Why restoration labor is hard to track
A restoration technician rarely spends a whole day at one address. A day might hold an early emergency extraction, three monitoring visits of twenty minutes each, a demo day on a reconstruction job and an after-hours call from a property manager. Between every stop is drive time, and whether it is paid, billable or neither depends on your policy and the job.
Each visit has a different value: the extraction is billed on the invoice, the monitoring visits are the evidence that the dry-out progressed, the demo day is labor cost against a fixed estimate. A visit with no recorded time and place is hard to bill, hard to defend and hard to cost.
What goes wrong with paper and memory
Paper and memory fail in predictable ways.
- Rounding. A visit from 9:40 to 10:05 becomes half an hour. Over a week the drift goes both ways, and neither is right.
- Forgotten visits. A quick stop to swap a dehumidifier is the easiest thing to leave off a sheet, and it still mattered to the claim.
- Unbillable labor on claims. When the adjuster asks who was on site on day four and for how long, a handwritten word with no time beside it is easy to deny.
- Disputes. The technician says eight. The customer says nine. Nobody can prove either.
None of these are honesty problems. They are memory problems, and memory does not improve with training.
What GPS-verified clock-in is, and what it is not
The term gets used loosely, so it is worth being exact. When a technician taps clock in on a job, the app reads the phone's position at that moment and compares it to a geofence drawn around the job address. The same check happens at clock out. Inside the fence, the entry is marked verified. Outside, it is still recorded but flagged for the office to review.
Some systems add breadcrumbs, location samples taken at intervals while clocked in. Verification says whether the person was at the job when the shift started. Breadcrumbs say where the clocked-in time went, which matters for drive time and multi-site shifts. Here is what the method is not.
- Not a camera. It records a coordinate and an accuracy figure, not what the technician was doing.
- Not always-on tracking. The app reads location around clock events and, with breadcrumbs enabled, only while a shift is open. Off the clock, nothing.
- Not a substitute for approval. A verified entry can still carry the wrong task or a missed break.
- Not proof of work. It is proof of presence; readings, photos and notes are proof of work.
When there is no signal
Basements and burned-out structures are the normal case. A phone can often get a fix with no data connection, because the fix comes from satellites rather than the network, though indoors it is slower, less accurate and can fail. What the phone cannot do offline is send the entry anywhere. A field app built for this work records what it could get and queues the event with its timestamp and position. On the AIFSOP iOS app, clock events wait in an offline queue and sync when the connection returns, each applied exactly once, so a retry never duplicates a shift.
How to size a geofence
A geofence is a circle around the job's coordinates, and the common mistake is making the radius too small.
Two things push the radius outward. The first is parking: the technician clocks in from the cab, which may be at a loading dock a few hundred feet from the affected area. The second is GPS error. A phone's fix is usually within a few meters in open sky and can be off by tens of meters next to tall buildings, under tree cover or inside a structure. A technician 30 meters from the pin with a reported accuracy of 50 meters is not outside the fence in any useful sense.
A radius that is too tight produces false flags, and false flags train the office to ignore flags. Too loose is the smaller problem: a technician clocking in from the right street is still at the job.
A starting point
- Residential, single address. Start around 100 to 150 meters (roughly 300 to 500 feet) and tighten only if the first week shows no flags at all.
- Commercial and industrial sites. Walk the site once and cover the lot and the entrances. Several hundred meters is normal.
- A job with a wrong pin. Geocoding puts some addresses on the wrong parcel. If one job flags every visit, check the pin before blaming the crew.
What a time entry should capture
A clock-in and clock-out pair is the minimum. The entry is far more useful when it carries what an adjuster, a payroll clerk and a project manager each need.
| Field | Why it matters |
|---|---|
| Job | Ties the hours to a claim, an estimate and a customer. Without it, the entry is payroll only. |
| Task or phase | Extraction, monitoring, demo, equipment pickup. Compare hours by activity and bill the right line. |
| Clock-in and clock-out, each with location and accuracy | Timestamp, coordinate and the phone's accuracy estimate at each tap. A clock-out from elsewhere is a flag in itself. |
| Breaks | Meal and rest breaks recorded as events, not estimated afterwards. |
| Mileage or drive segment | Where the policy pays drive time or mileage, the segment between two jobs belongs on the record. |
| Notes | A short free-text field: waited for the adjuster, returned for a forgotten air mover. |
| Approval state | Submitted, approved, edited, rejected. Every edit keeps who changed what and when. |
Proof of presence for adjusters and customers
On a water loss, the daily monitoring visit is the part of the job most often questioned and least often documented well. The adjuster wants to know the technician was there, what the readings were and whether the equipment was still justified.
