Nobody notices the theft because no single instance is worth mentioning. The file that takes forty seconds to open. The login that fails once a fortnight. The report someone rebuilds by hand every Monday because the export is not quite right. Each one is too small to raise, so nobody raises it, and the total never appears anywhere. Outages get fixed because they are loud, and their cost is well understood, which we cover in how internet downtime affects productivity. This is the quiet version, and across a small business it routinely costs more hours than the outages do.
Where the Hours Actually Go
- Friction beats failure. A system that is down gets fixed; a system that is slow gets tolerated for years.
- Workarounds are invisible by design. People build them to avoid complaining, which is exactly why nobody can see them.
- Context switching is the multiplier. The interruption costs two minutes and the recovery costs fifteen.
- You cannot ask your way to the answer. People systematically underestimate friction they have adapted to.
- Counting for two weeks beats guessing for a year, and it is the cheapest diagnostic available.
Written for owners, operations leads and office managers at businesses between ten and five hundred people.
The Four Common Thieves
Most of the loss concentrates in a small number of patterns, and they are recognisable once named.
Waiting on Systems
Slow application launches, sluggish file access, reports that take minutes to render, machines that need five minutes before they are usable in the morning. Individually trivial. Multiplied by every person, every day, for a year, it becomes a headcount’s worth of time. The tell is that people have adapted: they make coffee while something loads, they start the slow thing first and do something else, they know which times of day are worse. Adaptation is the signal that the loss has become permanent.
Authentication Friction
Password resets, sessions that expire mid-task, multifactor prompts arriving on a device somebody left at their desk, systems that each need separate credentials. Security is not the problem here and loosening it is not the answer, and much of the fix is configuration rather than purchase, as in improving productivity through Office 365 management. Badly implemented security is the problem, and single sign-on plus a managed password approach usually removes most of this without weakening anything.
Manual Bridges Between Systems
Someone exports a file, reformats it, and imports it somewhere else, every week, because two systems that ought to talk do not. These are the most expensive losses and the easiest to miss, because the person doing it has stopped thinking of it as a problem. It is simply part of their job now. Ask anyone what they do that a computer should be doing and you will find these quickly.
Being Interrupted to Fix Someone Else’s Technology
In most small businesses one or two capable people have become the unofficial help desk. Their own work is constantly interrupted, the interruptions are never recorded anywhere, and their cost shows up as that person being mysteriously behind. This is the single most common hidden cost we find, and it is usually invisible to the people paying for it.
How to Measure It in Two Weeks
Asking people how much time they lose produces bad data, because they have normalised the friction. Counting produces good data, and two weeks is enough.
Count Events, Not Minutes
Give everyone a simple tally: a shared sheet with one row per person and columns for waited on a system, could not log in, did it by hand, helped someone else with IT. One mark per occurrence, no estimating, no notes. Estimating minutes invites judgement and the judgement is what corrupts the data. Counts are easy and honest, and you can attach durations afterward from a handful of timed samples.
Watch for the Workaround Nobody Mentions
Spend an hour sitting with two or three people while they work, and ask them to narrate what they are doing. The workarounds surface within minutes, because they are visible in action even when they are invisible in conversation. The sentence you are listening for is a version of “oh, I always have to do this bit twice”.
Separate the Fixable From the Structural
Some of what you find is a purchase decision, some is a configuration nobody ever revisited, and some is a process that grew around a limitation that no longer exists. That last category is the most rewarding, because removing it costs nothing but attention. Sort the list before you spend anything, because the cheapest wins are usually not the ones that feel most urgent.
Put a Number on It
Multiply the counts by a representative duration, multiply by fully loaded hourly cost, and annualise. The number will be uncomfortable and it will also be conservative, because tallies undercount. That figure is what turns a vague sense that things are slow into a decision anyone can approve.
What to Fix First
Ranking matters more than ambition here, because the temptation is to start with the most visible annoyance rather than the largest cost.
Start With Whatever Touches Everyone Daily
A thirty-second improvement to something every person does ten times a day outranks a dramatic fix to something one person does weekly. This ordering is unintuitive because small universal friction generates no complaints, while a single person with a painful workflow generates plenty.
Retire the Manual Bridges
These convert directly into hours and usually into fewer errors, since every manual transfer is an opportunity to transpose something. Integration is often simpler than assumed, and where a real integration is not available, a scheduled export is still better than a person doing it by hand. Where the systems are Microsoft ones, a surprising share of these bridges already have a supported alternative, as we set out in top Microsoft 365 integrations that boost productivity.
Make the Help Desk Official
Whatever the arrangement, the interruptions should land somewhere they are recorded rather than on the shoulder of whoever is nearest. That alone recovers the hidden hours of your most capable people and produces, as a side effect, the first real data you have ever had about what breaks. It is also the point at which outside support starts to pay for itself, which is the subject of the impact of managed IT services on employee productivity.
