🔍 Read the full analysis: Why Workers Switch Apps 900 Times A Day—and Whether AI Can Help on ThorstenMeyerAI.com
Get hardware and tech essentials delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
A Harvard Business School preprint analyzed 103 million application events from 1,017 employees at eight companies in India and found an average of 908.3 app switches per workday. App use was less fragmented in the ten uses following an AI interaction, but workers switched more on days they used AI; the observational study cannot show that AI caused either pattern.
A Harvard Business School preprint analyzing 103 million application events from 1,017 office workers found they switched between apps an average of 908 times per workday. The study’s AI findings point in two directions: workers switched less in the ten app uses after an AI interaction, but made about 200 more switches on days when they used AI. The researchers did not test whether AI caused either pattern.
The paper, “Digital Fragmentation and Generative AI Use Across 103 Million Application Events,” was written by Sumer S. Vaid and Ashley V. Whillans of Harvard Business School. It draws on records from 1,017 employees at eight organisations in India, collected from November 2025 through January 2026. The dataset covers 62,693 person-days and 56.9 million recorded app switches. The researchers measured a switch when a worker moved to a different application; moving between two documents in the same app counted as one use.
Switching varied widely: the median worker made 814 changes per workday, and the middle half ranged from 547 to 1,130. A work app held attention for an average of 27.8 seconds before the next switch, though the median was 3.3 seconds. These measures describe app activity, not the quality or productivity of the work. The paper says 73% of switching was between work applications.
The AI analysis compared activity around recorded uses of tools including Claude and four other generative AI systems. In the ten app uses after an AI interaction, workers used fewer apps, switched less, stayed about half a second longer per app and followed more predictable sequences. Yet on days with any AI use, workers made roughly 200 more switches per day; days with more AI minutes were also more fragmented. Those findings can coexist because one measures a short sequence around AI use and the other compares whole days.
App switching and AI · one moment
What happens around one AI query?
The study zoomed in on every AI use: the ten app uses before it, and the ten after. Up close, work calms down a little. Zoom out to the whole day, and the picture changes.
- 1 · A workday of switchingWorkers changed apps 908 times a day on average. One app held their focus for 27.8 seconds on average, 3.3 seconds at the median.
- 2 · AI comes in at a busy momentPeople used AI a median of 13 times a day, for a median of 7 seconds. Fragmentation peaked in the hour of AI use.
- 3 · The next ten app usesFewer different apps, fewer switches, about half a second longer per app, more predictable. Real, but small.
- 4 · Zoom out: the whole dayOn days with AI use, the same people made about 200 more switches. Whether AI calms work or busy moments bring AI: unknown.
Source: Sumer S. Vaid, Ashley V. Whillans (Harvard Business School), “Digital Fragmentation and Generative AI Use Across 103 Million Application Events”, arXiv:2607.06681v2, preprint under review. Blocks are schematic, not to scale.
Graphic: Powered by Thorsten Meyer AI
AI’s Mixed Signal on Switching
The findings caution against treating an AI assistant as a proven focus tool. The study records an association: it does not show that introducing AI would reduce interruptions, improve concentration or raise output. Workers may use AI during already busy days, or its use may shift how tasks are organised; both may be happening.
The day-to-day pattern may also matter more than a worker’s identity or employer. The authors attribute 44.6% of variation in fragmentation to swings within the same person from day to day, compared with 35.8% to differences between people and 19.6% to differences between organisations. For employers, that points to a reason to examine workload and communication patterns across the week rather than assuming frequent switching is simply an individual habit. The data do not establish which workplace changes would help.
Products Worth Considering

LG 27G810A-B 27-inch UltraGear 4K UHD IPS Gaming Monitor

Logitech MX Master 4 Ergonomic Wireless Mouse with Haptics – Graphite

Apple MacBook Air 15-inch Laptop with M5 Chip, 16GB Memory, 512GB SSD, Midnight
As an affiliate, we earn on qualifying purchases.
How the Workday Was Measured
The paper appeared on arXiv on July 7, 2026, and was revised on August 7. It labels itself a preprint under peer review and warns that findings may change. A November 18 post helped circulate the claim that app switching drops “after AI,” but that wording captures only the short-window result, not the higher switching observed across AI-use days.
Activity records came from Flowace workforce-analytics software, which employers already used on employees’ computers. It logged the active application and timing, among other information. The employers determined which apps counted as productive. The researchers say they received anonymised data under an agreement, and that their human-subjects protocol was approved. The paper reports no external research funding and says Flowace had no role in study design, analysis or interpretation.
The sample included software, accounting, financial-services, e-publishing, legal and marketing organisations, all in India. The paper compares its brief app-use durations with earlier attention research, but differences in methods mean those figures should be read as broad context, not as a directly measured long-term decline.
““The manuscript is currently under review, and the findings may change based on the peer review process.””
— Sumer S. Vaid and Ashley V. Whillans, paper authors
What the Data Cannot Establish
Because the study is observational, not an experiment, it cannot determine whether AI led to more switching, whether workers turned to AI on especially fragmented days, or whether other factors shaped both. The authors do not report whether AI use improved the work, reduced time to complete tasks or changed workers’ wellbeing.
The evidence also has limits beyond causation. It comes from eight organisations in one country, relies on employer-installed monitoring software and covers computer applications rather than phone activity. The authors’ measure captures application changes, not every interruption or the purpose behind a switch. The employer-defined productivity categories may also affect how the activity is classified. Results may change during peer review.
Peer Review and Workplace Evidence
The next research milestone is peer review of the preprint; the supplied material gives no date for a final publication. Any revision could alter how the findings are reported. Further studies would need to test AI use under controlled or carefully compared workplace conditions and measure outcomes such as task quality, completion time and employee experience—not just app activity.
For now, the paper supports describing frequent switching and a mixed association with AI, not promising that an AI product will improve focus. Employers considering workflow changes can track when interruptions cluster and test changes against clear work outcomes, while recognising that this study does not show which intervention works. App switching is not itself a measure of productivity, and the current evidence does not settle whether AI helps workers concentrate overall.
Key Questions
Did the study find that workers switched apps 900 times a day?
It reported a mean of 908.3 app switches per day across the recorded workdays. The median was 814, so the mean does not describe every worker’s typical day.
Did AI reduce app switching?
The results were mixed. Switching was lower in the ten app uses after an AI interaction, while workers made about 200 more switches on days when they used AI. The study cannot establish that AI caused either result.
Who and where did the study cover?
The researchers analysed computer activity from 1,017 employees at eight organisations in India, spanning sectors including software, accounting, finance, publishing, law and marketing. The records were collected from November 2025 to January 2026.
Is the research peer-reviewed?
No. The paper was posted as a preprint under peer review and says its findings may change. It should not be treated as a final peer-reviewed result.
Does frequent app switching mean workers are less productive?
The study measures application activity, not work quality or output. It does not establish that more switches mean lower productivity, or that AI improved performance.
Source: ThorstenMeyerAI.com
