What a Tool Costs Besides Its Price

This post is a method, not a review. It names no product, quotes no price and carries none of this blog's three labels, because it is not making a claim about any particular tool.

It is about the part of a software decision that never appears on an invoice. The subscription is visible, recurring and easy to compare, which is exactly why it dominates the decision and why the decision is so often wrong. The costs that actually decide whether a tool was worth it are paid in time, attention and dependency, and nobody sends a bill for any of them. What follows is a way to put rough numbers on them before you commit.

The visible price is the one cost that cannot surprise you

A subscription is the best-behaved expense in an independent business. It is known in advance, identical every month, and trivially comparable between two products.

Which is precisely the problem. Comparison shopping goes where comparison is easy, so two tools get judged on the one number that is both the most legible and frequently the least significant. A tool that costs a few units a month more and saves an afternoon a quarter is not more expensive. It is cheaper, and the arithmetic is not close.

The six costs below are all harder to see than the price, which is the only reason they get ignored. None of them require precision. An estimate within fifty percent is enough to change most decisions, and an estimate of zero - which is what you are implicitly using when you ignore them - is the only one guaranteed to be wrong.

One: the hours before it works

Every tool has a setup cost paid once, in your most expensive hours, and it is routinely underestimated by a factor of several.

It is not only learning the interface. It is deciding how your work should be structured inside it, importing what already exists elsewhere, connecting it to the other things you use, discovering which of those connections does not do what the directory implied, and redoing the first structure once you understand the tool well enough to know the first attempt was wrong.

For an independent worker this cost has a particular shape: it is unbillable, and it arrives during a week you chose because it looked quiet. Setup time competes directly with paid work, so the honest unit to measure it in is not hours but client hours forgone.

• Estimate the hours, then double it. The second attempt at the structure is part of the job, not a sign you did it badly.
• Count the import separately. Moving existing records in is usually the longest single task and the easiest to forget.
• Price it in unbillable time, not in money - that is the currency it is actually paid in.

Two: the tax on every use

The second cost is small, invisible and never stops: the extra seconds a tool adds to work you were doing anyway.

A field that must be filled, a status that must be set, a record that must be created before the real task can start. Individually trivial. Multiplied by every occurrence for as long as you own the tool, it is frequently the largest line in the total, and it is the one no pricing page could possibly disclose.

The test is whether doing the work and recording the work are the same action or two actions. One action means the record stays true at no ongoing cost. Two actions means a tax, and - worse - a record that goes quietly wrong during busy weeks, which are exactly the weeks you most need it to be right.

This is also the cost that decides whether a tool survives. Tools are rarely abandoned because they lack features. They are abandoned because the per-use tax exceeded the benefit on a week when nobody had the time.

Three: the attention it takes while you are not using it

The third cost is paid in interruption, and it is the one most likely to be negative value outright.

A tool with default notification settings generates a stream of messages about things that do not need a human. The cost is not the seconds spent dismissing them; it is the decision required to determine that each one does not matter. That decision is small and it is made many times a day.

There is a predictable sequence here, worth naming because it ends badly. Notifications are too frequent, so you stop reading them, so you filter them, and now the one notification that did matter arrives in a folder you do not open. You have paid the attention cost and also lost the benefit.

The practical move is to turn almost everything off during the trial rather than after. A tool that is still useful with notifications silenced is a tool that earns its place. One that only works if it can interrupt you is renting your attention, and the rent is not on the invoice.

Four: the cost of the half-replaced system

The fourth cost appears when a tool does most of a job, and it is paid by keeping the old way alongside the new one.

A spreadsheet that still holds the real numbers. A folder that still holds the real documents. A habit of checking both because neither is complete. This is a common and expensive steady state, because you now carry the subscription, the per-use tax of the new tool, and the whole of the original problem.

It usually arrives through a reasonable decision - migrate the current work and leave the history where it is - and then nobody ever closes it out.

The defence is to decide in advance what will be switched off, and when. Not as an aspiration but as a date. If nothing can be switched off, the tool is an addition to your workload rather than a replacement, and it should be judged on that basis - which is a much higher bar than most tools clear. This is one of the questions the cancellation method is built to surface at renewal.

Five: the cost of leaving, paid at the worst moment

The fifth cost is the exit, and the thing to understand about it is when it falls due: never while things are calm.

You leave a tool because the price rose, a feature moved to a tier you do not have, the company was acquired, or the thing you depend on was discontinued. All four arrive on the vendor's schedule, not yours, and typically during a period when you have no spare capacity.

The size of this cost is set on the day you join, by one question: what can come out, and in what form? A tool that exports a flat list when what you stored was a structure has not really exported anything. The structure was the reason you bought it, and recreating it by hand is the actual bill.

So the honest time to test an export is during the trial, when it costs twenty minutes and you have no pressure. Run the export, open the file, and ask whether you could rebuild from it. That twenty minutes is the cheapest insurance available in this entire category.

• Export during the trial, not when you are leaving.
• Open the file. An export that exists and an export that is usable are different things.
• Ask what is missing - history, comments, attachments, relationships between records.
• Assume you will leave on the vendor's timing, during a week you cannot spare.

Six: the cost of having decided

The last cost is the strangest and the most often decisive: committing to a tool changes how you work, and some of that change is a cost rather than a benefit.

A tool encodes assumptions about how a job should be done. Adopt it and you adopt the assumptions, including the ones that do not fit you. Over months the work bends toward what the tool makes easy, and that drift is invisible from inside because each individual accommodation was sensible.

There is a financial version too. Money already spent argues for itself. A tool paid for through next year keeps being used past the point of usefulness, because stopping feels like waste - so the earlier commitment distorts the later judgement. That is a cost of the decision, not of the software.

The counterweight is a fixed review date set at purchase, when you are calm and have no sunk cost to defend. The reasoning behind reviewing at renewal rather than at the moment of frustration is in the cancellation method; the point here is only that the date has to be chosen before you have anything to justify.

Putting a number on it before you sign up

None of this requires a spreadsheet. It requires four estimates, written down once, before the trial ends - because afterwards you will be defending a decision rather than making one.

• Setup: hours to get it genuinely working, doubled, counted as unbillable time.
• Per-use tax: seconds added per occurrence, times occurrences per week, times fifty.
• What gets switched off, and the date it gets switched off. If the answer is nothing, say so plainly.
• The exit: run the export during the trial and note what did not come out.

Add the subscription last. It is usually the smallest of the five numbers, and seeing it in that position is the entire point of the exercise.

Two conclusions tend to follow, and both are useful. The first is that a more expensive tool with a lower per-use tax is frequently the cheaper purchase, and the second is that some tools which looked affordable are not worth adopting at any price, because the tax and the half-replaced system never go away.

Why this method has no prices in it: a figure here would be wrong for your rate, your volume and your month, and a wrong number is more persuasive than no number. The rules this blog works under, including why no figure appears without a source and a date, are set out in what I will and will not claim. The labelled reviews are where specific tools get assessed against their own documented plans - the time tracking review is the format.