Skip to content
Timesheet Systems

All notes

Everything here, in the order the work runs. If a product is already chosen, start at implementation. If a project has stalled, start at common failures.

All 50 notes, in order

The sections run in the order the work does. Skipping an early one makes a later one harder, which is why they are numbered rather than listed.

Requirements

10 notes

Four different products share this category name, and the rule engine decides which are viable at all. Plus the requirements that start with "not", which are the ones that get skipped.

Evaluation

4 notes

Convert every claim into a question with a checkable answer, then make the vendor demonstrate your own awkward case rather than their example.

The fields that decide disputes are the ones nobody looks at until there is one. Plus the three time cases that break products and are entirely predictable.

Integrations

6 notes

Every integration here is a join between systems, and the join fails on whatever identifier someone chose casually. The reconciliation is what turns a silent failure into a Tuesday email.

Installation is an afternoon. Rule discovery, reconciliation and organisational agreement are the project, and plans that allocate time the other way round run late in the same place every time.

Running it

9 notes

The project ends at go-live; the system generates work every period indefinitely. That work is routinely unassigned, and every failure below follows from it.

Two vendors quoting the same per-seat figure can differ threefold over a term. And the lock-in is not in the data, which exports, but in everything built around it.

Reference

4 notes

What no product resolves, the asks that should be declined, and a description of the end state to measure a proposal against.