
A subscription audit is a structured review of every recurring software charge the business pays, line by line, against the people who actually use it. The audit reads usage, seats and renewal dates from one register, then records a simple verdict for each line: keep it, cut it or step it down. This guide walks through the audit step by step, from gathering scattered records to setting the next audit date.
Key Points
- A subscription audit reviews every recurring software line against real usage.
- The register comes first, because lines outside it escape every later check.
- Seats, usage and renewal dates decide the verdict, not opinions.
- Each line ends with one verdict: keep, cut or downgrade.
- The audit closes by setting the next audit date and naming owners.
What A Subscription Audit Actually Reviews
A subscription audit is not a hunt for savings, although savings often follow from it. It is an evidence review. For every recurring line, the audit asks who uses the service, how much of the purchased capacity those people use, what the line costs across a full year, and when the next renewal gives the business a real chance to change the commitment.
Those four answers decide the verdict, which keeps the audit calm and repeatable rather than personal. Nobody defends a favourite tool, and nobody attacks one either. The record speaks, the owner responds, and the line earns its place or loses it. That tone matters, because an audit the team fears is an audit the team stops feeding with honest records.
Gather Every Record Into One Register
The audit starts with completeness, not judgement. Pull every recurring charge into one register: subscriptions, platform fees, add ons, domains, hosting plans and regular supplier services. Give each line its supplier, plan, billing cycle, annual value and renewal date. A line missing from the register escapes the audit entirely, and missing lines are usually the oldest and least questioned charges on the account.
Hunt for the awkward records deliberately. Check bank statements against the register, ask team leads what their people open every morning, and look for trials that converted quietly months ago. Applications holds this register with an owner and a renewal date on every line, which gives the audit a single list to work through instead of a pile of invoices.
Read Usage Before You Read Prices
Price is the worst first question in an audit, because a cheap tool nobody opens is still waste and an expensive tool the whole team depends on may be fair. Start with usage. For each line, read how many people hold assignments, how many of them used the service recently, and which work would stop if the service disappeared tomorrow morning.
Usage evidence lives in Consumption, where assignments sit beside purchased capacity. Where a service reports its own activity, compare the two views before believing either one. A tool with high assignments and low activity usually has an adoption problem or a training gap. A tool with low assignments and high activity may simply be licensed to the wrong group of people.
Compare Purchased Seats With Assigned Seats
Seat drift is the quietest finding in most audits. A team grows, buys seats for its new size, then changes shape again while the invoice keeps billing for the old one. Compare purchased seats with current assignments for every per seat line, and write the gap on the record where the owner has to answer for it at the review.
Treat every gap as a question rather than an automatic cut. Some spare seats are deliberate: a seasonal team, a client project starting next month, a handover in progress. The owner explains the reason once, in writing, and the audit accepts it or rejects it. Our article on reducing unnecessary business spend works through this seat check among its evidence led reviews of common waste.
Judge Each Line: Keep, Cut Or Downgrade
Every line leaves the audit with one of three verdicts. Keep means the usage evidence supports the current plan and price. Cut means the service ends at its renewal or cancellation deadline, with an owner responsible for the exit. Downgrade means the service stays but the tier, seat count or billing cycle steps down to match the real need the evidence showed.
Force the choice line by line, because a list of maybes is how audits die. Where evidence is genuinely missing, the verdict is a dated task: the owner produces usage proof before the next check, and the line returns to the top of that agenda. The spend control system guide shows how these verdicts feed the wider decision routine.
Check The Renewal Date Before The Verdict
A verdict without a date is only an opinion. Renewals decide when each decision can actually happen, so read the renewal date and the earlier cancellation deadline before recording any cut or downgrade. A cut agreed after the cancellation window closes is a wasted year, however strong the evidence behind it was.
Sort the audit by renewal date rather than by price. The lines renewing in the next ninety days get the deepest review, because their verdicts can still change real money. Lines renewing later get a lighter pass and a note to return. This ordering keeps the audit short enough to finish, which matters more than auditing everything with equal depth.
Question The Lines Nobody Owns
Every audit finds lines with no living owner: a tool bought by a colleague who left, a service inherited from a finished project, a charge nobody in the room recognises. Assign each orphan a temporary owner whose only job is to answer one question within a set period: does anyone still use this, and for what work?
An orphan that nobody claims after two checks becomes a cancellation candidate at its next window. Our article on business spending problems shows how orphaned costs grow into the commonest source of quiet waste in growing teams, precisely because no single person ever feels responsible for questioning them.
Write The Verdict On The Record Itself
Record each verdict on the line it belongs to, with the date, the reason in one sentence and the owner who will act on it. A verdict stored in meeting notes or a chat thread separates from the cost it describes, and the next audit starts by re-arguing decisions the business already made at full price.
