How to Audit Your Business Software Stack
Somebody on your team is paying for a tool you have never heard of. I would bet on it. It might be a design app a former intern signed up for on a company card, a second video meeting service that overlaps with the first, or a project board used by exactly one person who has since moved to a different role. The charge is small enough to ignore and regular enough to add up.
An audit of your software stack is how you find that tool, and the larger problems that tend to travel with it. This article is a method you can start on Monday and finish by Friday. It assumes you have no special software for the job. You need a spreadsheet, access to your company card statements, and a few conversations.
The goal is not a tidy inventory for its own sake. The goal is a short list of decisions. Keep this, cancel that, fix the connection between these two, and move the ownership of the other one to someone who will actually own it.
Before you start, set three rules
Audits die when they turn into witch hunts or into endless documents. Three ground rules prevent both.
- Nobody is in trouble. People who signed up for tools were trying to get work done. If you punish discovery, people will hide the next tool.
- Time box it. One week, a few hours a day. An audit that takes a quarter is a project, and projects get abandoned.
- Decide who decides. Pick one person who can approve a cancellation. Without that, every finding becomes a meeting.
Day one, find everything you pay for
You cannot audit what you cannot see, so the first day is about discovery. Use several sources, because no single one is complete.
- Card and bank statements. Go back twelve months and search for recurring charges. Annual charges hide here, since they appear only once a year.
- Accounting records. Look at vendor lists and expense categories for software, subscriptions, and hosting.
- Your identity provider or email admin console. If your team signs into tools using a Google or Microsoft account, the admin console often lists third-party apps that people have connected. That list reveals tools no invoice would.
- Expense reports. Staff reimbursed for software often never appear on the company card.
- A short question to the team. Ask everyone which tools they use in a normal week, and which ones they pay for personally. Remind them that nobody is in trouble.
- Browser extensions and desktop apps. Some tools live only as add-ons, and some are free tiers that hold company data without any invoice.
Free tools deserve special attention. A free account holding customer data is still part of your stack, and still a place where information can leak or get lost when someone leaves.
The collective name for tools adopted outside any approval is shadow IT. It is not a moral failure. It is a signal that the official tools did not meet some need. Treat each instance as feedback.
Day two, build the worksheet
Create one row per tool. These columns have worked well for small organizations, and I would keep them in this order.
| Column | What to record |
|---|---|
| Tool name | The product, plus the plan or tier |
| Purpose | One sentence on the job it does |
| Business owner | The person who answers for it, not the person who signed up |
| Users | Named seats, active seats, and the last login dates if available |
| Cost | Monthly or annual amount, and the billing date |
| Renewal date | When it renews, and the notice period to cancel |
| Data held | What kind of information lives in it, such as customer, financial, health, or none |
| Connections | Which other tools it exchanges data with, and how |
| Overlap | Other tools that do part of the same job |
| Access control | Single sign-on, multi-factor authentication, or a shared password |
| Export | Whether you can get your data out, and in what format |
| Score | The result of the scoring step below |
| Decision | Keep, merge, fix, or cancel |
Resist adding more columns. Every column is a question you must answer forty times.
The two columns people most often leave blank are business owner and renewal date. Fill them in even if it takes a phone call. A tool without an owner will not be maintained, and a tool with an unknown renewal date will renew.
Day three, ask what each tool actually does
For each row, do a short reality check. The question is not what the tool can do. The question is what it is doing in your business this month.
Open the usage reports, if the tool has any, and look at active users over the last 30 to 90 days. Compare paid seats with active seats. A seat that nobody has used in three months is a candidate for removal. Where there is no report, ask the owner and a regular user separately, since their answers sometimes differ.
Then trace the connections. For each tool, note where data enters it and where data leaves. Does somebody retype information from this tool into another? Does an automatic link move it? Is there an export someone runs by hand every Friday? Manual hand-offs show up here, and they are often where the real cost sits.
A hypothetical example, labeled as illustrative. A small consulting firm finds it pays for two e-signature services, one invoicing add-on, and a standalone calendar booking tool. Two of the four turn out to be used by a single person each. The booking tool duplicates a feature already included in the firm's main business suite. The audit is not finding waste in the large subscriptions. It is finding it in the overlap.
Day four, score each tool
Scoring turns opinions into something you can compare. Keep it simple enough that you can finish in an afternoon. Rate each tool from 1 to 5 on the following five questions, where 5 is good.
| Question | A score of 5 means | A score of 1 means |
|---|---|---|
| Value | Staff would complain loudly if it vanished | Nobody would notice |
| Usage | Most paid seats are active weekly | Most paid seats are idle |
| Fit | It does the job without workarounds | Staff use spreadsheets around it |
| Risk | Single sign-on, multi-factor authentication, tested export | Shared password, no export |
| Integration | Data flows automatically where it should | Data is retyped by hand |
Add the five scores for a total out of 25. You can weight risk more heavily if the tool holds sensitive data. A tool holding patient or financial records and scoring a 1 on risk deserves attention even if everyone loves it.
Be honest about the limits of this exercise. These are judgments, not measurements, and the numbers are only useful to rank tools against each other. Two people scoring the same tool may differ by a point, and that is fine. If you want to reduce that noise, have the owner score the tool and have one other person check it.
