failure mode becomes self-reinforcing. Enough that I think anyone that spends any time on the internet will be aware of it. (Communicating is fucking hard, communicating accurately is even harder)

3) (Perhaps the worst part) Scientists are primarily measured by their citations (count or "h-index"[3]). Unless you do something groundbreaking[4] (which is rare) then this is the main way to "measure" performance. A great way to boost citations is getting media attention. Unfortunately there's just a lot of papers published and a primary driver of citation count is knowledge of a paper's existence. You don't need to be an Avi Loeb type (it doesn't hurt) when we're talking about small numbers. If a Cal State grad student and a MIT grad student were to publish the same paper we'd expect the latter to have more citations due to the latter's greater visibility. MIT has a media wing and these papers are much more easily picked up by larger news orgs. This is why so many scientists use platforms like Twitter. Because your work doesn't mean anything (to your personal success and ability to continue doing your work) if you can't get enough citations. There's an obvious slippery slope here... One that can create a feedback loop to misreporting. The fiercer the competition the the more risky this situation becomes. It's really easy to do slight embellishments of your work. No one is checking at time of publication. Replication happens later and the system devalues replication. Plus, while replicating it is much easier to assume you've made a mistake rather than the paper was in error (or in serious error).

All this is to say that shit is messy. And I don't think any of it is particularly any one person's fault. More an emergent phenomena through compounding effects. Little things here and there add up as we talk about millions of events and many years. I know we all want things to be simple, and simplicity has a lot of benefits, but it also can be a big trap. "As simple as possible, but no simpler" does not mean something isn't extremely complex[5]. It's a trap that makes people think they can read a few lines from Wikipedia and understand something (read the whole article, it's still not enough). A trap with growing consequences to a world that grows in complexity[6]

[0] And I don't think w̶e̶ they should. Papers are a peer-to-peer communication network. Open and visible, but that's how the peer-to-peer communication takes place. Expert-to-public communication has traditionally been done through news or other science communicators. Asking scientists to write papers to the general public is like asking you to communicate to your coworker about your code as if your coworker knows nothing about code or the context it is running in (all because a layman may overhear). Good luck getting any work done...

[1] While news orgs and science communicators (especially pop sci communicators... ugh...) are doing harm here there are defenses anyone can take. Recognize your understanding is always wrong to some degree. Don't take in information as binary true/false statements but probabilities: e.g. likely true/maybe true/maybe false/likely false. Fundamentally the reason this is a good defense is because it is always a more accurate interpretation. Scientists don't find truth through confirmation but through negation. What I mean is w̶e̶ they rule things out. A scientist converges to truth[2]

[2] This also helps you sniff out conmen from the scientists. The scientist always has some doubt. At first they may show themselves as highly confident but as you press on detail they start weakening language. This isn't foolproof and isn't gonna work for every question, but it is common. Scientists are being more trained on media literacy because there is recognition that while this is the right way to talk to peers it gives the public a sense that they lack expertise rather than are aware of complexity. The best signal is just getting them to talk about their domain. They won't stop and will get very detailed.

[3] Number of papers with greater than N citations. An h-index of 10 is having 10 papers with >= 10 citations. h-index of 100 is 100 papers with >= 100 citations.

[4] If you do something groundbreaking no one gives a shit about citation count. But doing something groundbreaking will surely make your citation count skyrocket (often it also drives your h-index too, as you've simply gained more attention and more people are reading your other works. Your work doesn't change, but visibility does). This fact is often used to justify the usage of citation metrics. Citation metrics are fine, but they're also easy to hack and highly context driven. I mention Avi Loeb, and controversy is beneficial to this metric. Every paper that cites him to say he's wrong is a point for him. Controversy is a way to gather points from a whole new source! Not those building on your work, but those building against your work.

[5] "If you can't explain it simply, you don't understand it well enough" is laughable phrase. The reason you can't teach "a barmaid" Quantum Chromodynamics isn't because you don't understand it. You can't explain it "simply" because you don't understand it. You can't explain it simply because you can't even use the word "color" without a layman thinking a quark is red (can you even get into how it is impossible to have "red" at that scale?).

[6] Progress necessitates an increase in complexity. Look at a Taylor Expansion and it's relationship to computational difficulty. I'll let you all figure this out, my comment is already too long.

I interpret this and the article as saying that those 23 minutes are not spent trying to resume the original task, but on the interruption itself and the other intervening tasks that are worked on before returning to the original task.

If that interpretation is correct, those 23 mins are not wasted in confusion but simply spent on other things.

Do i read it correctly?

Unfortunately that's not the source. The author of this piece, Jaro Fietz aka oberien, is already familiar with that article, noting it in their diagram and linking to it toward the bottom of the page. They're looking for the underlying research itself rather than a quote about it.

>So in the end, where do the 23 minutes and 15 seconds come from? They are mentioned in interviews multiple times by Gloria Mark. But I wasn’t able to find a primary printed source. There are many more publications by Gloria Mark, but none of them turned up while searching for the 23 minutes 15 seconds figure. If someone knows a paper or study where that figure originally appears in, please tell me.

