Decide whether a little-used team app earns its cost by tracing who needs its output, testing alternatives and checking what a downgrade would change.
Direct answer: Keep the app provisionally if that person's output saves the wider team more effort than the app costs to run, or enables work an acceptable alternative cannot do. Before cancellation, trace a recent output to its users, trial the replacement process and check the exact downgrade terms. Cancel when the output is no longer needed or the replacement completes the same work at lower total cost without losing necessary access, quality or continuity.
A single login can support a whole team. The person preparing a client report, cleaning a dataset or managing specialist drawings may be the only one who needs to open the application. Counting everyone else as a failed adoption target confuses the producer of an output with its beneficiaries.
That does not mean the specialist gets a permanent exemption from review. My recommendation is to demand evidence about useful work while resisting pressure to manufacture more logins. The parent guide on whether your tech stack is actually working sets the broader principle; this article addresses the awkward one-user case.
Applies to: small-team subscription reviews. Examples are editorial scenarios; any real cancellation depends on your specific account, billing arrangement and contract.
Apply the dependency value test
The dependency value test is an editorial method for following value beyond the person using the interface. Start with a recent output, identify who needs it, and compare the work required to produce and use it without the app. The unit being assessed is a completed job, not a seat.
Ask the current user to choose an ordinary piece of work rather than their most impressive achievement. Obtain permission before examining client information, and use a redacted example if the review group does not need the original data. Gather the actual invoice, renewal date, current features used and an export that you can inspect independently.
Do not disable the account to discover whether somebody complains. That experiment interrupts live work and may hide the consequences until a deadline. Use a trial copy or parallel process while the existing route remains available. The review should expose dependencies without making colleagues experience avoidable failures.
Trace the output to a decision
Name the output precisely: a reconciled production schedule, a checked print file or a report used to approve expenditure. Ask each recipient what they do with it and what would happen if it arrived later, in a different format or with less detail.
Follow at least one example through to its final use. A document being emailed is not evidence that somebody reads it. Equally, a manager who never opens the app may rely on a figure from it every week. Record the specific decision or action, the recipient and the frequency of use.
If nobody can identify a current use, check whether it supports a seasonal task before deciding it is redundant. A quarterly reporting tool may legitimately appear idle between reporting periods. Mark the next real task and its deadline. Do not accept an undefined future possibility as sufficient reason for indefinite renewal.
Reproduce one job with the alternative
Choose an alternative you already have where possible: another licensed application, a built-in export or a manual method. Give it the same input and acceptance conditions. Include the recipient's effort to read, reconcile or correct the result, not just the producer's time creating it.
Record what the alternative cannot reproduce. Separate inconvenient formatting from a material loss of meaning, accessibility, required editability or evidence. A different-looking output is acceptable if it supports the same job. An attractive replacement that makes recipients reconstruct the figures may be worse.
Have the recipient review the trial without telling them which result they should prefer. Ask whether they can complete the normal next task and identify where they need clarification. This is a small operational check, not a scientific blind trial, and it should not be reported as a product benchmark.
Price the dependency rather than the login
Use actual invoices for recurring cost and include administration time that disappears only if the app goes. An annual payment already committed may not be recoverable today; record the avoidable cost at renewal separately. Read the contract or ask the supplier where the billing consequence is uncertain.
Consider this explicitly illustrative example. A small production agency pays an assumed £48 a month for a specialist reporting app. One coordinator uses it. The app releases an assumed 90 minutes of that person's time monthly, and four colleagues each avoid 20 minutes of reformatting the resulting reports.
Total time released is 90 + (4 × 20) = 170 minutes a month. Maintaining the app and its export instructions takes an assumed 30 minutes, leaving 140 minutes, or 2 hours 20 minutes. At a deliberately assumed internal capacity value of £30 an hour, that time represents 140 ÷ 60 × £30 = £70 of capacity. Subtracting the £48 fee leaves a £22 illustrative capacity margin.
That margin is not profit or a cash saving. Salaries may remain unchanged, and the released time may not produce extra revenue. More importantly, the decision is sensitive to the assumptions: if each colleague saves only five minutes, net time becomes 90 + 20 − 30 = 80 minutes, valued at £40 on the same basis. The assumed £48 fee then exceeds that capacity value by £8.
Collect the recipient evidence. The difference between those cases is downstream usefulness, not the number of people logging in. Quality, deadlines and necessary capabilities may still justify keeping the app, but state those reasons separately instead of manipulating an hourly value until the purchase looks economical.
