File sharing frequency

File sharing frequency is the average number of files a user shares with others over a defined period, usually a week or a month. If a team of 50 active users generates 400 share events in a week, file sharing frequency is 8 shares per active user per week. The metric tells you how often work actually moves between people, rather than how many files simply exist in the system.

This matters because sharing is where collaboration tools earn their value. A file that is created and never shared is a private document. A file that is shared turns the product into a place where teams coordinate. File sharing frequency is therefore a leading indicator of stickiness. It tends to move ahead of retention rate, because a user who shares regularly has woven the product into how the team works and is far less likely to leave.

The metric also separates genuine collaboration from passive storage. Two products can have identical daily active users and very different sharing frequency. The one with higher sharing frequency has stronger network effects, because every share pulls another person into the product and widens the surface area for feature adoption.

The core calculation divides total share events by the number of active users in the same window. The judgement sits in how you define a share event and which users belong in the denominator. Pin both down before you report a number, because changing either definition shifts the result without any change in behaviour.

1. 1

   Total share events

   Every action that grants another person or group access to a file: a direct share, a link created and sent, an invite to a shared folder. Decide whether re-sharing the same file to a new recipient counts as a fresh event. Most teams count it, because it represents new collaboration.

2. 2

   Active users

   The users who were active at least once in the period. Using total registered users instead of active users understates the metric, because dormant accounts dilute the average and make a healthy product look quiet.

3. 3

   Time window

   Weekly suits high-frequency collaboration tools, monthly suits slower document workflows. Keep the window fixed so the trend is comparable. Switching from monthly to weekly mid-quarter breaks the series.

4. 4

   Internal versus external shares

   Sharing inside the team behaves differently from sharing with outside guests. Segment the two so an uptick in external sharing is not mistaken for deeper internal adoption.

A worked example makes the definitions concrete. Suppose 200 users are active in a month and they generate 1,600 share events. File sharing frequency is 8 shares per active user per month. If 1,200 of those events come from just 20 power users, the average hides a thin middle. Reporting the median alongside the mean, or the share of users who shared at least once, prevents a handful of heavy sharers from masking weak adoption across the rest of the base.

A metric tree decomposes file sharing frequency into the drivers that actually move it, so a dip becomes a diagnosis rather than a mystery. The headline rate is the product of how many users share at all and how much each sharer shares, and each of those splits again into operational levers.

The first level separates participation from intensity. Participation is the share of active users who shared at least once. Intensity is the average number of shares among those who did. A falling headline number can come from either branch, and the fix is different. If participation is dropping, the problem is onboarding or discoverability of the share action. If intensity is dropping among active sharers, the problem is more likely friction in the sharing flow or a change in how work is structured.

KPI Tree models this by connecting each branch to the team that owns it. Product owns the visibility and friction of the share action. Growth owns the invite and onboarding flows that turn new users into sharers. When a node moves, the platform pushes the change to the accountable owner on that branch, so the right person sees it without scanning a dashboard, and the verified impact loop confirms whether their fix actually lifted the rate.

There is no universal benchmark, because sharing frequency depends heavily on product type. A real-time collaboration tool will show far higher numbers than a formal document archive. The useful comparison is against your own trend and against products in the same collaboration category. The ranges below describe weekly shares per active user for typical collaboration products.

Read these ranges in context. A surge in sharing frequency is only healthy if recipients act on what they receive. Pair the metric with downstream engagement, such as whether shared files are opened and edited, so a rise driven by spammy bulk sharing does not get mistaken for genuine collaboration.

Improving file sharing frequency means lifting either the share of users who share or the amount each sharer shares. The metric tree tells you which branch has the larger gap, so effort lands where it moves the number most. The cards below map to the main branches.

The most efficient lever is usually whichever branch sits furthest below its potential. If most active users never share, fixing participation will move the headline number more than squeezing extra shares from existing power users. If participation is already broad, the gain comes from removing friction so each sharer shares more often. KPI Tree keeps this honest by holding an owner accountable for each branch and checking, through the verified impact loop, whether a change to the share flow actually lifted the rate rather than just looking plausible.

1. 1

   Counting saves and edits as shares

   A save or auto-sync is not collaboration. Counting private actions as shares inflates the metric and hides whether work is actually moving between people. Count an event only when access is granted to someone else.

2. 2

   Using registered users in the denominator

   Dividing by total registered accounts instead of active users drags the average down with dormant logins. The result makes a healthy product look quiet and obscures the real trend among engaged users.

3. 3

   Reporting only the mean

   A small group of power sharers can carry the average while most users share nothing. Show the median or the share of users who shared at least once, so a thin middle does not hide behind a healthy mean.

4. 4

   Treating all volume as good

   Bulk or spammy sharing lifts the number without lifting real collaboration. Pair the metric with whether shared files get opened and acted on, so noise is not mistaken for engagement.

5. 5

   Mixing internal and external shares

   Internal team sharing and external guest sharing behave differently and signal different things. Lumping them together lets a rise in one mask a fall in the other.