What was the bad news?

This is what kills my days. Other interruptions from coworkers usually are on topic and it’s always a pleasure to help out.

Oh how I hate a last minute rescheduling of a meeting. Feeling like I only have 30 minutes, I didn't start anything deep. Such a waste.

Pre-pandemic, I tended to keep an erratic home-work/office-work schedule. Sometimes I would be at home working, and then leave to go to the office for a meeting, but I'd arrive a half hour before the meeting was to start. That time turned out to be mostly wasted. There was no way I was going to get any deep work done in that time. Sure, sometimes I could do some idle research, stuff that I could get into and out of quickly, but largely it was wasted time.

I could have just spent all my time in the office, but the open office plan environment is not conducive to deep work, even without direct interruptions (which there would be, too).

I guess the best solution for me would have been to arrive at the office 5-10 minutes before the meeting start time, but it was surprisingly hard to get that right, even with a predictable commute time (bus or walking).

On top of all that, I would occasionally completely miss meetings if I was in a flow state before that. I would somehow miss desktop and phone notifications that a meeting was coming up. I guess the solution to that would have been to set an actual alarm on my phone, something that I couldn't ignore, but I never ended up doing that for some reason.

Yep. For me, a meeting is half a day lost.

Oh, half an hour till that meeting, better not start anything serious.

And some smartypants will schedule meetings with 1h gaps between so absolutely nothing will get done

I've been trying to articulate why switching our OSS project from using public GitHub issues to private Jira issues (behind auth that requires logging in with 2FA again every 48 hours) is bad for programmer productivity. Upper management quite literally doesn't believe me. To them, "flow state" and the visceral pain getting pulled out is it causes is completely fictional.

I agree with you on that from the perspective of a senior developer. Far too many times people would come and interrupt me for something that they could have figured out themselves. Sure, it might have been something that I knew offhand and could tell them in 15 seconds, but would take them 2 minutes to research. But it was so very very frustrating to me.

I would get that with junior folks for deeper questions, even, stuff that would take 15-30 minutes to go over with them. I started asking them questions up-front, to try to determine what research and steps they'd taken on their own before asking me. I would generally be pretty gentle with my "go away" patter: "ok, so through the questions I've asked you, I think you should have a few avenues of research that you can pursue on your own; if you're still stuck after that, let me know". But the repeat offenders who just didn't seem to get it... sigh. It was also clear to me from how some of my questions were answered, that often I wasn't suggesting anything they didn't already know; they just didn't feel like doing the research on their own, and wanted someone else to tell them how to do their jobs. So exhausting.

I find these moments to be teaching a person to learn to fish. Not only tell them how to get the answer, but then ask questions to see if they put in the work needed before coming to you. If it persists, maybe address it directly in a one on one meeting. If it's systemic across different people, maybe there is a documentation or organizational problem. Sometimes it's training the person to help themselves, other times it's seeing things that are broken that prevent them from helping themselves. I see this especially in highly complex systems where there are large and multiple teams, and the person needs to be able to work across these areas.

If you're a manager, and those requests are coming from your team, isn't that just part of the job and not an interruption? Giving direction and priority is not a full time job, surely.

I always assumed that would be an average, meaning it takes 5s for some and 2h for others.

Again, the actual quote is not specific enough, but that would be a very wild and easily dismissed claim.

Or those tasks just got prioritized lower and lower.

I think it's unfortunately very common for people to come here into the comments without having read the article.

I'll admit I do that sometimes, mainly to get an impression as to whether or not the article is worth reading in the first place, but sometimes I'll end up getting drawn into the discussion here, having not read the article at all.

But I can't imagine posting a toplevel comment without reading the article.

This is a valid point. But another aspect of interruptions is if they add to the experience of stress, and thus contribute to a toxic work environment.

> Good for the company.

Unless you’re C level, this thought shouldn’t even cross your mind. Look after yourself.

I think interruptions are similar for both types of work.

Physical work does require some concentration as well, but imagine on top of that having to take off all your gear and walk into some makeshift office on site, deal with whatever bullshit they're bugging you about, and then go back out to your spot and get ready all over again.

Interruptions for any reason other than a true emergency is just poor planning caused by bad management.

Knowledge and physical work is not the right distinction. It's fine for two broad categories, but within knowledge work there is — at the very least — the manager vs. maker schedule.

When I'm doing "physical labour" I find interruptions almost as impactful. (Though I'd accept that I'm the type to just overthink things.)

"Take this pile of stuff and pack it in that moving truck."

I mean, sure, I might just grab the nearest thing, carry it and set it in the first available spot and repeat... But usually I'm putting a little more into it than that. Unless the destination is comically oversized, I need to make sure I'm making good use of the space. Even just packing boxes, putting a 150lb plastic tub full of ammo on top of a bunch of boxes full of dishes is... probably not going to turn out well at the other end.

