Productivity metrics for yourself: what to measure and what to ignore

Most pages about productivity metrics are written for managers who want a dashboard of their staff. This one is for the person being measured, and for small teams that want to know whether the week worked without installing anything on anyone's laptop. It covers the difference between activity, output and outcome metrics, a one-person adaptation of the SPACE framework, a metric-design table you can copy, what not to measure, how our Focus Planner logs focus minutes without sending them anywhere, and the European rules that limit what an employer may track. It connects to our guides on time blocking, which produces the plan you measure against, Getting Things Done, which produces the lists, and work-life balance, because the point of measuring is to stop when the number says so.

Activity, output and outcome

In the economic sense productivity is output per hour. The US Bureau of Labor Statistics reported that non-farm business productivity rose at a 4.9 percent annualised rate in the third quarter of 2025 because output grew 5.4 percent while hours grew 0.5 percent ^[1]. That definition is the right starting point for a person too: productivity is the ratio, not the hours.

Level	What it counts	Examples	Strength	Weakness
Input (activity)	Time and effort spent	Hours worked, focus minutes, messages sent, keystrokes, apps open	Easy to collect automatically	Measures cost, not value; the easiest level to game and the one surveillance tools sell
Output	Things finished	Drafts shipped, tickets closed, pages written, calls made	Visible, countable, tied to your work	Rewards small finished items over large important ones; ignores quality
Outcome	What the output changed	Client renewed, exam passed, bug reports down, revenue, a skill demonstrably learned	The only level that matters in the end	Slow, noisy, partly outside your control

A usable personal set has one metric from each level, with the outcome metric decided first. Activity alone tells you how busy you were; the economist's definition says busy is the denominator.

The phrase usually attributed to the economist Charles Goodhart, that a measure which becomes a target stops being a good measure, applies with full force to self-measurement. A focus-minutes target produces long sessions of light work; a tickets-closed target produces small tickets. The defence is the one the SPACE authors recommend: never use a single metric, and never use an activity metric on its own.

What the activity data show, and why they should not be targets

The best-known numbers about how knowledge workers spend their time come from vendors who sell tracking software, and they illustrate the problem with activity metrics as much as the problem with attention.

• RescueTime's 2018 analysis of close to 50,000 users reported that the average knowledge worker had about 72 minutes a day of uninterrupted focused work, and spent about 40 percent of productive time multitasking with email and chat ^[2]. Its 2017 analysis of 225 million hours reported about 40 checks of communication apps per working day ^[3]. These are vendor blog posts about a self-selected user base; a commenter noted that the chart shows a median where the text says average. We grade them C C.
• Microsoft's 2025 Work Trend Index, combining a survey of 31,000 workers with Microsoft 365 telemetry, reported interruptions every two minutes, about 275 a day including off-hours activity ^[4] A.
• Atlassian's 2025 survey of 12,000 knowledge workers reported that teams spend about a quarter of their time searching for information ^[5] A.

Each figure describes an input. None says what was produced. If your week was 72 focused minutes a day and the proposal shipped, the week worked; if it was 300 focused minutes a day and nothing shipped, it did not. The data are useful for diagnosing where attention goes, not as goals.

SPACE for one person

The SPACE framework was written for software teams by researchers at GitHub, Microsoft and the University of Victoria as a corrective to single-number productivity metrics ^[6]. It names five dimensions: satisfaction and wellbeing, performance, activity, communication and collaboration, and efficiency and flow. Its design rules are the useful part: measure at least three dimensions, include at least one perceptual measure (how it felt, not only what the system logged), and do not use activity metrics alone. It is a framework paper, not an experiment, so we grade it B and use it as structure rather than evidence B.

