The standup ends, you click the same reporting link, and it hangs on the same spinning circle it hung on yesterday.
You open the spreadsheet workaround, paste the numbers by hand, and the thought arrives on schedule: maybe I should just look at what else is out there.
That thought has a headline behind it. The 91% figure making the rounds is not in any 2026 survey supplied here. The real shares differ by survey, because one asks technology professionals and another asks desk workers.
What pushes is digital drag: friction that repeats across the tools you use all day. It shows up as minutes in your week, not as a national percentage.
By the end you will have a one-week tech log, three thresholds that sort fixing from searching from leaving, and one move for today: a 15-minute check-in with your manager, or one job search opened.
Tech That Creates More Work Than It Removes

The tech problem behind the exit talk is digital drag: friction that repeats across the tools you use all day. Duplicate data entry, a system that stalls, and rechecking what an AI wrote all count. A single outage does not.
The difference is repetition. One glitch costs nothing you remember. A weekly pattern means the same task, the same workaround, and the same lost minutes come back, week after week, until the task itself starts to feel like a losing fight.
Check the headline first. No 2026 survey in this article's sources reports 91%, so treat that number as a survey artefact rather than a measurement of you.
Standley Systems did find 85% of desk workers hit a tech issue that slows them down at least once per workday. Most, in other words, not 91%.
Start the one-week tech log. The columns, in this order: Date | Task | Minutes lost | Workaround used | Happened before (Y/N). Write one row at the moment the workaround happens, not from memory on Friday.
The log comes first because it turns a mood into a sortable list. Once the minutes sit in a column, the biggest repeatable line is the single issue worth acting on. Everything else waits.
⚠️ COMMON MISTAKE
Never Rebuild the Log on Friday
A log written from memory records the worst moment, not the pattern. One row per workaround, typed while the tab still spins, keeps the minutes honest and gives you a number a manager cannot argue with.
The Daily Tax Adds Up Before Burnout

The Standley Systems figures describe at scale what the minutes column already shows on your desk. Nearly 30% of desk workers lose an hour or more of productivity each week. About 6 in 10 first try fixing issues themselves.
That self-fix habit hides the cost. Three-quarters avoid contacting IT at least sometimes. So the friction never becomes a ticket, and it never appears in any system a manager reads.
The sum is easy to check. A 10-minute workaround every day is 50 minutes a week. Three logged rows of 10, 15 and 20 minutes come to 45 minutes. Two daily workarounds clear an hour.
To convert your own log, add up the minutes column for seven days and divide by 60. Hold that figure against the one task that shows up in the most rows.
Then put one question to IT or a manager about that task:
The Tool Gap Splits Bosses and Staff

Management skips the fix because management does not see the problem. WalkMe's 2026 State of Digital Adoption report found 88% of leaders believe their employees have adequate tools.
Among employees themselves, 21% said the same. One stack looks fine from the executive floor and broken from the desk.
The gap has a second layer now. WalkMe also found only 12% of workers are fully confident AI tools grasp the context of their work.
So correcting the output is not sloppy use on your part. It is built into the job, and it belongs in the log.
WalkMe warned of a third: 45% of workers used unapproved AI tools in the past 30 days, and 36% of those users put confidential data into them. Keep customer lists, screenshots and account details out of chatbots and tickets.
A report a leader can act on has four parts: the task, the minutes lost, the workaround used, and the deadline at risk. That shape of evidence moves a manager who already believes the tools are adequate.
AI Pressure Turns Productivity Into a Trap

The tool gap only explains why a fix never arrives. The other push comes from what AI now promises on top of broken tools: more output, faster.
Modern Health's 2026 survey found 67% of employees say AI raised the productivity expectations on them, and 64% of that group tie extra stress straight to those expectations.
Upwork's research on AI users found the heaviest AI users reporting burnout at 88%, and twice as likely to be weighing an exit. The people AI helps most are the ones closest to leaving.
The plain reason: AI rarely replaces a step. It adds review, correction and switching between systems. A task that saved 20 minutes can hand back 25 minutes of checking when the tool misses the context around the job.
The test: for one week, log one repetitive task in two columns – minutes the AI runs and minutes you spend reviewing and fixing.
Keep the tool only if your total drops. A higher number is a documented reason to switch it off, not a personal failure.
Keep the AI Tool or Switch It Off
Keep it
- Total minutes fall below the by-hand time
- You can name the check that catches its errors
- The task repeats enough to be worth the setup
- Output needs light edits, not rewrites
Switch it off
- Review takes longer than writing it yourself
- It invents names, dates or sources you must hunt down
- It lacks the context of your job and you re-brief it daily
- You keep it only because a leader asked for AI use
Burnout Is a Workload Story, Not a Buzzword