I started by doing a quick walk through and looking at labels to figure out roughly what I have to deal with. If the boxes are somewhat standard, I've probably found a pattern to how they best fit to minimize wasted space. I may have some piles I'm working on on the side that are "very heavy" or "very light" to try and at least roughly sort them into "heavy on the bottom light on top". That also lets me cut down the number of trips since the "very light" boxes can probably be carried two or three at a time. The boxes stack 6" short of the top of the truck, so I should probably put these bed frame boards aside and once I have solid base of boxes I can slide those right on top...

So I'm walking around with the context of a couple of temporary piles, a couple things to go once I have the right spot for them, the rough shape of the source pile, vaguely what the inside of the truck looks like and what I need next for each stack, the next couple moves to make, and more all floating around in my head. Losing that is not without consequence.

"Get all these pallets off of the loading dock and into the store."

They probably weren't put into the truck and then unloaded based on any sort of knowledge of the layout of the store at all. And each individual stack of pallets was stacked for the lower one supporting the upper one sufficiently to survive a semi ride across the continent. (If you have two pallets of TVs and two pallets of toilet paper, you don't make a stack of each.) I could just grab the next stack, set it down where the first pallet belongs, grab the one on top, drive that to where the next belongs...

Or I could see "Hey there's two stacks here with TVs on the bottom and toilet paper on top. Those are opposite corners of the store. Let me spend a few seconds on the loading dock to restack these, and cut my driving way down."

I _could_ take this one with that one I see back there... but that one way back there is deep into the pile, and I've got limited "scratch" space to drop stuff. In all likelihood, if I don't just take this one now and eat the drive across the store, it's going to just be in my way for the next hour. Screw it.

If someone pulls me on to something else for a bit and I lose all the context, the next few steps I have planned, why I put that pallet to the side for now, etc... yeah, I'm gonna come back and operate less efficiently.

(And if you're thinking "well yeah but you're clearly an insane person" or "that's just bringing a CS background to forklift driving"... I had and have no CS background, and while I picked this stuff up way more intuitively this was something the "lifers", who in some cases had literal brain damage, were teaching to the new guys.)

"Count these recyclables by type and write down the number on this paper."

Okay this one there's no real crazy context to hold in your head, but you definitely get into a flow where your brain is almost checked out besides your hands moving containers and your brain keeping tally marks.

While I _can_ do basic math, there were a number of slightly embarrassing incidents where I'd done things like looked at a paper where I'd written "100", "100", "12" and decided that added up to "112" because I was so deep into "put containers in fingers, put in correct bin, make tally mark" apparently the math part of my brain had fully turned off.

> Knowledge work is not the same as physical work.

I do disagree. I don't mean to give you any shit about it though. If I hadn't started in it when I was 14 years I have no doubt I'd be here emphatically agreeing with you.

I think the main difference between "knowledge work" and "labour" is basically where the floor is to get the job done.

Anyone can move boxes from A to B--a good mover can do it much faster, fit more things in less space and get it to the other end in one piece much better.

Anyone not a complete idiot could be shoved in a backhoe and dig a hole without tipping it over. An experienced operator can do shit that borders on magic. (I mean, go look up some YouTube videos.)

Anyone can grab a tub of drywall mud and a can of paint and patch and repaint a room. A good drywaller/painter can take your wavy wall full of holes and turn it into a piece of glass.

Anyone can glue two pipes together. A good plumber can look and get a picture of the entire plumbing _system_ the house and figure out how the gurgle in that drain means there's an air lock and the plumbing vent on your roof is blocked. Or figure out that the reason your reverse osmosis system's pump keeps cycling on and off is because the transformer supplying it power has failed. (Yeah, I had a plumber pull out a bench power supply to fix my water system...)

Anyone can drive a forklift around. Someone with a minimal amount of sense and training can do it safely. An experienced operator is trying their best to optimize for some sort of traveling salesman problem involving the locations being queued in multiple stacks on a 2D plane with imperfect knowledge.

Someone at the floor of these roles... yeah, they haven't put enough thought into it for an interruption to matter. But if you ever deal with a "really good" mover, equipment operator, tradesperson, or anything else... that shit makes a difference.

The difference between a knowledge worker and a labourer is that no one's asking their forklift driver making $35k/yr how to make their job more efficient. They're asking their MBA making $150k/yr and software engineer making $300k/yr to create a program to design software to give the driver a more efficient route.

This is the source: https://news.gallup.com/businessjournal/23146/too-many-inter...

Gloria Mark said it in an interview in 2006.

Original source: <http://web.archive.org/web/20131030072159/https://heeris.id....>

I've been cited on Wikipedia once... and that's when I truly understood why you can't trust Wikipedia.

So is it on xkcd?

When some of those plates are unnecessary, it causes problems. I may be responsible for 3 plates worth of work, and be capable of 6 on a really good day. On a bad day though, 3 may be all I can handle and an interruption=dropped plate.

Works fine for me?

  > The way it is written implies the research is all wrong

I think you read the words from the article but I think you missed their entire point. I think you're actually demonstrating their thesis...

Someone yelling at me costs themselves however long it takes to hire my replacement. I did my time working for overgrown children, I won't put up with that kind of nonsense anymore.