Adapted for an individual or a team of three to five:

SPACE dimension	Personal metric	Type	How collected	Gaming risk	Cadence
Satisfaction and wellbeing	End-of-day energy, 1 to 10	Perceptual	One number in a note or the weekly review	Low	Daily, reviewed weekly
Performance	Weekly big three completed, 0 to 3	Outcome or output	Count at the weekly review against what you wrote on Monday	Choosing easy outcomes; counter by writing the definition of done in advance	Weekly
Activity	Focus minutes logged	Input	Focus Planner timer, or a time tracker	High; use only as a denominator or diagnostic	Daily, read weekly
Communication and collaboration	Blocked time that survived meetings, as a share	Process	Compare the planned deep-work blocks with the calendar at the end of the week	Low	Weekly
Efficiency and flow	Self-reported interruptions per focus block	Process	A tally mark per block	Low	Per block, read weekly

Five numbers, two of them perceptual, one of them activity, and the activity number is explicitly not a target. The outcome lives in the performance row and is written before the week starts, because specific goals outperform vague ones and are harder to redefine after the fact ^[7].

A minimal weekly scorecard

If five rows are four too many, keep two: the big three completed, and average end-of-day energy. The first tells you whether the week produced what you intended; the second tells you what it cost. A week scoring 3 and 4 is a warning, and a week scoring 1 and 8 is a question about the plan rather than the person.

Add a third number, estimated against actual hours for the big three, after a month. People underestimate how long tasks take even when they know they did so last time ^[8], and the ratio between your estimates and your actuals is the most practically useful personal metric there is: it tells you how many outcomes a week can hold.

Goodhart and gaming yourself

Self-measurement has a built-in conflict: you set the target, you report the number and you feel the consequence. Three safeguards.

1. Fix the definition before the week. Write what "done" means for each outcome on Monday ^[7]. Friday is too late to be honest.
2. Pair every input with an output. Focus minutes alone drift upward while the work drifts toward whatever is pleasant to focus on. Minutes per finished outcome is harder to flatter.
3. Expect the benefit to be control, not speed. Time-management behaviours are moderately related to performance and more strongly to wellbeing across the literature ^[9], and the mechanism runs through perceived control of one's time ^[10] A. If a metric raises your sense of control, keep it; if it raises your anxiety and not your output, drop it. A metric you dread is already failing.

Tools: automatic versus manual

Automatic trackers record which applications and sites were active; manual trackers record what you say you were doing when you pressed start. The first is complete and unreflective; the second is incomplete and deliberate. For personal metrics the manual kind is usually enough, because the act of starting a timer is itself a decision about what matters. Prices are from our catalogue, checked 2026-11-09, in US dollars.

Tool	Model	Best for	Price	Note
RescueTime	Automatic, with focus sessions	Seeing where the hours actually went before designing a plan	Free plan; Solo $7 per month billed annually ($9 monthly)	Produces the activity data above; useful for a two-week diagnosis, risky as a permanent target
Toggl Track	Manual, with reports	Tracking focus blocks and client work by hand	Free plan; Starter $9 per user per month billed annually	The start button doubles as a commitment to one task
Clockify	Manual, team-oriented	Small teams that need shared timesheets without monitoring	Free plan; Basic $3.99 per seat per month billed annually	Hours per project, not screenshots
Our Focus Planner	Manual session timer	Logging focus minutes next to the day plan	Free, no account	Stores sessions in your browser only; see below

The full comparison, including each product's stance on surveillance features, is in our guide to the best time tracking software.

How the Focus Planner logs focus minutes

The planner's timer runs at 25, 50 or 90 minutes and only while the page is open. It makes no sound and sends no notifications. When a session completes, or when you press "stop and log" after at least one minute, the minutes are written to a log in your browser's local storage: the date, the minutes and a timestamp, nothing else. The weekly dashboard shows the last seven days as a bar per day with a seven-day total. Nothing is sent to us, there is no account, and a "clear session log" button deletes the record. The log is the activity row of the scorecard above and nothing more; the planner's own limitations note says that evidence for exact block lengths is weak and that the inputs are self-reported.

Employer monitoring: the EU limits

Self-measurement is legal because you consent to it and control it. Measurement by an employer is a different legal object, and in the European Union the perimeter is narrow.