Fix the Login Experience
Single sign-on, a managed password approach and sensibly configured multifactor remove a category of daily friction without reducing security. This tends to be one of the higher-return changes available, and it is one people notice immediately, which helps with everything you want to do next.
Why the Loss Stays Invisible
Understanding why nobody reports this is the difference between running the count once and building a habit that keeps the friction from coming back.
Small Costs Do Not Clear the Reporting Bar
Every organisation has an unwritten threshold for what is worth raising, and it sits well above forty seconds. Anything under that line accumulates permanently, because the mechanism that would surface it never triggers. The count works precisely because it lowers the bar to a tally mark, which costs nothing to make and therefore captures what a conversation never would.
People Protect Their Colleagues
A surprising amount of friction goes unreported because naming it feels like naming a person: the colleague who set the system up, the manager who chose it, the provider somebody recommended. Framing the exercise around events rather than causes removes that, and it is worth saying out loud when you introduce the tally. You are counting what happened, not who is responsible.
Adaptation Looks Like Competence
The person who has built an efficient workaround looks capable rather than obstructed, and they are usually proud of the workaround, reasonably. That is why sitting with people beats surveying them. In conversation the workaround is background; in action it is visible in seconds.
Nobody Owns the Total
Individual managers see their own team’s friction and assume it is normal. No single person holds the whole picture, so the total never gets assembled and therefore never gets funded. Assembling it is most of the work, which is why the number at the end of the count changes conversations that years of complaints did not.
Keeping It From Growing Back
Friction returns, because every new system, integration and policy adds a little. The firms that stay ahead of it treat the measurement as recurring rather than as a project.
Repeat the Count Quarterly
The same four columns, two weeks, once a quarter. It is cheap, it takes almost nothing from anyone, and it tells you whether the last round of fixes actually landed. A comparison against the same categories is far more convincing than an impression that things feel better.
Ask the Friction Question When Buying
Before adopting anything new, ask which existing step it removes and which new one it adds. Most tools add at least one, and knowing which one in advance is the difference between a planned cost and a surprise. A tool that adds a login is not free even when the licence is.
Give the Workarounds Somewhere to Go
A standing, low-ceremony place for people to record the thing they have to do twice keeps the tally alive between counts. It works only if things on the list visibly get fixed, so pick one item a month and close it. The list dies quietly the first quarter nothing on it moves.
Frequently Asked Questions
How much time does the average small business lose to IT friction?
The honest answer is that it varies enough that any published average is a poor guide to your own situation, which is why counting for two weeks beats citing a study. What is consistent is that the total is larger than anyone expects and that most of it sits in small recurring events rather than in outages.
Why not just ask staff how much time they lose?
Because people underestimate friction they have adapted to, and because nobody wants to sound like they are complaining. Counts of events are honest in a way that estimates of minutes are not, which is why the tally asks for marks rather than judgements.
Is this really an IT problem or a process problem?
Both, and the split is usually about half. Some friction is a system that needs fixing and some is a process that grew around a limitation nobody has revisited. The process half is often free to fix, which is a good reason to sort the list before spending anything.
What is the cheapest fix that usually pays off?
Making the informal help desk official, so interruptions land somewhere recorded rather than on the nearest capable person. It recovers that person’s hours and it produces real data about what actually breaks, which makes the next decision easier.
How long before we see the difference?
Friction fixes show up faster than most technology work because people feel them immediately. A login improvement or a retired manual export is noticeable within a week. The measurable version takes a second two-week count a quarter later, comparing the same tally against the same categories.
Who Is Behind This Guidance
Our team supports small and mid-sized businesses on managed IT, which means we usually meet this problem as a symptom rather than as a complaint. Someone says the team feels stretched, or that a capable person is inexplicably behind, and the count afterwards finds the hours. The consistent finding is that the biggest losses were never reported to anyone, because each individual instance was too small to be worth a conversation. Matt Rosenthal, Mindcore’s CEO, keeps the practice focused on making technology costs legible to the people accountable for them, and hours lost to friction are the least legible cost most businesses carry. That is the standard this guide is written to.
Count Before You Buy Anything
Run the tally for two weeks with four columns and no estimating. Sit with two or three people for an hour each and listen for the sentence about doing something twice. Sort what you find into purchases, configuration and process, and notice how much of it is the third. Multiply the counts by a sample duration and your loaded hourly cost so the number is real. Then fix whatever touches everyone daily first, retire the manual bridges, and give the interruptions somewhere official to land. Most businesses skip the counting and buy something instead, which is why the friction survives the purchase. If you would rather have the count run properly and the results costed, our managed IT services start exactly there, and the support side is in help desk services. Bring two weeks of tallies to a free strategy call and we will tell you which line is costing you most.
