Why nobody wants to log hours
There are three reasons, and all three can be fixed:
- It feels like surveillance. If hours are only used to ask “why did this take so long?”, people pad them or forget them.
- It takes time. A fifteen-field timesheet at the end of the day is one more chore, and a boring one.
- Nobody sees the result. If hours disappear into a spreadsheet nobody discusses, logging them feels pointless.
Explain what the hours are for, and what they are not for
Project hours are good for three things: checking whether the estimate was realistic, catching variance in time and quoting the next project better. Say exactly that, before you ask for the first hour.
Also say what they will not be used for. If you compare people by hours logged, the data goes bad within a month. Whoever takes longer may be doing the hard part.
Starting Monday, everyone logs their hours.
We want to know whether we quote our rollouts correctly. For two months, log the time for each task. At the wrap-up, we will show you what we learned.
Log time where the work happens, not on Friday
One hour logged on the spot is worth more than eight rebuilt from memory. Memory rounds off, forgets interruptions and dumps everything on the most visible task.
- Against the task, not just the project. “12 hours on the project” does not say where the time went. “12 hours on data migration” does.
- From wherever the work is. The computer at the office, and the phone on the job site, on the plant floor or at the customer’s location. If people have to go back to the office to log time, they won’t.
- When each task ends. Even better, with a timer started at the beginning, if the work allows it.
- With as few fields as possible. Task, time and, if needed, a note. The task already knows the customer, the project and the phase.
How much detail to ask for
| Detail | When it makes sense | Risk |
|---|---|---|
| Project only | Short, small fixed-price projects | You cannot tell which part ran over |
| Project and phase or line item | Most cases | Low, if phases are few and clear |
| Each task or document | If you quote by task or bill by the hour | High if there are too many tasks: people log to the first one they find |
A rule of thumb: if nobody will ever look at a line item on its own, do not ask people to log time to it on its own.
From hours to the real project cost
Hours are not cost. To get to cost, multiply each hour by the cost of the person who worked it, then add everything that is not hours:
Actual cost = each person’s hours × their hourly cost
+ expenses: travel, per diem and subcontractors
+ materials
Hourly cost has to be the fully loaded one, not salary divided by hours. It is explained in what a field technician hour really costs, and the math works for any role.
Example
An engineering firm quotes a building systems project at 120 hours, at a cost of $55 an hour, plus $1,000 in travel. At closeout:
| Item | Budget | Actual |
|---|---|---|
| Engineering hours | 120 h × $55 = $6,600 | 150 h × $55 = $8,250 |
| Travel | $1,000 | $1,400 |
| Total | $7,600 | $9,650 |
The overrun is $2,050, or 27%. With hours logged by line item, you can see that 25 of the 30 extra hours went into drawing reviews with the customer. The next quote knows what to adjust. The figures are an example.
And if the project is billed by the hour, time that is not logged is not billed: the hidden cost of paper work orders runs that math.
Give the data back to the people who log it
- At project wrap-up, compare budget with actual, phase by phase, with the team in the room.
- Use actual hours to quote the next project, and say so.
- Watch the load: if someone logs time on four projects at once, that is a conversation about workload, not performance.
Logging time is not the same as recording hours worked. Hours worked show when each person worked; logged time shows what they worked on. If you need both, nobody should enter the same thing twice.
Start with one project, for two months, with just enough detail. If the data helps you decide at wrap-up, the team will keep logging without anyone asking.
Last updated: