Switching Costs Nobody Quotes
Any comparison between what you have and what you might buy shows two subscription prices. The difference between those two numbers is rarely the answer, because switching costs something and the something is not on either page. For a related reference point, Monitask has a guide to time tracker manipulation methods, which is useful when comparing whether a dedicated tool is warranted.
No affiliate links on this site, and nothing here is ranked.
The six costs
One. Migration. Getting data out of one thing and into another. Where the export is good this is an afternoon; where it is not, it is a fortnight or a decision to lose history.
Two. Configuration. Setting up the new thing to match how you work. Frequently longer than the migration and always done by whoever is most stretched.
Three. Learning. Everybody, not just the person who chose it. An hour each is optimistic and it is not the real cost — the real cost is the fortnight of doing everything slightly slower.
Four. Running both. Which is normal, necessary and should be time-boxed — and during it you are paying twice and maintaining two records.
Five. Rebuilding the things around it. Integrations, saved reports, bookmarked views, the spreadsheet somebody built that pulls from it. These are invisible until they break.
Six. The mistakes. Things missed during the change, entered twice, or done in the old system by somebody who forgot.
Estimating it in an hour
Write down each of the six and put a number of hours against it.
Be pessimistic on configuration and learning, which are the two most consistently underestimated.
Multiply by what an hour of your time is worth — not the billing rate, but what you would otherwise be doing.
Add the double-running months.
That total, divided by the monthly saving, is the payback period. If it exceeds a year, the switch is being made for reasons other than money, which is fine as long as it is known.
When the switch is worth it anyway
When the current tool is genuinely failing at something specific. Not untidy — failing, with a named consequence.
When the price is rising steeply, which happens at renewal and is a legitimate trigger.
When somebody is leaving who was the only person who could work it. A tool that needs one particular person is a risk, and the moment they announce a departure is when it becomes visible.
Or when the tool is disappearing — acquired, sunset, or a company in obvious trouble. This one has a deadline attached and the deadline is not yours.
The switch not to make
To a marginally better tool.
Where the improvement is real but small, the switching cost eats it entirely, and frequently for several years.
The bar for switching is considerably higher than the bar for choosing, and confusing the two is why firms migrate repeatedly and never settle — each move justified on its own terms, none of them ever repaid.
Reducing it in advance
Two decisions made at purchase time.
Ask the export questions before buying, which converts the largest of the six costs from unknown to known.
And keep the configuration simple. Elaborate setup is elaborate switching cost, and most of it was built to satisfy a preference rather than a requirement.
A tool used plainly is a tool you can leave. That is worth something on the day it matters and costs nothing until then.
The cost that is not a cost
Losing history you were never going to look at.
Firms delay switches for years over archives nobody has opened — three years of closed tasks, old chat messages, superseded documents.
Ask when you last needed something from more than a year ago. For most categories the honest answer is never, and where it is not never, the answer is usually a document rather than the whole record.
Export it, keep the file, and stop treating it as a constraint. An archive somewhere safe is not the same as an archive inside a live subscription, and confusing the two keeps tools alive for no reason.
Doing the switch well
Four things that reduce the pain more than the choice of tool does.
A date for switching the old thing off, decided before starting.
One person responsible, because a migration owned by everybody stalls.
Migrate the active, not the archive — current customers, open work, live records — and leave the rest in an exported file.
And tell people what is changing and when, once, in writing. Most of the friction in a switch is surprise rather than difficulty.
What vendors do to help
Some genuinely do, and it is worth asking.
Migration assistance, import tooling, a discount for the double-running period — these exist and are commonly offered when requested rather than advertised.
Ask directly: "What do you do to help people move from what we have now?" A vendor with a real answer has moved people before, which is itself useful information.
The switch you should have made and did not
The other error, and it is real.
A tool that has been failing for two years, kept because switching seemed expensive, has cost more than any migration would have — in workarounds, in the second tool somebody quietly started using, in the work that goes wrong.
The signal is a workaround that has become permanent. Where people have built a routine around a defect, the defect has been priced in and nobody is counting it any more.
Ask once a year what everybody's least favourite tool is. The answer is usually immediate and unanimous, and it is the one worth doing the arithmetic on. For broader context, see Tom's Guide.
The short version
- Any comparison shows two subscription prices and omits six real costs
- The six: migration, configuration, learning, running both, rebuilding the things around it, and the mistakes made during the change
- Configuration and learning are the two most consistently underestimated
- Total hours divided by the monthly saving is the payback; beyond a year the switch is for other reasons
- Switch when the tool is failing specifically, when the price rises steeply, when the only person who can work it leaves, or when it is being shut down
- The bar for switching is much higher than the bar for choosing, and a marginal improvement never repays the move