How Much Time People Spend Writing and Typing at Work

People who work with words spend a large part of the week writing, and email alone takes a big share of it. Across the 15 data points ¹ we collected from 8 sources ², email has a median of 8.8 hours per week ³. Writing documents has a median of 10.7 hours per week ⁴. The two studies of all writing and typing have a median of 13.6 hours per week ⁵.

These are medians of published figures for very different groups, not the average worker's week. Read them as a plausible range for people whose jobs depend on text, and use the sections below to find the group closest to you.

Email takes a median of 8.8 hours a week, but the range is wide

The 8.8 hours ³ median hides a spread from 17 hours per week ⁶ to three-plus hours per week ⁷. The low end comes from software that logged the time of RescueTime users. The high end comes from consumers who told Adobe how long they spend checking work email.

Two rows come from Microsoft's logging of its heaviest email users: 8.8 hours per week ⁹, and the same finding reported by a news outlet, 8.8 hours per week ¹⁰. They are one measurement counted twice, because we recorded both the original and the press copy. That pulls the email median toward the Microsoft figure. It is also the top quarter of users, so it is not an average.

Adobe's personal email row, two-plus hours per week ⁸, is not work writing at all. Keep it out of any estimate of your working week.

Method explains most of the gap between email figures

Self-reported surveys give higher numbers than software logs. The logged figures are 17 hours per week ⁶ for RescueTime users and 8.8 hours per week ⁹ for the heaviest Microsoft users. The self-reported Adobe figures are three-plus hours per week ⁷ for work email and two-plus hours per week ⁸ for personal email.

Several things push the survey numbers up:

• Both Adobe rows are lower bounds. Respondents chose "three-plus" and "two-plus" hours a day, and we converted the daily answer to a week.
• Adobe asked about time spent checking email, which includes glancing at an inbox. Logging software only counts time in the program.
• The Adobe sample is general consumers, not a defined workforce, and the post does not say whether weekends are counted.
• The RescueTime figure is itself a conversion from a share of the workday. It cannot tell reading from replying, and it misses time away from the screen.

Neither approach is wrong. They measure different things: what people believe they spend, and what a tool can see.

Writing documents ranges from 5.6 to 20.4 hours depending on the job

The documents rows cover professional writers and academics, and the gap between them is large. The lowest are faculty at public four-year colleges, 5.6 hours per week ¹², and private four-year colleges, 5.7 hours per week ¹³. Faculty at public universities report 10.7 hours per week ¹⁴, and at private universities 14.5 hours per week ¹⁵.

The pattern follows how much each institution expects research. Colleges that emphasize teaching report the least writing, and research universities report the most. All of these come from the same national faculty survey, so they compare cleanly with each other.

The highest row, 20.4 hours per week ¹⁶, comes from a survey of business people who already write at least a set amount each week for work, with email excluded. That selection guarantees a high average. It describes heavy writers, not the typical office worker.

Share-of-week estimates for whole professions reach 42 hours

Two rows state all writing and typing as a share of the work week. We converted each to hours. Programmers sit at 26 hours per week ¹⁷, the lowest share of any field in that survey. Marketing professionals sit at 42 hours per week ¹⁸, the highest.

Even the lowest profession writes a lot. A programmer's figure is higher than the Microsoft heavy-user email figure, though the activities are not the same thing. The marketing figure counts email, so it overlaps with the email rows and the two cannot be added together.

The studies disagree because they measure different things

Four differences account for most of the disagreement:

1. Activity. One row counts only research writing, one counts email checking, one counts everything typed. A reader who compares them directly is comparing different objects.
2. Population. Faculty, consumers, software users and heavy writers differ in job and in how they were selected. The Microsoft figure is the top quarter, and the WOBS figure is self-selected heavy writers.
3. Basis. Self-reports, software logs and conversions from shares of a day or week each carry their own bias. Most rows here are self-reported.
4. Season and date. The faculty figures are for teaching semesters only, and the Adobe survey predates current tools.

For email alone the logged figures are lower than the surveyed ones. For writing, the groups differ so much that method matters less than who was asked.

What the data cannot tell you

The data does not give a single average for the working population. No row measures a random sample of all workers, and several have no stated head count. The email rows do not separate reading from writing, so they say little about how much of that time is typing. The faculty rows exclude email and messaging, so they understate those people's total.

There is also a gap on messaging and chat. We found no figure that met our rules for time spent in chat tools, so this page cannot say how much of the week goes to them. Time diaries, the most careful way to measure this, are absent as well: every row here is a self-report, a software log or a conversion.

Several of the figures are conversions, and each states its conversion in its caveats. Treat those rows as approximate.

What this means if you choose how to write

Your own week is a better guide than any median here. A few steps follow from the data.

1. Find your group. If you are a heavy email user, the logged figure for the top quarter of Microsoft users, 8.8 hours per week ¹⁰, is a realistic marker. If you write documents for a living, the faculty and business-writer rows show how wide the range runs.
2. Measure before you decide. A week of logging with a tool shows how much of your time is typing, and how much is reading or waiting.
3. Match the tool to the task. The longest stretches are drafting documents and replying to email, and those are where a faster input method pays off most. Our guide to typing speed covers how fast people type, and the overview of dictation at work covers where speaking instead of typing helps and where it does not.
4. Do not treat email as writing time alone. Part of that time is reading and triage, which a faster way of entering text does not shorten.

For someone near the heavy end of these ranges, even a modest gain in drafting speed adds up over a year. For someone at the low end, the choice matters less, and the built-in tools on your computer are probably enough.