The moment the time AI frees gets filled with more work, the cost leaves the minutes column and lands on the body.
The burnout figures say the same: the strain is capacity, not any one app. ISACA's 2026 survey of technology professionals found 59% reporting work-related stress or burnout, and named workload as the top cause at 48%.
Lack of resources followed at 43%, unsupportive management at 42%. ITRM's Tech Anxiety Index 2026 asked a wider group and got the same direction: around 80% had felt anxiety about workplace technology, and 88% had felt frustration or stress from it.
Two surveys, two audiences, one pattern. Neither blames a single app. Both point at too much to do with too little backing.
Three signs tell you this has crossed from annoying to health: sleep lost to a workaround, Sunday dread about Monday's tools, and work that follows you home.
One boundary to set: the evening fix gets logged as a ticket for the morning and is not repeated that night.
Satisfied Workers Still Watch the Door

Because a workload problem can sit alongside a job people still rate highly. Health signals are strong exit evidence, yet the same surveys find most workers calling themselves satisfied, which is the puzzle worth untangling.
ISACA's 2026 survey of technology professionals found 76% were satisfied with their jobs, and 65% would still consider changing jobs within two years.
The Employee Benefit Research Institute's Workplace Wellness Survey found 55% of workers fantasize about quitting. In the same survey 60% report being satisfied, and 90% plan to stay with their employer at least two more years.
Both can be true because they answer different questions. Satisfaction describes the job; exit thoughts describe the day. A satisfied worker can carry a 50-minute daily tax without either number being wrong.
The split: a bad week changes when the tool changes. A bad job survives the fix and shows up again the Monday after. Ask one question: if this tool worked perfectly tomorrow, would I still want to leave?
The Market Is Quiet, So People Job-Hug

The timing is not a mood, and the numbers explain the gap. The Bureau of Labor Statistics puts the quits rate at 1.9% in August 2026, down from 3.0% in April 2022. The hires rate fell to 3.3% from 4.6% in July 2021.
Intent sits far ahead of action. According to Robert Half, 46% of white-collar workers planned to look for a different job in the second half of 2026, up from 27% a year earlier. The quits rate stayed at 1.9% through it all.
Monster found 48% of workers describe themselves as job hugging, staying put while they watch. In the Employee Benefit Research Institute survey, 33% said they were not confident they could quickly find another job if they left.
That is why an afternoon matters less than it feels. A resignation typed while the screen is frozen is priced against a hires rate of 3.3%, not against the afternoon. The check is cheap: refresh one saved search and message one person in your field.
The Words in This Argument
Security Rivals Pay When the Door Is Open

Pay, benefits and stability are what hold people in place, and none of them is the same thing as liking the tools. In the EBRI survey, 66% named pay and benefits as their reason to stay, and 64% named job security or stability.
Robert Half's reasons for hunting look different. Better benefits drew 47%, career advancement 43%, remote work just under 40%, and a higher salary only 35%.
The lists do not match. Benefits top both. Pay and benefits pull 66% toward staying, while salary pulls 35% toward looking. Security, 64% of the stay case, barely appears as a reason to hunt.
Write the stay list as five lines and score each one. Leave the labels exactly as they are and put a number after each.
Your stay list: Pay – / Benefits – / Security – / Flexibility – / One tech fix –
Score each line 1 to 5 for how much would have to change before leaving became worth it. The ranking gives you a real answer to "what would have to change." That turns the exit question into something you can check, not a mood.
One Fixable Issue Beats a Resume Rewrite

