🔍 Read the full analysis: Workers Juggle Apps 900 Times A Day. What Role Can AI Play? on ThorstenMeyerAI.com
Get business pricing on office and shipping supplies
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
A Harvard Business School preprint analyzed 103 million application events from 1,017 employees at eight companies in India. Workers switched apps a mean of 908 times per day; AI use was linked to fewer switches immediately after use but about 200 more switches on AI-use days overall. The observational study cannot establish whether AI caused either pattern or whether app switching affected work quality.
A Harvard Business School preprint analyzing 103 million application events found that 1,017 office workers switched apps a mean of 908 times per workday, while the relationship between AI use and switching varied by timeframe. Workers switched less in the ten app uses after an AI interaction, but made about 200 more switches on days they used AI—an observational pattern that does not show AI caused the change.
The study, 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. The researchers examined records covering 62,693 person-days, collected between November 2025 and January 2026 at eight organisations in India. The paper is a preprint under peer review, and its authors warn that findings may change during that process.
Across the sample, workers changed applications a mean of 908.3 times a day; the median was 814. The middle half of employees recorded between 547 and 1,130 switches per workday. An application remained active for a mean of 27.8 seconds before the next switch, although the median was 3.3 seconds. The paper counts a switch when the active application changes, not when a worker opens another document within the same app.
The AI results point in different directions depending on the measurement. In the ten app uses following an AI use, employees used fewer apps, switched less, remained in each app about half a second longer and moved through apps in a more predictable pattern. Across whole days, however, AI-use days involved about 200 additional switches, and days with more AI minutes were more fragmented. The authors say the data cannot establish whether AI changed work patterns, whether workers used AI during days that were already different, or both.
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 Link to App Switching
The findings complicate the idea that adopting AI automatically reduces workplace interruptions. The immediate sequence after an AI interaction looked less fragmented, but the broader daily comparison went the other way. Those results can coexist: a short period after using a tool is not the same measure as an entire workday.
For employers and workers, the paper is a reason to distinguish tool use from work outcomes. It measures application activity, not whether employees produced better work, finished tasks faster, or felt more focused. It also does not test an AI product against a control group. The study therefore cannot support a claim that buying an AI system will improve concentration—or that AI itself makes the workday more fragmented.
The researchers also found that day-to-day variation within the same person accounted for 44.6% of variation in fragmentation, compared with 35.8% linked to differences between people and 19.6% between organisations. This suggests that daily conditions may matter substantially, although the analysis does not identify which conditions are responsible. The paper’s results can inform questions about meeting load and workflow, but are not a direct measure of performance or wellbeing.
Products Worth Considering

FLEXISPOT 48 x 24-Inch Solid Rubberwood Electric Standing Desk, Black Frame

Dell UltraSharp U2424H 24-Inch FHD Monitor

Canon SELPHY CP1500 Compact Photo Printer Kit with 2x KP-108IN Ink and Paper Sets
As an affiliate, we earn on qualifying purchases.
How the Workday Data Was Collected
The records came from Flowace, a commercial workforce-analytics service used by all eight participating organisations. According to the paper, its software records when an employee changes the active application, along with information such as the app name and window title. The data were anonymised before researchers received them, under a data-use agreement; Harvard’s human-subjects committee approved the research protocol.
The sample included companies in software, accounting, financial services, e-publishing, law and marketing. Employers, rather than the researchers, decided which applications counted as productive. The paper states that it received no external research funding, and that Flowace provided no funding or commercial support and had no role in the study design, analysis or interpretation. One author disclosed equity in an unrelated private company that uses AI to create enterprise digital twins.
The authors compare the measured app-use periods with earlier attention research, including estimates of about 2.5 minutes two decades ago and 47 seconds in the last decade. Because research methods differ, these figures are best treated as broad context, not a directly comparable measurement of a steady decline.
“The manuscript is currently under review, and the findings may change based on the peer review process.”
— Vaid and Whillans, in the preprint
What the Study Cannot Establish
The analysis is observational, not an experiment. It shows associations between AI use and switching, but cannot tell whether AI caused the patterns. The authors say workers may turn to AI when their days are already consolidating, AI may reorganise the day, or both could be happening.
The evidence also has limits in scope. Participants were employees at eight organisations in India, tracked by employer-used monitoring software from November 2025 through January 2026. The study did not cover phone use, and the paper does not establish whether switching improved or harmed work quality. The authors’ definitions of productive apps depended on employer choices.
It remains unclear whether the findings generalise to other countries, industries, workplaces or AI tools. The preprint is still under peer review, and its authors say the results could change. The study does not establish that app-switching totals alone reflect distraction, lost productivity or employee experience.
Peer Review and Further Evidence
The paper was posted to arXiv on July 7, 2026, and revised on August 7, 2026; it remains under peer review. The authors say the manuscript may change through that process. No further publication date or follow-up study is specified in the supplied material.
To determine whether AI changes fragmentation, future research would need stronger comparisons, such as tracking workers before and after a tool is introduced or comparing similar teams with and without access. Researchers would also need to assess outcomes beyond app activity, including task completion, quality and employee experience. Until then, the study supports treating AI’s relationship with switching as unresolved rather than as evidence that an AI tool improves focus.
Key Questions
Did workers switch apps 900 times a day?
The study reports a mean of 908.3 switches per day across its sample. The median was 814, and the count records changes between applications rather than every change of document or window.
Did AI reduce app switching?
The results depend on the timeframe. Workers switched less in the ten app uses after an AI interaction, but made about 200 more switches on AI-use days overall. The observational study cannot show that AI caused either pattern.
Who and where did the study cover?
Researchers analyzed activity from 1,017 employees at eight organisations in India, across software, accounting, financial services, e-publishing, law and marketing. Records covered November 2025 to January 2026.
Is the Harvard study peer reviewed?
It is a preprint under peer review, not a completed peer-reviewed publication. The authors say the findings may change during review.
Does the study show AI improves productivity?
No. It measures application activity and switching, not whether work quality, task completion or productivity improved. It also does not establish that AI caused the observed patterns.
Source: ThorstenMeyerAI.com