A verified time entry closes the first question with no extra work from the technician: on day three a named technician clocked in at 8:14 inside the geofence, logged readings in four rooms, took six photos and clocked out at 8:41. The readings and photos are already in the job record; the time entry is the frame around them. It also answers the customer who says nobody came on Tuesday, and holds up months later. The documentation checklist for insurance claims covers what else belongs in that file.
Payroll
Time captured at the moment still has to be approved: a supervisor reviews flagged entries, corrects a task, adds a missed break and signs off on the week. Nobody should be able to silently change a time; every edit keeps the original and the editor's name.
Approved hours then go to a payroll provider or an accounting system, so the practical requirement is an export those systems accept: employee, date, hours and pay type on each line. Overtime, shift differentials and on-call rules are the employer's to configure and differ by company and state; the time system records hours and should not guess at the rules. On the AIFSOP web app, the office approves entries and exports the approved period for the payroll run.
Federal wage and hour law requires covered employers to keep records of the hours their nonexempt employees work, and the U.S. Department of Labor publishes those recordkeeping requirements. Read them, and ask your payroll provider or counsel how they apply to your crews, drive time and on-call arrangements. A system that captures time at the moment and keeps edits visible is a foundation, not a legal opinion.
Job costing
Materials come with a receipt and equipment has a rental or depreciation figure, but labor is whatever the timesheet says, a week late.
Time entries tied to a job change that. Hours per task, multiplied by the loaded hourly rate, become a labor cost that updates as the job runs, and the drift from the estimate shows up while there is still time to act: a demo taking twice the estimated hours, a reconstruction phase quietly eating the margin. On AIFSOP that figure sits on the job's profitability view next to materials and equipment, with no wait for the payroll run.
Privacy and policy
Technicians accept location verification when they understand it and it does what it says. They work around it when it is vague or does more than it says. The policy matters more than the software.
The page should state:
- What is collected. Timestamps and coordinates at clock-in and clock-out, the accuracy figure and, if enabled, samples while clocked in.
- When. Only between clock-in and clock-out. Nothing before, nothing after, nothing on days off.
- Who sees it. The technician, their supervisor and the office staff who run payroll.
- What it is used for. Payroll, job costing and claim documentation. If it also feeds performance reviews, say so.
- How long it is kept, and how to raise a concern. A name, not a form.
Laws on employee location tracking vary by state and by country. Some require written notice, some require consent and some restrict what can be collected on personal devices. Review the policy with counsel before the first crew clocks in, and again when you expand into a new state.
Rollout
A change to how people get paid is not a software update.
- Start with one crew that trusts its supervisor. Four to six technicians is enough to find the problems.
- Explain the why in person. Lead with disputes and unbilled monitoring visits, not with control.
- Hand out the policy page and walk through it. Set geofences generously; early false flags cost credibility.
- Run paper and the app in parallel for one pay period. The differences are your training material.
- After the first week, review every flagged entry with the crew. Most will be parking, a wrong pin or a tight radius. Fix the radii, then the pins.
- Turn on breadcrumbs, if at all, only once the policy covers them. Then expand one crew at a time.
Paper, a time-clock app, or a job-aware field app
There are three ways to collect field time, and the difference is not only accuracy.
| Paper timesheet | Generic time-clock app | Job-aware field app | |
|---|---|---|---|
| Time captured | From memory, at day's end | At the moment, on the phone | At the moment, on the phone |
| Location | None | A point at clock-in, sometimes a fence | Checked against the job's own geofence |
| Tied to a job and task | If someone writes it in | A manual code or a separate list | Chosen from the job's own tasks |
| Beside readings and photos | No | No | Yes, same record, same timeline |
| Works offline | Yes | Varies | Queued on the phone, synced later |
| Labor cost per job | Manual | Needs a separate mapping | Computed from the entries |
A generic time-clock app solves payroll. It does not solve the claim file or the job cost, because it does not know what a job is. That is the gap a field app built for restoration fills.
Metrics to watch
- Hours per job type. Average labor hours for a Category 1 water loss versus a Category 3, or for a two-room mold job. These feed better estimates.
- Unbilled labor. Hours on a claim job that appear on no invoice line. Monitoring visits show up here first.
- Dispute count. Time or attendance disputes reaching the office per month. It should fall quickly and stay down.
- Flag rate. The share of entries outside the geofence. If it stays high after the first weeks, the radii or the pins are wrong.
- Estimate versus actual labor. Per job and per phase. The single most useful number for the next estimate.
Verified time is not about catching anyone. It is about the twenty-minute visit on day four that was real, that mattered to the claim and that nobody could prove happened. Record it once, at the moment, with the location, and the rest of the job file gets stronger with it.