Time tracking at work: a practical guide

You run a shift-based operation. Some of your people are paid by the hour. You need to know what hours they actually worked, partly because they want to get paid for them, partly because in many jurisdictions you legally have to keep records, partly because if you’re invoicing a client for staff time, the numbers have to be defensible.

So time tracking isn’t really optional. The question is how to do it.

The market for time-tracking software has split into two camps. One sells productivity surveillance: screenshots every ten minutes, keystroke logging, application monitoring, AI-powered “workforce analytics” that looks suspiciously like activity scoring. The other sells lightweight clock-in/clock-out tools, mostly for shift workers and small businesses.

If you’re running an event staff operation, a cleaning crew, a hospitality venue, a volunteer programme, or an on-demand workforce, the second camp is the right one. The surveillance approach was designed for desk-based knowledge workers being managed from a distance. It doesn’t fit your operation, and it will burn your team’s trust faster than you can replace them.

This guide is about doing the lightweight version well: when you actually need to track time, what the minimum viable approach looks like, what the surveillance trap looks like and how to avoid it, what to put in writing, and what your team will worry about.

When time tracking is necessary, and when it isn’t

The honest cut.

You need to track time when:

Workers are paid by the hour, the day, or the shift. Payroll needs defensible numbers.
You’re invoicing a client for staff time. The hours you bill have to match the hours you can document.
Your jurisdiction’s labour law requires it. Most do, especially for hourly and shift work. The UK Working Time Regulations require records of weekly hours and night work. EU member states have similar requirements under the Working Time Directive. US federal law (FLSA) requires hourly-employee records. Specifics vary by state and country; the principle is universal.
You’re reporting to a grant funder or regulatory body that requires hours documentation.
You’re trying to understand your own operation’s labour cost per job, event, or shift.

You don’t need to track time when:

Workers are salaried with no client-billable hours.
The work output is what matters and the hours are incidental.
You’re using time tracking as a proxy for trust, productivity, or performance management. That’s not what it measures, and pretending it does makes everything worse.

If you’re tracking time for the second list, stop. You’re spending team trust on data that won’t answer the question you actually have.

What “doing it well” actually means

The minimum viable approach for shift-based work is short.

Clock-in and clock-out at the shift level. The worker records when they started and stopped. Tap on phone, tap on a shared device, supervisor confirmation, sign-in sheet: the mechanism matters less than the consistency.

Breaks, where legally required. Most jurisdictions require unpaid break tracking past a certain shift length. Capture it the same way as the shift bookends, no more invasively.

Overtime. The system needs to flag when worked hours cross into overtime so payroll can apply the correct rate.

Worker-visible hours before payroll. The worker should see their own logged hours and be able to flag corrections before the pay run, not after. The principle: errors should land in your inbox, not their bank account.

A correction process that doesn’t require a meeting. A tap, a note, a confirmation. Nothing more.

Export to payroll. The data goes where it needs to go without being re-typed.

That’s it. Everything beyond this list needs justification specific to your operation.

The surveillance trap

The line between time tracking and surveillance is sharp, and most of the well-known time-tracking products have crossed it. Specifically.

Continuous background GPS. Tracks where the worker is at all times the app is running, not just at clock-in. If your operation needs location at all, it needs it at the moment of clock-in, possibly at clock-out, never in between.

Screenshot capture. Some platforms take periodic screenshots of the worker’s screen. There is no shift-based operation where this is necessary. If a tool offers this, it isn’t built for your audience.

Keystroke logging and mouse-movement tracking. Same principle. Productivity surveillance for desk workers, irrelevant to shift work, corrosive to trust.

Application and website monitoring. Records which apps and sites the worker uses during tracked time. Productivity surveillance.

Productivity scoring. Numerical “activity ratings” based on the above. Often presented as helpful coaching data; in practice, a covert performance ranking.

“Time theft prevention” as primary framing. A small number of workers do clock fraudulent hours. A much larger number show up, do their work, and clock honest hours. Software that frames itself primarily around catching the first group will treat the second group like suspects, and the second group will notice.

Mandatory biometric or facial-recognition clock-in. Sometimes legitimate (high-security sites, regulated industries). Often unnecessary friction that telegraphs distrust. If the worker can clock in by tapping a button, biometric clock-in is solving a problem you don’t have.

The data you get from surveillance-style tracking is also less useful than people expect. Workers who know they’re being watched closely learn to perform the metrics rather than the work. The screenshots get taken when work is happening; the actual decisions and judgement that matter happen between screenshots and don’t show up in the data.

How this plays out for different teams

The minimum viable approach adapts depending on what you run.

Event staff. Clock-in and clock-out at the venue. Break tracking if shifts run long enough. Hours export for payroll on a per-event basis. Location at clock-in is fine and often useful (you do need to know everyone made it to the venue). Location continuously is overkill. (Event staffing apps cover the coordination side directly.)

Cleaning crews and dispatch. Per-job time, not continuous. Worker confirms start and end at each property. Location at job start is legitimate (you and the client both want to know the job was attended), location between jobs is not your business. (Cleaning crew dispatch handles this lane.)

