How Many Hours a Month Go to Looking Things Up at Work? | What Surveys Say About Searching and Asking, and How to Cut It

"Where did I put that file?" "Who would know about this?" Each search takes only a few minutes, but repeated several times a day, it adds up to a real block of time. How much it adds up to at your own company is something you only know once you measure it.

This article collects survey numbers from Japan and abroad on the time spent looking for files and information, and the time spent asking colleagues or waiting for answers. It covers, in order, a monthly estimate, how to read the numbers, how to measure your own team, and steps to reduce the time.

Bottom line

  • A survey in Japan found office workers spend an average of 1.6 hours a day looking things up. At 8 hours a day and 20 days a month, that is about 32 hours a month.
  • Surveys abroad also find people spend around a quarter of their working time searching for information. One survey found 5.3 hours a week go just to waiting on colleagues for information and recreating knowledge that already exists.
  • Taken together, searching and asking come to roughly 20 to 30 hours a month. This is an estimate built on the assumptions below, not a measurement that fits every company.
  • Every number here is self-reported, and most of the surveys come from companies that sell search or knowledge-sharing tools. If you need a number to decide on, measuring your own team for two weeks is the reliable way.

The survey numbers

Survey Finding Respondents Published
OKWAVE Research Institute 1.6 hours a day on average looking things up 1,000 office workers in Japan April 2019
Panopto (with YouGov) 5.3 hours a week waiting on colleagues for information and recreating existing knowledge. 42% of workplace knowledge is held only by the individual 1,001 U.S. workers (at organizations with 200+ employees) July 2018
Atlassian, State of Teams 2025 25% of working time spent searching for information (23.5%, about 10 hours a week, in Australia) 12,000 knowledge workers in the U.S., U.K., Australia, and other countries, plus 200 executives at large companies 2025
Coveo, EX Relevance Report 2025 An average of 3 hours a day searching for information 4,000 people in the U.S. and U.K. at companies with 5,000+ employees April 2025

The OKWAVE Research Institute survey (Japan)

This survey asked 1,000 office workers in Japan. They spent an average of 1.6 hours a day looking things up, and the most common answer was "1 to under 2 hours" (69.2%). 62.9% said looking things up takes time away from their work.

The most common reason it takes long was "the information I need isn't in one place" (56.4%). The most common way to look things up was "internet search" (93.3%), followed by "a boss or colleague" (44.2%), so nearly half look things up by asking someone.

"Looking things up" in this survey includes internet searches. Note that it is not limited to searching for internal information.

The Panopto survey (U.S.)

This survey asked 1,001 people in the U.S. They spent 5.3 hours a week waiting on colleagues for information and recreating knowledge that already exists in the company. 42% of workplace knowledge is held only by the individual. 60% said getting the information they need for their job from colleagues is difficult, very difficult, or nearly impossible.

It is one of the few surveys that counts the time spent relying on people, rather than searching.

The Atlassian and Coveo surveys

Atlassian's survey found people spend 25% of their working time searching for information. For respondents in Australia, it was 23.5%, about 10 hours a week. Coveo's survey covered large companies with 5,000 or more employees and found an average of 3 hours a day searching for information. The bigger the company and the more systems people use, the longer searching tends to take.

What it adds up to in a month

Assuming 8 hours a day and 20 days a month (160 hours a month), here is each survey's number per month.

Survey Original number Per month Share of working time
OKWAVE Research Institute 1.6 hours a day About 32 hours 20%
Panopto (waiting and recreating) 5.3 hours a week About 23 hours 14%
Atlassian 25% of working time About 40 hours 25%
Coveo 3 hours a day About 60 hours 38%

The Japanese survey includes internet searches, and the two surveys abroad lean toward large companies. Both push the numbers above the time spent only on finding internal information and asking colleagues, so we treat roughly 20 to 30 hours a month (10 to 20% of working time) for searching internally and asking people combined as the working estimate.

Asking someone also uses the time of the person asked. At companies where someone answers the same questions again and again, the total is larger than this estimate.

How to read the numbers

  • They are self-reported. None of these surveys measured actual working time, and perceived time tends to run long.
  • Most surveys come from companies that sell search or knowledge-sharing tools. Numbers that make the problem look big are more likely to be published, so read them with some discount.
  • What counts as "looking things up" or "searching" differs by survey. Some include internet search, some include time waiting on colleagues.
  • Some surveys are old. The Japanese number is from 2019, and Panopto's from 2018. The time may have changed since chat and cloud shared folders became common.

How to measure your own team

A number you measured yourself beats any outside number. For about two weeks, ask a few people to fill in the table below.

Date What you looked for or asked Where you looked / who you asked Time taken Found it?
11/4 Last month's quote template Shared folder → asked a senior on the sales team 15 min Yes
11/5 A client's billing cutoff date Past emails → asked accounting 10 min Answered next day

Add up two weeks and divide by 10 to get the time per day. The "where you looked" column also shows where people looked and didn't find things. If the people who get asked keep the same table, you also learn how much time answering takes.

Steps to reduce searching and asking

  1. Decide where things go. Pick one place for each kind of document, and share the list. "Not in one place" was the top reason searching takes long.
  2. Write down what people ask often. Starting with the questions your log shows most, write one answer per page. The steps are in How to Spend Less Time Answering the Same Questions.
  3. Answer with a link to the document. The person asking learns where it lives and can find it themselves next time.
  4. Keep exchanges out of direct messages. An answer in a direct chat never reaches others with the same question. Ask in the team channel, and answer in the channel.
  5. Measure again a month later. Keep the same table for a week and check whether the time went down.

How to do it in Samurai App

In Samurai App, you can bring internal documents in as knowledge and have questions answered with sources attached. Since each answer comes with the documents it was based on, the person asking can check the original themselves.

Rules and procedures go in SIROE pages. You can choose who sees each page: just you, a group, or the whole tenant. If you build a bot that reads those pages and add it to an internal channel, it answers from the pages when someone calls it with "@" in the channel.

Learn more on the knowledge-sharing page.

Frequently asked questions

How much time looking things up counts as a lot?

Going by the surveys, more than 20% of working time (around 1.6 hours a day) is on the high side. But how much looking up a job needs varies with the work. Checking whether your own measured time went down after a month is more useful for decisions than comparing with outside numbers.

Shouldn't time spent asking people be left alone?

Time spent discussing background or judgment calls is not time to cut. What you want to cut is time spent on questions a document could answer (where things are, cutoff dates, procedures). Keeping a log shows which kind of question is more common.

How do I turn search time into labor cost?

Multiply the hours per month by the hourly labor cost and the headcount. For example, 20 hours a month, 3,000 yen an hour, and 50 people come to 3 million yen a month. Replace these assumptions with values you measured.

Sources

Confirmed against each survey's release and press coverage, and the Samurai App documentation, as of November 15, 2026.

Related featureWhy knowledge sharing fails, and a way that builds itself from workKnowledge sharing tools fail when writing things down becomes extra work. In Samurai App, conversations, files, and meeting minutes stay behind as you work, and internal information is shared and searched within each person's permissions.

More in this category

Back to Productivity articles

Everyday work comes together, connects, and becomes the company's strength.

Contact usOpen Samurai App

See the demoContact us