The written verdict also protects good decisions. When a kept tool is questioned months later, the record shows the usage evidence that justified it and the date it will be read again. Over a year of audits, those notes become the memory of the whole software estate, and each new audit starts from evidence instead of from scratch.
Set The Next Audit Date Before You Close
An audit that ends without a next date becomes a one time cleanup, and cleanups decay. Prices rise, teams change shape and new tools arrive every month, so the findings of this audit start ageing immediately. Before closing, set the next audit date, name who runs it, and note which lines carry open tasks into that session.
Quarterly suits most small teams: often enough that seat drift and orphan lines stay small, rare enough that the work stays proportionate. High growth teams may audit their fastest changing categories monthly. The monthly budget review gives the smaller monthly rhythm a meeting format, while the audit itself stays a quarterly habit.
Prepare The Room Before The Session
An audit succeeds or fails before it starts. Book the session long enough to finish the register in one sitting, invite the owners of the largest lines rather than the whole company, and circulate the register a few days ahead with the renewal dates highlighted. Owners who arrive having seen their own lines bring answers instead of surprises, and the meeting spends its time deciding rather than discovering basic facts.
Preparation also means deciding the evidence standard in advance. Agree what counts as proof of use for each kind of tool: assignment lists, activity reports or the owner's written statement where a tool reports nothing. Arguments about evidence standards consume more audit time than any verdict, and settling them beforehand keeps the session's judgements consistent from the first line to the last.
Handle The Verdicts Nobody Wants To Hear
Some lines carry history. A founder's favourite tool, a platform a senior manager championed, a service woven into one team's identity. The audit treats these exactly like every other line, because an exemption granted for seniority teaches the whole room that evidence is negotiable and the next audit gets fiction instead of records.
Where a verdict genuinely needs a higher decision, escalate it as a written question with the evidence attached and a date for the answer, rather than letting it stall the remaining lines. The line stays flagged on the register meanwhile, and the escalation itself becomes part of its record. Handled this way, difficult verdicts strengthen the audit's authority instead of quietly dissolving it.
What The Second Audit Teaches That The First Cannot
The first audit maps the estate; the second one measures the system. Repeat findings are the real information. The same team producing orphan lines twice has a handover problem, not a memory problem. Seat drift returning in the same pools points at contracts that cannot flex mid term, which is a buying problem to fix at the next signature.
Compare the two audits line by line for the lines they share. Prices that rose without a review, verdicts from last time that never landed, owners who changed without the register noticing. That comparison converts the audit from an inspection into a management instrument, and it is the strongest argument for keeping the rhythm when a busy quarter tempts the team to skip it.
Final Verdict
A subscription audit works when it stays an evidence routine: one complete register, usage read before prices, seats compared with assignments, and a dated verdict of keep, cut or downgrade on every line. Run it quarterly and the software estate stays honest without drama or blame. Our cost control guide places this audit inside the wider routine that holds all business costs at their chosen level.
Ready to audit from connected records instead of scattered invoices? Spends Control keeps every subscription with its owner, seats and renewal date in one workspace, so the next audit starts from live figures.
Frequently Asked Questions
What Is A Subscription Audit?
A subscription audit is a line by line review of every recurring software charge a business pays. It compares each line with real usage, assigned seats and renewal dates, then records a verdict of keep, cut or downgrade on the record itself. The audit repeats on a set rhythm, usually quarterly, so findings stay current as prices, teams and tools change.
How Long Does A Subscription Audit Take?
For a small business with a tidy register, the first audit usually takes one focused working session, and later audits take less because the records already exist. Audits drag when the register is incomplete and the session becomes a search through statements and inboxes. Gathering records beforehand is what keeps the audit itself short.
Which Subscriptions Should An Audit Review First?
Review the lines renewing in the next ninety days first, because their verdicts can still change real money. After those, read the largest annual values and any line with no named owner. Lines renewing later in the year can take a lighter pass with a note to return before their window opens.
What Should Happen To A Subscription Nobody Uses?
Give it a temporary owner and one dated question: does anyone still use this, and for what work? If nobody claims it after two checks, schedule its cancellation at the next renewal or cancellation window, and record the verdict on the line so the decision survives staff changes and memory gaps.
How Is A Subscription Audit Different From A Budget Review?
A budget review compares totals with limits for a period. A subscription audit examines each recurring line on its own evidence: usage, seats, owner and renewal date. The audit produces verdicts about individual services, while the budget review produces decisions about category totals. Healthy teams run both, because each catches what the other misses.