Hospitality shift pools. Standard clock-in and clock-out at the venue, break compliance, overtime flagging. Tips reporting if applicable. Lean toward the simplest possible mechanism; venues that try to import office-style tracking onto a kitchen team tend to give up on tracking within a quarter.

Volunteer programmes. Hour logging is optional in most contexts and required in some (grant reporting, in-kind contribution accounting, awards). When required, capture it the same way as paid hours, but emphasise transparency: volunteers should see what’s logged about them and why. (Volunteer signup apps handle this distinctly from generic time-tracking products.)

Brand ambassador and activation work. Activation time plus photo evidence is usually enough. The photos serve as both work product and a light verification of presence, replacing the need for continuous location tracking.

On-demand staffing. Shift attendance, completion confirmation, hours export. The agency or platform owns the time record; the client venue doesn’t run its own parallel system. (On-demand staffing apps cover this category specifically.)

Care, healthcare, and other regulated work. Compliance requirements may mandate additional record-keeping (medication times, patient interactions, mandated breaks for safety). Layer those on top of basic time tracking, don’t replace it. Vendor must be compliant with relevant frameworks (HIPAA in the US, equivalent rules elsewhere).

What your team will actually worry about

The concerns workers raise about time tracking are mostly variants of concerns about work apps generally. Worth addressing each specifically rather than waving them off.

”Will you know where I am all day?”

No. Sensible time tracking captures location at the moment of clock-in or clock-out, not continuously. Anything more is surveillance disguised as scheduling. If your operation legitimately needs location at additional moments (start of each job for a cleaning crew, for example), capture it at those moments, log it transparently, and stop there.

”Will you see what apps I’m using on my phone?”

No. Apps that monitor website and application activity are productivity surveillance, not time tracking. A coordination app for shift work doesn’t need this and shouldn’t have it. If your tool offers screenshot or activity monitoring features, they should be off and stay off.

”What happens if the system records my hours wrong?”

Your hours should be visible to you before payroll and editable (with a flag, not a unilateral edit). A system where you only see your hours after payroll is one where errors land in your bank account instead of your inbox. The correction process should not require a meeting; a tap and a note should be enough.

”Can I see what’s being recorded about me?”

Yes. Workers should be able to see their own hours, any location records, and any related data on demand. This is GDPR-required in the EU and good practice everywhere else. If your vendor can’t show the worker their own data on request, that’s a sign the data model assumes worker-facing transparency is optional.

”Do I get paid for travel time or prep?”

Depends on what you and your employer have agreed. The policy should be written, and your tracker should capture what’s agreed: paid travel between sites, paid prep time, paid setup time, paid breaks, etc. The most common source of dispute on temp and shift work is uncertainty about whether the time between formal clock-in and actual paid work counts. Decide. Write it down. Track accordingly.

”What happens to my hours data when I leave?”

Pay records typically need to be retained for some years (six years in the UK, four years federally in the US, longer in some jurisdictions). After that, the data should be deleted or anonymised. The vendor should be able to tell you their retention policy. Workers should be told what’s kept and for how long.

What to put in writing

Even small operations should have a one-page statement covering time tracking. It doesn’t need to be a formal policy document. It needs to answer the questions a thoughtful worker would ask before clocking in.

A reasonable one-pager covers:

What’s tracked (shift start and end, breaks, overtime, location at clock-in if applicable)
What isn’t tracked (continuous location, application activity, anything else)
How hours flow to payroll
How workers can see and correct their own hours
What time is paid (and what isn’t): travel, prep, breaks, downtime
How long the data is retained, and what happens at the end
Who to contact with corrections or concerns

Send this with the onboarding pack, not buried in a contract. Transparency is the entire point.

When to skip time tracking entirely

There are cases where any time tracking is overkill.

A handful of regular volunteers on a stable rota. If hours aren’t required for grants and you’re not paying them, the trust cost of tracking exceeds the value of the data.

A salaried management team with no client billing. Time tracking for salaried workers without specific client or project requirements is performance theatre.

Knowledge work that’s measured by output. If the work is judged on what gets produced, not when, time tracking captures the wrong thing.

Short engagements where pay is agreed by the day or by the gig. A one-day flat-fee event doesn’t need hour tracking; it needs attendance confirmation.

For everyone else (event, cleaning, hospitality, on-demand, regulated work) the question isn’t whether to track but how. The answer is: lightly, transparently, and never as surveillance.

Where this fits

Time tracking sits inside the broader question of how you run a shift-based operation. The same principles that apply elsewhere (clear expectations, light-touch tools, transparency by default) apply here.

For the broader BYOD and app-rollout question, see work apps on personal phones. For managing hourly staff specifically, how to manage hourly employees covers the wider operational frame. For the scheduling side, self-scheduling and autonomy and accountability cover the workflow. For team comms, shift team communication sets out the signal-vs-noise discipline that applies just as much to time data as to messages.

Zelos isn’t a dedicated time-tracking product, by design. For shift-based operations, though, the data you need for time tracking is often the data Zelos already captures: who signed up for which shift, who claimed it, when they started, when they completed, location at task start if you configure it. No background tracking, no screenshots, no productivity scoring, no app monitoring. The shift history is the time record. Built in Estonia, GDPR-compliant by default. The free plan covers unlimited team members and 25 concurrent active tasks, with no per-person fees on any plan.