Before tech drag gets treated as an exit signal, a two-week test of one fix should come first. The test needs a written request, about the worst repeated line on the log, and that line is usually your cheapest open item.
Write it out with the log open in front of you, so every number comes from a row and not from memory. Fill the brackets, keep it under a hundred words, and send it to your manager or IT.
Prompt to copy
Write a short, plain request to my manager about one recurring tech problem, using these details: when I [task], [tool] fails about [X] times a week; each failure costs [Y] minutes and causes [Z]. Ask whether we can test [fix] for two weeks. Keep it under 100 words. State the task, the minutes and the impact only. Leave out blame, frustration and any confidential data. End with one question asking for a yes or no.
Fix the review in advance: one date two weeks out, and one measure of success. Either the workaround is gone, or the weekly minutes lost fall below 30.
If the fix lands, stay and log whether it holds for a month. Keep the exit question closed on unrelated bad days instead of reopening it over the next tool that stumbles.
🗓️ The Two-Week Test, Start to Review
Day 0: pick the line
Choose the row with the most minutes lost per week, not the one that annoys you most.
Days 1 to 7: log only
Log workarounds without changing anything. Anything you fix this week ruins your baseline.
Day 8: send it
Send the request Monday morning. A Friday send can sit unanswered past your test window.
Days 8 to 21: freeze one thing
Start no new tools and no desk moves. Two changes at once and neither gets credit.
Day 22: read the number
Recount the same rows. Success is under 30 minutes lost a week or zero workarounds.
Day 22 on: the tiebreak
No fix and no answer means escalate once with the log attached. Never escalate twice.
The Exit Decision Needs a Rule

A request that survives two escalation cycles is no longer a request. It is a record, and the record now decides. Three thresholds do the deciding, and the log picks which one you are standing in.
Threshold one: the fix arrives and holds for fourteen days. Lost minutes fall, the workaround stops being typed, and no exit decision is owed on one bad quarter.
Threshold two: the same task breaks after a second escalation with no fix date. Or the minutes now cost sleep and health. Start a quiet search now, one updated saved search, while the fixing continues.
Threshold three: the only real fix is less work, and management will not cut it. That is the case for leaving deliberately, and the log, the written request and both escalation dates travel with you.
The 91% headline is not a diagnosis of your desk. No supplied 2026 survey reports that figure, and the surveys that exist ask different groups different questions. So it cannot be a personal verdict, and it is not one.
Then the one action, chosen by your threshold: book a 15-minute check-in with your manager, or open one saved job search. One of the two, today, before the afternoon decides for you.
Frequently Asked Questions
Where does the 91% figure come from?
Not from any 2026 survey in this research. The real findings point the same direction but measure different things, like Standley Systems' 85% of desk workers hitting a tech issue daily. Treat 91% as a headline, not a finding about you.
How long should I track my tech problems before deciding anything?
Seven days is enough to see a pattern. Log every workaround with its minutes as it happens, then total the column and divide by 60. Two weeks of the same result is stronger evidence than a single awful afternoon.
Will reporting a tech issue actually get it fixed?
Sometimes, and only when the report is specific. A task, a failure count, the minutes lost and a deadline at risk gives a manager something to act on. Vague complaints about slow systems do not survive a budget meeting.
Can I use AI to make up the time I lose to tech problems?
Only if you count the review minutes too. Log how long the AI takes and how long your checking takes for one week. Keep the tool if the total drops, and switch it off if checking costs more than writing did.
Is it worth job hunting if most workers say they want to leave?
The market decides that, not the crowd. The quits rate was 1.9% in August 2026 against a hires rate of 3.3%, which is why many people job-hug. Update one saved search and talk to one person in your field before you do anything bigger.
What if the tech problem is really about my workload?
Then say so plainly. ISACA found workload leads burnout causes at 48%, above resources and management. A tool fix will not solve a capacity problem, so name the capacity issue in the same request.
Which Door Your Own Log Opens
Fix and monitor if it is one documented tool issue that survives a two-week test.
Start a quiet search if the same issue has outlasted two escalation cycles or is costing you sleep and health.
Leave deliberately if the fix would need a workload change management will not make.
This morning's third freeze is now a row with a date and the workaround you used. The 91% is a survey artefact that belongs to no single desk.
The tool is what can change this month; the job is what two weeks of rows will settle.
Nobody leaves over a frozen screen. They leave over the fortieth one, and that number is finally written down.