• Consent does not work at work. The EU data-protection authorities' 2017 opinion on data processing at work holds that consent is highly unlikely to be a valid legal basis for monitoring employees, because they cannot refuse without consequence; monitoring must be strictly necessary for a legitimate purpose, proportionate and transparent ^[11] A.
• National law may add rules, and dignity is the test. GDPR Article 88 lets member states set specific rules for employee data, which must include measures to safeguard human dignity and address "monitoring systems at the work place" in particular; Article 22 gives people the right not to be subject to decisions based solely on automated processing with significant effects ^[12] B.
• An automated score can itself be the decision. In its 2023 SCHUFA judgment the Court of Justice held that automatically producing a probability score counts as automated decision-making when a third party relies on it strongly ^[13] A. A productivity score that feeds a performance decision sits close to that line.
• Emotion recognition at work is banned; performance-monitoring AI is high-risk. The AI Act prohibits AI systems that infer emotions of people in the workplace, outside medical or safety uses, a rule that has applied since 2 February 2025. AI used to monitor and evaluate workers' performance and behaviour is listed as high-risk in Annex III, with those obligations applying from 2 December 2027 after the 2026 Omnibus amendment ^[14][15] A.

For a small team the practical reading is simple: shared outcome metrics (what shipped, which goals were met) are ordinary management; shared activity metrics (minutes, keystrokes, app logs) are monitoring, and they need a legal basis, a necessity test and transparency before they need a dashboard. The metrics in this guide are designed to be owned by the person they describe.

Demand note: who searches for this

In the United States "productivity metrics" draws about 170 searches a month with a keyword difficulty of 5 and a cost per click of $34.53, down 46 percent year on year and declining about 18 percent a year over three years; "productivity tracker" draws about 480 a month, down 64 percent year on year (DataForSEO, retrieved 2026-11-09) ^[16]. Neither term is among the forty we tracked in UK, German, French, Spanish and Czech data, so this page has no localised version yet.

What to do next

Choose the two or five numbers, write Monday's definitions of done, and give the review a slot. The OS Builder returns a weekly rhythm with a fixed review block, a four-week checklist that includes a measurement step in week three, and a suggested app stack; bring the scorecard from this page to that review so the metrics are set before the tools are.

Frequently asked questions

What are good productivity metrics for an individual?

One from each level: an outcome (weekly big three completed against a definition of done written on Monday), a perceptual measure (end-of-day energy, 1 to 10) and, as a diagnostic only, an input such as focus minutes. The SPACE framework's rules apply: at least three dimensions, at least one perceptual, never activity alone ^[6].

Should I track my hours or my focus minutes?

As a denominator, yes; as a goal, no. Minutes measure cost, and a minutes target produces long sessions of light work. Pair them with an output (minutes per finished outcome) and use an estimate-versus-actual ratio, since people underestimate task duration even when they know they tend to ^[8].

What is the SPACE framework?

A framework from researchers at GitHub, Microsoft and the University of Victoria for measuring developer productivity across five dimensions: satisfaction and wellbeing, performance, activity, communication and collaboration, and efficiency and flow ^[6]. Its rules (several dimensions, at least one perceptual measure, no activity-only metrics) transfer to individuals and small teams; it is a framework paper, not an experiment.

Is it true that knowledge workers only focus for 72 minutes a day?

That figure comes from a 2018 RescueTime blog post analysing its own users, where the chart appears to show a median rather than the average the text claims ^[2]. It is vendor data from a self-selected group, which we grade C. It is a reasonable illustration of fragmented attention and a poor basis for a personal target.

Can my employer track my productivity in the EU?

Within narrow limits. EU data-protection authorities hold that consent is almost never a valid basis for monitoring employees and that monitoring must be strictly necessary, proportionate and transparent ^[11]. GDPR Article 88 requires safeguards for dignity around workplace monitoring systems ^[12], the AI Act bans emotion recognition at work and classes performance-monitoring AI as high-risk ^[14][15], and an automated score that drives a decision can itself count as automated decision-making ^[13].

