Time clock buying guide: face, fob, PIN, QR, geofence or WhatsApp?

1. The six ways to clock in

Here's what each method actually is, in one paragraph each.

Face recognition. A wall-mounted or handheld device with a camera. The worker looks at it and the device matches their face to their record. Nothing to carry and nothing to forget. Better devices also save a photo of each punch, which settles arguments later.

Key fob or RFID card. The worker taps a fob or card on a reader. It's fast, it works with gloves on, and many staff already carry a door card. The weak point is that a fob can be handed to someone else.

PIN. The worker types a short code on a keypad or screen. Cheapest to set up and easy to explain. Also the easiest to share: a code is just a number someone can tell a friend.

QR code. You print a code and post it at a site. Workers scan it with an app on their own phone. The code itself proves where they were, so it works where GPS is poor, such as inside large buildings.

Geofence. You draw a circle around each site on a map. The phone app only lets a worker clock in when its GPS position is inside that circle. Good for crews that move between sites. There's a full guide on how this works in Geofence time tracking.

WhatsApp or text message. The worker sends a word like "in" or "out" from their own phone number. No app to install and almost no training. Good for drivers, carers and contractors who never visit a base. On its own it proves the phone, not the place, so check what extra proof the system can ask for.

2. Pros and cons side by side

Two honest points that the brochures skip:

• No method is perfect on its own. Face clocks can struggle with harsh backlight. Fobs get passed around. Phones run out of battery. The right question is "what's our fallback when this fails?", not "which one never fails?"
• The strictest method isn't always the right one. A face clock at a two-person office is overkill. A PIN at a 200-person factory gate is an invitation to buddy punching. Match the strictness to the size of the risk. For more on that risk, see Buddy punching: how it happens and how to stop it.

3. What it really costs

The sticker price of a clock is usually the smallest part of the bill. Add up five things over three years.

1. Hardware. Devices, wall brackets, any spare units.
2. Connectivity. Wi-Fi at the site, or a data SIM in each device.
3. Consumables. Fobs and cards, and replacements for lost ones.
4. Software. Usually a monthly fee per employee or per device.
5. Admin time. The hours your managers spend fixing missed or wrong punches.

Worked example: three sites, 40 staff, over three years

These are example figures to show the method, not quotes. Plug in the prices you're actually given.

Option A: a wall clock at each site with face and fobs. Device US$ 400 each. Data SIM US$ 10 a month each. Fobs US$ 3 each, plus 10 replacements a year. Software US$ 4 per employee per month. One repair or spare US$ 300.

That's 8,550 ÷ 40 staff ÷ 36 months = about US$ 5.94 per person per month.

Option B: phones only (geofence or QR), plus one shared tablet for the five staff without a suitable phone. Software 5,760 plus one tablet at 250 = US$ 6,010, or about US$ 4.17 per person per month.

Option B is cheaper on paper. But it moves some cost onto your staff (their phone, their data, their battery), and it depends on everyone having a working phone every day. Option A costs more but gives you a fixed point nobody can argue with.

Don't forget admin time. If managers fix 20 missed punches a week at 3 minutes each, that's 60 minutes a week. At US$ 15 an hour that's US$ 15 a week, or US$ 780 a year. A method that staff find easy pays for itself here.

4. Sites with no Wi-Fi

Plenty of workplaces have weak or no internet: building sites, farms, beach bars, warehouses with metal roofs. Ask three questions.

• Can the device run on a data SIM? A clock with its own mobile data doesn't need your Wi-Fi at all.
• What happens when the signal drops? A good device stores punches and uploads them when the connection comes back. A poor one simply refuses to clock anyone in. Ask the vendor to show you, not tell you.
• Do phones need signal to punch? Phone-based methods (geofence, QR, WhatsApp) usually need data at the moment of the punch. GPS itself works without data, but sending the punch doesn't. If your site has no mobile signal either, a phone-only setup won't work there.

5. Gloves, dust, masks and the real world

Test the method in the conditions your staff actually work in, at the time of day they actually clock in.

