How to choose an automatic time tracker
Almost every time tracker now says it tracks automatically. The phrase has been diluted to the point where it tells you nothing, because it is used to describe at least four different behaviours:
- A timer that starts when you open the app.
- A background recorder that logs which applications you used, which you later read and turn into entries by hand.
- A recorder that drafts entries for you, which you then confirm or correct.
- A tracker that creates the billable entry itself, against a project, as the work happens.
All four are “automatic”. Only the last one removes work from your week. The rest move it from the start of the session to the end of it, which is a real improvement in one specific way, because you cannot forget to do it retroactively, and no improvement at all in the amount of admin you do.
So the first question is not “is it automatic”.
1. Where does the recording stop and the entry begin?
Ask what you have to do between the software noticing you worked and there being a billable line with a project and a rate on it.
If the answer involves opening a timeline and dragging things onto projects, that is the product’s actual workflow and you will do it every week for as long as you use it. It might be fine. Some people like the weekly review and find the correction step reassuring. But price it honestly: fifteen minutes a week is thirteen hours a year.
If the answer is that entries are drafted and you approve them, ask how often the drafts are wrong, which you can only really find out during a trial. A drafting tool that is right most of the time is excellent. One that is right half the time is worse than nothing, because checking is slower than entering.
2. What happens to the time between the sessions?
Every tracker handles the obvious block correctly. The differences are all in the edges:
- Idle time. You walked away for twenty minutes. Does the tool notice, and does it ask you what to do about it, or silently keep counting, or silently discard it?
- Sleep and lid close. If the machine suspends for two hours, does the session end at the moment of suspension or at the moment you reopened it? Getting this wrong adds hours that never happened, and it is a surprisingly common bug.
- Context switching. Ten minutes moving between two clients is real work. Is it captured, discarded, or attributed to whichever client you happened to land on?
These sound like small details. They are the difference between a number you can put on an invoice and a number you have to audit before you can put it on an invoice.
3. Is the output an invoice, a payroll report, or a chart?
This is the question that sorts the category, and it is not usually on the pricing page.
Products built for freelancers and agencies end at an invoice: a document you send to a client that asks for money. Products built for employers end at a payroll report: hours verified so staff can be paid. Products built for self-improvement end at a chart: a picture of where your attention went, with no billing concept at all.
All three track time. All three are legitimate. Buying the wrong one means you will be exporting CSVs into something else forever, which is the specific pain that makes people give up on time tracking altogether.
Work out which of the three you are actually shopping for before you compare features, because the feature lists overlap enough to be misleading.
4. Who is the data for?
Related, but distinct. Some of these products are designed so that a third party can review your activity. Some are designed so that only you can.
You can usually tell from the pricing page without reading a word of marketing copy. Look for whether capture volume is metered by tier. If a product charges more for more screenshots per user per month, the thing being sold is capture, and the buyer is not the person being captured.
If you are a solo freelancer, a monitoring product will work, in the sense that it will record your hours. You will be carrying an oversight apparatus with nobody at the other end of it, and paying per seat for the privilege. Our comparison pages mark which products sit on which side of this, and each one includes a section on what that product does better than we do.
5. What does it cost when you are one person?
Three pricing patterns to watch for, all of which make the headline number misleading:
Seat minimums. A price advertised per seat with a two-seat floor is double that price for a solo operator. This is common and it is usually a footnote.
Individual plans that do not scale. Some products sell a single-user plan and a separate team product at a different rate. You cannot buy five of the individual plan, so multiplying the advertised price by five gives a number nobody can actually be charged.
Free tiers with headcount. Some free plans cover several users, which can make a paid comparison at small headcount irrelevant. Worth checking before you compare anything.
Compare like for like: same billing cadence, same headcount, both sides. Most published comparisons quote the competitor’s monthly rate against their own annual rate, which flatters the author by roughly twenty percent for free.
A shortcut
If you want to skip the research: work out which of the three outputs you need (invoice, payroll report, or chart), then check the seat minimum and the free tier at your actual headcount, then trial the two survivors for a week each and see how much correcting you do on Friday.
The correcting is the whole thing. Every one of these products can record a block of time. What separates them is how much of your week you spend fixing what they recorded.
We keep dated, sourced comparisons against the main products in this category, including the rows where the other one wins, which on most of those pages is several. If you would rather just try ours, Stunda is free for one person with no trial clock.