Day five, turn scores into decisions
The audit should produce a decision for every row. Four outcomes cover nearly everything.
- Keep. High value, good usage, acceptable risk. Record the owner and the renewal date and move on.
- Merge. Two tools do overlapping jobs. Pick the one that better fits your needs and plan the migration of the other, including how you will export data and when the cutover happens.
- Fix. The tool is valuable but something around it is broken. That might be a missing connection, an idle seat count, a shared password, or an absent export. Write the specific fix and assign it.
- Cancel. Low value, low usage, or high risk with no good reason to continue. Before canceling, confirm what data lives in it, export what you must keep, and check the cancellation notice period.
A rough decision rule. A tool scoring below about 12 out of 25 is a cancel or merge candidate. A tool scoring above 20 but with a risk score of 2 or below is a fix candidate. Anything else is probably a keep, with the lowest-scoring dimension noted for later. These thresholds are my rules of thumb, not a standard, so adjust them to your tolerance.
Worked example with illustrative numbers
Here is a made-up scenario so the arithmetic is concrete. A 15-person company finds 22 tools. Here are four rows, simplified.
| Tool | Paid seats | Active seats | Monthly cost | Score | Decision |
|---|---|---|---|---|---|
| Project board A | 15 | 4 | $150 | 11 | Merge into the main suite |
| Video meetings B | 15 | 14 | $225 | 22 | Keep |
| Design app C | 5 | 1 | $75 | 8 | Cancel after export |
| Customer database D | 8 | 8 | $320 | 21, risk 2 | Fix, move to single sign-on |
The savings from the first and third rows are $225 per month, or $2,700 per year, before counting the time people spent switching between overlapping tools. The fourth row produces no savings at all. It produces a safer system. That is the honest picture of most audits. Some findings save money, some reduce risk, and the second kind is often more valuable.
Seat true-ups and renewal timing
Two habits keep the audit from decaying after the first week.
A seat true-up is a periodic comparison between the seats you pay for and the people who actually use them. Do it before each renewal. Many vendors let you reduce seats only at renewal time or at the start of a billing cycle, so the timing is part of the savings.
A renewal calendar lists every renewal date and the notice period. Put the date to start deciding on your calendar, which is the renewal date minus the notice period minus two weeks. This one habit stops most accidental renewals. Several vendors renew annual plans automatically and ask for notice thirty or sixty days in advance, and you should read the terms for each tool rather than assuming.
Offboarding hygiene
Every time someone leaves, the stack is audited by accident. Make it deliberate. Keep a short checklist of tools by role, and when a person departs, remove their access from each one, transfer ownership of any documents or accounts they held, and reassign any paid seat. Tools that sit behind single sign-on make this a one-step job. Tools with their own passwords make it a scavenger hunt.
Watch especially for accounts where the departing person is the only administrator, or where the billing runs on their personal card. Fix these during the audit, not on their last day.
Negotiation basics, for what remains
When you keep a tool, the audit gives you bargaining strength in a conversation with the vendor, simply because you now know your real usage. A few plain suggestions.
- Ask for the plan that matches your active seat count, not your historical one.
- Ask whether annual billing lowers the rate, and weigh that against the flexibility you lose.
- Ask what the renewal price will be and when you will be told. Prices commonly step up at renewal.
- Ask whether you can pay a short-term rate while you evaluate an alternative.
If any of those terms depend on a contract you have signed, read it carefully or have counsel review it. I am describing general practice, not legal advice.
Things that go wrong in audits
A few patterns appear repeatedly.
- Canceling before exporting. Always confirm you can get your data out, and do it before the cancellation date, not after.
- Missing the free tools. They hold data too.
- Trusting one source. Card statements miss anything paid by invoice or by a personal card.
- Treating cheap tools as unimportant. A $12 per month tool can run the one process that invoices your largest client.
- Auditing once. The stack drifts. A short review every quarter is worth far more than a big one every three years.
A note on Puerto Rico and resilience
Add one column to the worksheet if your operation depends on staying online during storm season. Record what happens to each critical tool when your office loses power or internet. Cloud tools that work from a phone on mobile data are easier to live with than software that only runs on a desktop in one room. This is also a fair test of an export, since an export you cannot reach during an outage is not much of one.
The Monday morning version
If you have only an hour to start, do this.
- Download twelve months of card statements and highlight every recurring software charge.
- Open your email or identity admin console and list connected third-party apps.
- Start the worksheet with just three columns, tool, owner, and renewal date.
- Write the renewal date and notice period for the five most expensive tools on your calendar.
That hour alone often finds money and risk.
An invitation
If you would like help running the audit, or would just like someone to look over your worksheet when it is done, I am happy to. Ozcorp InfraOps includes this kind of review for small teams. The Base service starts at $500 per month, with vendor and license charges billed separately. Send me the worksheet, even half finished, and I will tell you which three rows I would look at first.
Discuss your infrastructure requirements
Review the current service offer or schedule a conversation about the scope you need.
Stay in touch
Follow Ozcorp
Leave your details for occasional notes about new articles and Ozcorp updates.