• Gloves. Fobs and cards work through gloves. Face recognition does too. Touch screens and PIN pads often don't, unless the buttons are big and physical.
• Dust, heat, rain and salt air. Ask about the device's rating for dust and water and its working temperature range. A clock built for an office won't last on a building site.
• Masks, hard hats, glasses. Some face systems handle these well. Some don't. Try it with your own people wearing their own kit.
• Light. A camera facing a bright doorway at sunrise can struggle. Mounting position matters as much as the model.
• Queues. At a 6:00 a.m. shift change with 60 people, a method that takes 5 seconds each takes 5 minutes for the whole line. One that takes 15 seconds takes 15 minutes, which staff will notice and resent. Time it.

6. Shared devices or staff phones

A shared device on the wall gives you one fixed place, a consistent experience, and no question of whose phone or whose data. The downsides: one device per site, and a queue at busy times.

Staff using their own phones means no hardware and no queues, and it works for people who never come to a base. The downsides are real, though:

• Not everyone has a smartphone, or one with enough space and battery.
• Some staff object to installing a work app on a personal phone. That's a fair concern, so explain what the app does and doesn't collect.
• In some places, asking staff to use their own phone and data for work may come with an expectation of an allowance. Check local rules.

A common middle path: a shared device at the main site, and phones only for the people who genuinely work away from it.

7. Mixing methods per site

You don't have to pick one method for the whole business. A sensible mix might look like this:

• Head office: face recognition at the entrance, PIN as a fallback for anyone the camera won't read that day.
• Warehouse: fobs, because gloves stay on.
• Security posts: a QR code at each post.
• Delivery drivers: geofenced phone app at the depot, or a message from their registered number on the road.
• Office staff at desks: clocking in from the browser on their computer.

What matters is that every punch lands on the same timesheet, against the right site, whichever method made it. If each method feeds a different system, you've just created a new copying job. Ask to see a mixed week in the vendor's dashboard.

8. Getting hours into payroll

A time clock that doesn't connect to payroll is a data-entry project waiting to happen. Retyping hours from one screen into another takes time and creates errors. Typing 8.5 as 85 isn't rare.

Ask:

• Do punches turn into hours automatically, with breaks, overnight shifts and overtime rules applied? (See Overtime, explained for why those rules matter.)
• Can a manager fix a missed punch, and is the fix recorded with who changed it and when?
• Does it run payroll itself, or export to your payroll? If it exports, in what format, and does someone still have to tidy the file?
• Can staff see their own hours? People who can see their hours raise problems early, not on payday.
• How long are records kept, and can you get them out if you leave? See Time records and the law for why that matters.

If you're still on paper, Moving from paper to digital timesheets walks through the switch.

9. Evaluation checklist and vendor questions

Use this as a scorecard when you compare vendors. Ask for a trial on one real site for at least two pay periods.

• Each site has a primary method and a fallback for when it fails.
• The device (or phone setup) has been tested in real conditions: gloves, masks, light, dust, the morning queue.
• Punches are stored and uploaded later if the connection drops.
• The device can run on a data SIM where there's no Wi-Fi.
• There's a way to stop one person clocking in for another, proportionate to the risk.
• Every method feeds one timesheet, with the site recorded on each punch.
• Hours flow into payroll without retyping.
• Manager edits are logged with who, when and why.
• Staff can see their own hours.
• You've checked the local rules on face data and location, and written a short notice for staff.
• You have a three-year cost for your real headcount, including SIMs, fobs and support.

Questions to ask every vendor:

1. What happens to a punch when the internet is down? Show me.
2. What's the full three-year cost for our headcount and sites?
3. Who owns the punch data and photos, where are they stored, and how do we get them out?
4. How do you stop buddy punching with each method?
5. How do staff without a smartphone clock in?
6. What does a manager see when a punch is refused or missed?
7. How long does it take to set up a new site or a new starter?
8. Can we disable a site or a lost fob straight away, without losing its history?

10. Frequently asked questions