Four measurements, all of them actually measured
Time to accept, time to start, completion time and turnaround — each one a subtraction between two timestamps the product already stores. Nothing here is estimated, weighted or turned into a score.
No timer, no timesheet, no activity monitoring.
The problem
Most "productivity" numbers measure the wrong thing
A figure that cannot be traced back to a record is an opinion with a decimal point. The damage is not that it is inaccurate — it is that people start optimising for it.
- Activity counts reward whoever creates the most tasks, not whoever finishes the most work.
- A composite "score" hides its own formula, so nobody can argue with it and nobody can learn from it.
- Manually written summaries drift from what the database could actually produce — and once one figure is wrong, none of them are trusted.
- Averages that treat a missing timestamp as zero flatter whoever has the most incomplete records.
- Numbers arrive monthly, which is far too late to help anyone.
How TaskIt handles it
Every value traces to a real column
The analytics module computes employee, department and company figures strictly from stored task timestamps. If the data cannot produce a number, no number is produced.
- Time to accept — accepted at, minus created at.
- Time to start work — started at, minus created at.
- Completion time — completed at, minus started at. The hands-on duration.
- Turnaround time — completed at, minus created at. The total elapsed duration a requester experiences.
- A duration is averaged only when both ends exist and the result is non-negative, so a missing timestamp shrinks the sample instead of counting as zero.
What you get
Two figures that answer different questions
Completion time says how long the work took once it was started. Turnaround says how long the requester waited. A team with a good completion time and a bad turnaround has a queueing problem, not a capability problem.
Counts that cannot exceed the total
Attribution follows the current assignee, and each task has at most one, so per-person counts summed across the workspace can never exceed the workspace total — the invariant a hand-written report will break sooner or later.
Windows short enough to act on
Alongside the lifetime aggregate, every subject gets four time-zone-aware windows: today, the last seven local days, month to date and the previous month.
Rankings with a floor
The completion-rate ranking requires at least five assigned tasks, so someone who completed their only task cannot outrank someone who completed eighteen of twenty.
How it works
In the order a person using it meets them.
- 1
The team just works
Accepting, starting, handing over and completing all stamp their own timestamps. Nobody logs hours and nobody fills in a timesheet.
- 2
Figures are computed on demand
Company, department and employee aggregates, plus rankings, are calculated from those stored columns when they are requested — filtered by time zone, date range, department or person.
- 3
They reach you where you already read
The module is API-side today and surfaces through the scheduled WhatsApp reports, where the performance block can be embedded in the daily, weekly or monthly message.
- 4
You read them alongside the tasks
The People report group covers employee and department performance, productivity over 30 days and workload — the tabular view of the same underlying records.
Reading four numbers correctly
The same department, four figures, four different conclusions.
- 1
Time to accept: 4 hours
Work is sitting in a queue before anyone picks it up. That is a routing or notification problem, not a performance one.
- 2
Time to start: 4 hours 5 minutes
Almost identical to acceptance, which is expected — accepting starts the task in one step.
- 3
Completion time: 90 minutes
Once someone starts, the work itself is quick. Capability is not the constraint.
- 4
Turnaround: 6 hours
What the requester experienced. The gap between this and completion time is queue, and queue is the thing to fix.
The result
The fix is faster pickup — a department queue, or a notification the team actually sees — not pressure on the people already doing the work in ninety minutes.
What is counted, per subject
Alongside the four averages, each employee, department and the company get counts of assigned, accepted, started, completed, pending, in progress, overdue and blocked tasks, and a completion rate.
Rankings
Computed only from the values above, so a ranking cannot contain a factor the figures do not.
- Fastest and slowest responder, by time to accept
- Fastest turnaround
- Most completed, lifetime
- Most completed in the last seven days
- Highest completion rate, with a minimum of five assigned tasks
What this deliberately does not measure
There is no timer, no timesheet, no effort log, no screen monitoring and no activity tracking of any kind. It measures elapsed time between recorded events, which is not the same as effort — a task that sat over a weekend has a long turnaround and nobody was slow.
The blind spot, stated plainly
Attribution follows the current assignee. That is what keeps the counts internally consistent, and it means collaborators and prior owners in a handover chain earn no credit in these figures. Handover History and the collaboration reports are where that work is visible; use them alongside these numbers rather than instead of them.
Frequently asked questions
No. There is no timer, timesheet or effort log anywhere in the product. It measures elapsed time between events it already records — created, accepted, started and completed — which is a different thing from effort and should not be read as one.
The task is excluded from that average. A duration is only averaged when both ends exist and the result is non-negative, so a missing start time shrinks the sample rather than counting as a zero-length task.
Not in the per-employee timing figures, which follow the task’s current assignee. Collaboration is visible in the Tasks by Collaborator and Collaboration Statistics reports, and prior owners are visible in Handover History.
Through the People report group, and embedded in scheduled WhatsApp reports via the performance placeholder. The analytics module is API-side today and does not yet have a dedicated page of its own in the web app.
Where to go next
The capabilities either side of this one, and the problems it is bought to solve.
Get figures you can actually defend
Run a month of real work through a workspace and compare the four averages against what you assumed they were.