Choose between keeping, reducing and removing
| Option | When it fits | Evidence required | Main follow-up |
|---|---|---|---|
| Keep the current arrangement | A necessary feature or reliable output justifies its total cost | A recent accepted output and a credible alternative comparison | Document a backup operator and export route |
| Reduce the plan or seats | The needed work survives the lower entitlement | A checked feature and permission comparison for the actual account | Confirm the next charge and effective date |
| Remove the app | The job has ended or an alternative produces acceptable results | Completed replacement trial and usable retained records | Transfer ownership and verify cancellation |
A downgrade is a change in capability, not simply a smaller bill. For a documented example, Notion's downgrade guidance describes effects on features including history, uploads and security controls, depending on the destination plan. That is a reason to check what your actual workflow uses before reducing access, not a general recommendation for or against Notion.
Investigate continuity if only one person understands the workflow. Keeping useful software can be the correct decision while leaving its operation dependent on one colleague is not. Have a second authorised person locate the inputs, interpret the output and follow the documented process. Do not share credentials or buy unnecessary seats without first checking the supported access model.
Reach a decision before the next renewal
- Within two working days, collect the invoice and one ordinary output, then identify its recipients and next use. Allow about thirty minutes for the initial discussion.
- During the following week, reproduce that job through an existing alternative and record both producer and recipient effort. Keep the live route available.
- Before the cancellation deadline, choose keep, reduce or remove. Record the reason, billing effect and person responsible for the change.
- If removing the app, verify that retained files remain usable and that a successor can complete the next scheduled job. Stop the cancellation if essential access, rights or recovery remain unresolved; ask the supplier for the missing fact instead of guessing.
Related guides
Frequently asked questions
Does low adoption mean we bought the wrong software?
Sometimes, but low adoption alone cannot establish that. A specialist application may deliberately have one operator while several colleagues consume its output. Investigate whether the original buying decision expected broad participation or a specific specialist result. If the requirement was shared project coordination and everyone except one person works elsewhere, that is stronger evidence of a mismatch. If the requirement was preparing technically valid deliverables, assess those deliverables. The exception is a tool whose value genuinely depends on all relevant people contributing; then participation is part of task completion rather than a decorative usage metric.
What if the user says the app saves time but cannot quantify it?
Ask them to show a recent job and time a comparable one through the alternative. Avoid treating an unsupported estimate as a lie; people often know that a process is easier without having measured the difference. Include awkward cases and recipient corrections so the comparison remains credible. Where exact timing is impractical, record observable constraints such as whether a required output can be produced at all. A short provisional extension can be reasonable when it allows a real task to supply evidence. Give that extension a specific decision date rather than allowing uncertainty to become automatic renewal.
Should I move the only user to a cheaper personal account?
Only if the provider's terms, data ownership and access model support the actual business use. A lower headline price is not equivalent when billing, exports, recovery or organisational control change. Ask who would retain the files if that colleague left, and whether another authorised person could obtain access without their private credentials. Check your organisation's requirements before transferring client material. A personal account may suit a genuinely personal task, but moving team work there solely to reduce the subscription can replace a visible fee with a less visible continuity problem.
How should I assess an app used only once a quarter?
Evaluate a quarter or a complete reporting cycle, not an arbitrary month with no scheduled work. Record the effort and consequences attached to the actual task, including preparation and checking. Then compare a recurring subscription with any supported seasonal or lower-tier arrangement using verified billing terms. Do not assume cancelling and rejoining preserves every setting or retained record. If access is needed between reporting dates to answer questions, include that requirement. The recommendation changes when a usable archive and a different existing tool cover those occasional enquiries without keeping the subscription active.
What if the app makes work more accessible for one colleague?
Treat accessibility as a functional requirement, not an optional preference that disappears because other people do not need it. Ask the colleague which specific tasks become possible or less difficult and assess alternatives against those needs with their involvement. Do not demand disclosure of unnecessary medical information to justify ordinary configuration choices. A superficially cheaper replacement can transfer effort or exclude someone from work. Where a decision involves UK employment duties or a disputed accommodation, obtain appropriately qualified local advice; requirements vary by jurisdiction and circumstances. The operational review should establish usability, not make unsupported legal determinations.
Is it fair to count a manager's time when they never use the app?
Yes, if a demonstrated output changes the manager's work and you avoid double counting. For example, a report may remove the need to reconcile several separate documents. Compare the time required to reach the same decision with and without that report. Do not attribute every minute spent reading it as time saved, or assume that a convenient summary improves accuracy. Ask the recipient to identify the actual difference and check that the alternative could not provide it equally well. Value follows a defensible change in work, not proximity to a senior person's salary.
Sources and verification
- Notion: what happens when a plan is downgraded, checked on 9 September 2026 for plan-dependent feature and access changes. No vendor prices are quoted and no application was tested.
The dependency method, scenarios and numerical assumptions are editorial, not market benchmarks. The parent was read in the publication's project; its public category URL could not be retrieved. Internal links use the supplied article register.
This article is practical guidance. Apply it in proportion to your tools, evidence, risks, and responsibilities.
