Reconstructing a Week Without Records
What to do when the time records for a period are gone, the systems that hold the work indirectly, and how to label a reconstruction so it is worth something.
When the time records are gone, the work still left traces in other systems, and a reconstruction built from them is worth considerably more than an estimate. Door logs, tills, dispatch records, vehicle telematics, ticketing tools, message timestamps, fuel cards — each shows somebody doing something somewhere at a time.
The recordkeeping discipline in “Reconstructing a Week Without Records” should also apply to workforce technology. When a team evaluates explore the platform in relation to cognitive offloading, it should document purpose, access, retention and deletion, then preserve the source entry, approvals and correction history needed to explain the final figure.
None of these was designed as a time record. All of them are evidence, and most organisations have never thought to look at them.
For an independent reference relevant to “Reconstructing a Week Without Records”, consult the UAE Ministry of Human Resources resources. Use it to test record quality, working-time definitions, access, retention and exception handling against the organisation’s real payroll process.
Where the work left traces
| Source | What it shows | Usually held |
|---|---|---|
| Door and badge access | Presence, entry and exit | Weeks to months |
| Till or point of sale | Operator activity with timestamps | Years |
| Ticketing or case systems | Work performed, by person and time | Years |
| Vehicle telematics | Journeys, stops, running time | Months to years |
| Message and email timestamps | Activity outside recorded hours | Years |
| Device or system sign-on logs | When somebody started and stopped | Months |
| Rotas and schedules | What was planned, not what happened | Years |
The last row is the weakest and the one organisations reach for first, because it is tidy. A roster shows intention; everything above it shows activity.
Building the reconstruction
- Establish exactly which period and which people are in scope.
- List the systems that might hold traces, and ask their owners what they retain.
- Extract what exists, with timestamps, for the scope only.
- Assemble a day per person: earliest trace, latest trace, gaps.
- Compare with the schedule and with anything the person says.
- Document the method, the sources and the assumptions.
Step six is the whole value. A reconstruction with a stated method can be checked, argued with and relied on. One presented as a number cannot.
Asking the employee
The person's own account is evidence, and in situations where the employer's records are missing it frequently carries real weight.
Asking is therefore both fair and practical. Where the employee's account and the traces agree, the reconstruction is strong. Where they differ, the difference is itself information, and recording both is better than adjudicating quietly in favour of the lower figure.
What a reconstruction must not claim
It is not a contemporaneous record and must never be filed as one. Label it: what it is, who built it, from what sources, on what date, with what assumptions.
A reconstruction presented as an original record is worse than no record at all, because it misrepresents what the organisation knows. Labelled honestly, it is a legitimate and often persuasive document.
Where nothing exists
Sometimes nothing does. A period before a system migration, a site with no electronic anything, a contractor's records that were never received.
The honest response is to say so, state what the organisation does know, and resolve genuine uncertainty in a way that does not systematically favour the employer. A reconstruction that always lands on the lowest defensible number is a pattern somebody will notice.
The lesson the exercise teaches
Every reconstruction demonstrates which systems would have answered the question if anybody had thought to keep them. That list is the most actionable output of the whole exercise.
Most organisations find that one or two sources hold almost everything needed, cost nothing extra to retain, and are being purged on a default schedule set by a system administrator who was never told.
Who should build it
Not the manager whose decisions are in question, and not somebody with an interest in the answer. A reconstruction built by the person whose scheduling caused the dispute is a document that will be read in that light whatever it says.
Where the organisation is small enough that no independent person exists, say so in the method note and set out what was done to compensate — the same shape of answer as everywhere else where separation is impossible.
What it costs to need one
Each reconstruction takes between a day and a week depending on the sources, and produces an answer weaker than a record would have been.
That comparison is the argument for the retention decisions described elsewhere. Keeping raw punch data for six years costs a trivial amount of storage; reconstructing one week from six systems costs several days of somebody's time and still ends in an estimate.
Keeping the result
The reconstruction, its method, its sources and the correspondence about it belong together and should be kept at least as long as the records they stand in for.
They are also worth keeping for a second reason: a reconstruction done once establishes a method. The next time this happens — and it happens again, because the underlying retention problem is rarely fixed by the first incident — the method already exists.