Does the Focus Planner send my focus data anywhere?

No. The timer logs the date, minutes and a timestamp to your browser's local storage only, shows the last seven days as a bar chart, and has a button that clears the log. There is no account and nothing is transmitted.

Sources and evidence

1. Productivity and Costs, Third Quarter 2025, Preliminary — US Bureau of Labor Statistics (2026), dataset A www.bls.gov/news.release/archives/prod2_01082026.htm (retrieved 2026-11-09)
2. The true cost of email and IM: You only have 1 hour and 12 minutes of uninterrupted productive time a day — RescueTime blog (2018), vendor page C blog.rescuetime.com/communication-multitasking/ (retrieved 2026-11-09)
3. What we learned about productivity from analyzing 225 million hours of working time in 2017 — RescueTime blog (2018), vendor page C blog.rescuetime.com/225-million-hours-productivity/ (retrieved 2026-11-09)
4. 2025 Work Trend Index Annual Report: The year the Frontier Firm is born — Microsoft WorkLab (2025), report B www.microsoft.com/en-us/worklab/work-trend-index/2025-the-year-the-fro (retrieved 2026-11-09)
5. State of Teams 2025 — Atlassian (2025), report B www.atlassian.com/blog/state-of-teams-2025 (retrieved 2026-11-09)
6. The SPACE of Developer Productivity: There's more to it than you think — Forsgren, N.; Storey, M.-A.; Maddila, C.; Zimmermann, T.; Houck, B.; Butler, J. (2021), review B doi.org/10.1145/3453928 (retrieved 2026-11-09)
7. Building a practically useful theory of goal setting and task motivation: A 35-year odyssey — Locke, E. A.; Latham, G. P. (2002), review A doi.org/10.1037/0003-066X.57.9.705 (retrieved 2026-11-09)
8. Exploring the "planning fallacy": Why people underestimate their task completion times — Buehler, R.; Griffin, D.; Ross, M. (1994), RCT B doi.org/10.1037/0022-3514.67.3.366 (retrieved 2026-11-09)
9. Does time management work? A meta-analysis — Aeon, B.; Faber, A.; Panaccio, A. (2021), meta-analysis A doi.org/10.1371/journal.pone.0245066 (retrieved 2026-11-09)
10. Time management: Test of a process model — Macan, T. H. (1994), observational B doi.org/10.1037/0021-9010.79.3.381 (retrieved 2026-11-09)
11. Opinion 2/2017 on data processing at work (WP249) — Article 29 Data Protection Working Party (2017), report A ec.europa.eu/newsroom/article29/items/610169 (retrieved 2026-11-09)
12. Regulation (EU) 2016/679 (GDPR), Articles 22 and 88 — European Parliament and Council (2016), report B gdpr-info.eu/art-88-gdpr/ (retrieved 2026-11-09)
13. Judgment of 7 December 2023, OQ v Land Hessen (SCHUFA Holding), Case C-634/21 — Court of Justice of the European Union (2023), report A collab.dpa.gr/wp-content/uploads/2023/12/CJEU-C-634-21.en_.pdf (retrieved 2026-11-09)
14. Regulation (EU) 2024/1689 (AI Act), Article 5 and Annex III — European Parliament and Council (2024), report B artificialintelligenceact.eu/article/5/ (retrieved 2026-11-09)
15. AI Act | Shaping Europe's digital future (application timeline as amended) — European Commission (2026), report A digital-strategy.ec.europa.eu/en/policies/regulatory-framework-ai (retrieved 2026-11-09)
16. Keyword database: DataForSEO Labs keyword overview and historical search volume, US/UK/DE/FR/ES/CZ (data/research/keywords) — Productivity Signal research (2026), dataset A dataforseo.com/apis/dataforseo-labs-api (retrieved 2026-11-09)