A productivity score is the share of tracked time spent in applications and websites that the company has classified as productive. If 6 hours of an 8-hour tracked day were in productive tools, the score is 75%. It measures where time went. It does not measure how good the work was, so it is useful for spotting patterns and misleading as a grade for a person.
Key takeaways
- The score is only as good as the rules behind it.
- The same site can be productive for one team and unproductive for another.
- There is no universal "good" score. Compare a team with its own past.
- Use it to start a conversation, never to end one.
How is a productivity score calculated?
Productivity score = productive time ÷ tracked time × 100
Example: 5h 45m productive out of 7h 40m tracked = 345 ÷ 460 minutes = 75%
Tools differ in the details. Some leave neutral time out of the calculation, and some weight it. What matters is that you know which method your tool uses and that it stays the same from month to month.
What does a day look like?
Idle time, 20 minutes in this example, is kept apart from tracked time and is not part of the score. What idle time means is covered in idle time at work.
Where do the three classes come from?
From rules. Each application and website is assigned to one of three classes.
Productive
Tools the role exists to use: an editor for a developer, a CRM for a salesperson.
Neutral
Tools that support work without being the work: email, chat, calendars, file managers.
Unproductive
Things unrelated to the job: entertainment, shopping, personal social media.
A rule can match the application name, the website domain or words in the window title. The last one matters for browsers, where one application holds both the project board and the video site.
Why do rules need to differ by team?
| Application or site | Engineering | Sales | Marketing | Support |
|---|---|---|---|---|
| Code editor | Productive | Neutral | Neutral | Neutral |
| CRM | Neutral | Productive | Productive | Productive |
| Social networks | Unproductive | Productive | Productive | Unproductive |
| Design tool | Neutral | Neutral | Productive | Neutral |
| Video platform | Unproductive | Unproductive | Productive, for content work | Unproductive |
One list for the whole company gives the marketing team a low score for doing its job. Score each team against rules written for its work.
What can a productivity score not tell you?
It can show
- How the tracked day was split between kinds of tools
- A change in a team's pattern over weeks
- Applications nobody has classified yet
- Where a process wastes time
It cannot show
- Whether the work was any good
- Work done away from the computer
- Thinking, reading and conversations
- Who deserves a raise
This is the central limit. The US Bureau of Labor Statistics defines labor productivity as the ratio of output per hour of work. A score built from application use says nothing about output. It is a measure of attention, and a rough one. Real output measures are covered in how to measure employee productivity.
What is a good productivity score?
There is no reliable universal benchmark, and any tool that quotes one is guessing. A support agent in one application all day will score higher than a manager who lives in meetings and email, and both may be doing the job well.
Build a baseline instead. Look at each team's score over a month, then watch for changes: a steady fall, a sudden jump or one person far from their own normal.
How do you set the rules up well?
Start from the defaults
Most tools classify common applications from day one.
Review what has no rule
New tools appear constantly. Classify them weekly at first.
Adjust by team
Change the class where a tool means something different for a role.
Ask the team
They know which tools are work. Their input also earns their trust.
Apply changes to history
Re-run new rules over past data so that trends stay comparable.
Read weekly, by team
Daily, individual scores are noise.
How should managers use the score?
Good uses of a productivity score
- Spotting a team whose pattern changed after a new process or tool
- Finding applications that take more time than anyone realized
- Checking that rules still match how the team works
- Opening a conversation with "what changed?"
What it should not be is a league table. Ranking people by score rewards whoever keeps the right window in front, and people learn that quickly. See mouse jigglers and fake activity.
How does productivity scoring work in BirdsEye?
BirdsEye classifies tracked time as productive, neutral or unproductive and turns the split into one score you can follow over time.
- Starter rules. Common development, office, communication and entertainment applications and sites are classified from day one.
- Custom rules. Match on application name, website domain or words in the window title. Add rules in bulk.
- Rule suggestions. BirdsEye surfaces the applications and domains seen most often that no rule covers yet.
- Uncategorized review. Separate lists of unclassified applications and unclassified websites.
- Re-run over history. Apply new or changed rules to past data, on the Team plan and above.
- Idle kept apart. Idle time is booked separately once keyboard and mouse stay untouched past your threshold.
Employees can see their own activity, and the rules screen in the agent shows what is tracked. See productivity tracking.
- VS Codeproductive
- Figmaproductive
- Outlookneutral
- Video streaming sitesunproductive
Frequently asked questions
What is a productivity score?
The percentage of tracked time spent in applications and websites that the company has classified as productive.
How is a productivity score calculated?
Productive time divided by tracked time, multiplied by 100. Tools differ on how they treat neutral time.
What is a good productivity score?
There is no universal figure. It depends on the role and the rules. Compare each team with its own history.
Is a productivity score a fair way to judge employees?
Not by itself. It shows which tools were in use, not the quality or value of the work. Use it to find patterns and combine it with output measures.