According to McKinsey research, cited by document management specialists, the average professional spends 1.8 hours a day — 9.3 hours a week — just searching for and gathering information to do their own job. That's not time spent producing anything: it's time spent trying to find something that, in most cases, already exists somewhere inside the company.
Another compilation, from the market-research firm Cottrill Research, pulls together numbers from different sources that all point the same way: knowledge workers dedicate as much as around 30% of the workday to internal searches, and nearly 20% of work time gets lost to exactly this. The numbers vary from study to study, but the conclusion is always similar — too large a slice of the workday goes to "where is that thing", instead of to the work itself.
The odd part is that, in most cases, "that thing" was already written down. The company has a process manual, a reimbursement policy, a service procedure, a current price list. The problem is rarely the absence of the document. It's that the document is stored somewhere nobody remembers, under a name nobody searches for, in a version that may have already changed.
This reality isn't any different at Brazilian and Latin American companies — in practice, it tends to be even more informal. A large part of operational knowledge circulates through team chat groups, in shared spreadsheets with no version control, or simply in the memory of whoever has been in the role the longest. When that person goes on vacation, changes teams, or leaves the company, the question that used to have a quick answer stops having any answer at all.
The document exists, but nobody reaches it
Any mid-size company or larger accumulates a fair volume of written material: internal rules, operating manuals, HR policies, security procedures, service scripts. In theory, that body of work should resolve half of the everyday questions on its own.
In practice, it tends to be scattered across a shared folder, a document-management system, an old email, and a slide deck someone put together once for a meeting. Finding the right version requires knowing where to look, what file name to search for, and — most importantly — whether that copy is still current. Each of those requirements is one more reason to give up the search.
A common symptom shows up in any alignment meeting: two people cite different versions of the same policy, each one convinced their copy is the current one. Neither is wrong on purpose — each one just opened a different file, saved in a different folder, on a different date. The problem isn't a lack of discipline from whoever wrote the policy. It's the absence of a single place where "which version is current" has one reliable answer.
The result is exactly what the research measures: hours a week spent digging through folders and systems, often only to discover, in the end, that the answer was right there all along — it was just easier to ask someone than to search alone.
Asking a coworker is cheaper — it just doesn't scale
Faced with a hard-to-find document, the rational move for anyone is the same: ask whoever probably knows. It's faster, it's more comfortable, and it usually works. The problem is that this individual solution becomes a collective cost.
A Panopto report on workplace knowledge and productivity, cited by the publication HR Executive, estimated that an organization with 17,700 employees loses $47 million a year in productivity because of inefficient knowledge sharing — a 1,000-person company would lose roughly $2.7 million, and a 100,000-person one would reach $265 million. The same research points to $24.2 million a year lost specifically to inefficient onboarding of new hires, who spend an average of 28 hours a month, for 6.5 months, trying to learn their own job.
Asking a coworker has three effects that a "it's all documented" policy doesn't solve: whoever answers loses their own productive time; the answer might be outdated, because the person answering also didn't check the current version; and none of it gets recorded — the next person with the same question will interrupt another coworker, from scratch, months later. Writing yet another document doesn't fix this. The document keeps existing in the same hard-to-find way; it just gets company.
This dynamic gets worse when the question arrives through an informal channel — a direct message, a team chat thread. The answer is born and dies right there: nobody else in the company will ever find it again, even if the same question comes up in another team a few months down the road.
What has to be in place
What changes this picture isn't having more written content — it's changing the path between a person's question and the answer that already exists.
Knowledge with an owner, a current version, and scope by area. Every document needs an identified owner and a version clearly marked as "the current one" — not a history of similarly named files with uncertain dates. And access follows the right scope: the finance policy shows up for people who handle finance, not for the whole company.
An answer that cites the source, not just a generic summary. When someone asks about an internal rule, a useful answer isn't a loose sentence — it's a sentence accompanied by the document and the passage it came from, so the person asking can verify it and the same question doesn't need to be asked again unnecessarily.
Two-step review and approval for what's critical. Before a manual or a policy becomes the basis for any automated answer, someone reviews and approves that content — with whoever writes it kept separate from whoever approves it. This keeps an outdated draft, forgotten in a folder, from becoming an official answer by mistake.
Access according to each person's role. The same body of knowledge doesn't need to — and shouldn't — be equally visible to everyone. Each person sees what their own role authorizes.
This is how Skyller was designed: documents with an owner and a current version, answers that can cite the source used, and two-step review for anything critical before it becomes everyday knowledge.
The gain when the answer already comes with the source
The most immediate effect shows up exactly where Panopto measured the biggest loss: onboarding someone who just joined. A new hire who can ask and instantly receive the current policy with its source attached spends far less of the 28 monthly hours the research ties to the traditional learning curve — because they don't depend on finding, among their coworkers, someone with time to stop and explain.
There's also a consistency gain that rarely shows up on a spreadsheet but matters day to day: when the answer always comes from the same current document, two people in two different conversations get the same information, backed by the same source. Today, each coworker's answer can vary depending on what they remember, how long ago they last checked the policy, and how much they trust their own memory.
This effect compounds the time savings. Every question answered with a cited source is a question that doesn't need to be answered the same way again — over months, the volume of repeated interruptions drops, because the earlier answer stays accessible and stays correct, instead of getting lost in the conversation where it was given.
A starting checklist
Before writing yet another document, it's worth reviewing what already exists and how it's found:
- List the five documents most often asked about informally — asked of coworkers instead of searched for. These are the starting point with the fastest payoff.
- Confirm who owns each one and which version is current. If the answer isn't immediate, that's already the reason nobody trusts searching for it alone.
- Separate what's critical enough to require two-step review and approval — security policy, price list, regulatory procedure — from what can be made available directly.
- Measure how many repeated questions land on the same coworker every month. Each one is time that could have been an answer with a source, instead of an interruption.
- Check whether the answer changes depending on who answers. If two people on the same team give different answers to the same question, the problem isn't discipline — it's access to the right source.
