SaaS spend management is the practice of finding every software subscription your company pays for, assigning an owner to each one, and making a deliberate renewal decision before the notice deadline passes.
The decision in front of you is simple: build one inventory of software spend, and attach a named owner and a renewal rule to every line. Most of the savings come from that discipline, not from a new tool.
The evidence points the same way. Spendesk found that 72% of companies using AI tools pay for two or more competing products. Zylo found companies use only 49% of the licenses they provision.
This guide gives you the definition, the calculations, how to read the results, what to do next, and a checklist you can copy from this page.
Key takeaways
- Software spend is a cash and runway question first, and an IT question second.
- Duplication and unused licenses are common, so look at categories, not only single tools.
- Fixed subscriptions need renewal-cycle discipline. Usage-based AI spend needs monitoring while the spend is happening.
- Every tool needs one data owner and one decision owner.
- The inventory checklist below is ungated. Copy it and use it today.
Why software spend is a finance decision, not a procurement chore
Software subscriptions sit inside gross burn next to payroll and cloud infrastructure. Fiscallion's net burn rate methodology lists them explicitly among the cash outflows to capture.
Scale makes this harder to see. In the Spendesk European Spend Report 2026, companies in the sample spent close to EUR 250,000 a year on software on average. The data covers June 2025 to June 2026 and a mostly small-company European panel, so treat it as a reference point, not your budget.
Cost placement matters too. In one Fiscallion gross-margin review, data pipeline, cloud infrastructure and vendor API costs sat in an engineering budget line instead of cost of goods delivered. That hid a gross-margin gap against the peer benchmark the founders had been using.
In Fiscallion's experience, a recurring audit of the SaaS stack and cloud costs is a spend-recovery lever that tends to surface spend that crept in as teams grew. That is Fiscallion's own practice, not an industry benchmark.
Two tempos: renewal discipline and real-time control
Not all software spend behaves the same way. Treating it as one pool is the first design error.
- Fixed subscriptions (per-seat tools, annual contracts) are controlled at the renewal date. The lever is the decision you make before the notice deadline.
- Usage-based spend (AI tools, consumption pricing) moves month to month. The lever is monitoring while the spend happens.
The Spendesk report shows why the second tempo matters. AI spend grew 340% in twelve months, and the median company's monthly AI bill swung by about 61% from one month to the next. An annual budget assumption cannot absorb that.
Spendesk recommends three finance levers: benchmarking, usage monitoring (cost per output), and real-time controls such as spending limits with alerts at the point of spend. You can use all three without buying anything first.
How to build the inventory
The inventory is the calculation layer. You cannot decide on spend you cannot see.
- Pull from every payment source. Export vendors from your accounting system, card statements and bank feeds. Subscriptions paid on a personal card and expensed show up in expense reports.
- Record one row per vendor. Capture vendor, owner, monthly spend, usage evidence, notice deadline and renewal decision.
- Flag shadow IT. Zylo reports that more than one-third of a company's applications are shadow IT. Any vendor with no owner is a candidate.
- Group by category. Put design, project management, collaboration, AI and analytics tools side by side so overlap is visible.
- Check where the cost is booked. Software tied to delivering the product belongs in cost of goods, not in a general engineering or marketing line.
Then calculate three simple ratios:
| Metric | Formula | What it tells you |
|---|---|---|
| Software spend per employee | Monthly software spend / headcount | Whether you pay more per head than disciplined buyers |
| License utilization | Active users / provisioned licenses | How many seats you pay for but do not use |
| Category overlap | Categories with two or more tools / total categories | Where consolidation is possible |
How to read the numbers
Start with price variance. Spendesk found that businesses buying the same software at a similar scale often paid two to four times more than the most disciplined buyers. Its size-controlled per-employee gaps for common per-seat tools make the point.
| Vendor | Lower per-employee bill (EUR/month) | Higher per-employee bill (EUR/month) |
|---|---|---|
| Atlassian | 2 | 12 |
| Notion | 2 | 7 |
| HubSpot | 6 | 24 |
Source: Spendesk European Spend Report 2026, June 2025 to June 2026, controlling for company size. Consumption-priced cloud vendors were excluded.
Headcount does not explain a gap of this size. Plan tier, seat hygiene, negotiated terms and billing frequency are the usual suspects to check in your own contracts.
Next, read duplication and utilization:
| Finding | Value | Source and period |
|---|---|---|
| AI-tool buyers paying for two or more competing products | 72% | Spendesk, June 2025 to June 2026 |
| Design-tool buyers doing the same | 64% | Spendesk, June 2025 to June 2026 |
| LLM-using companies running two or more competing tools | 29% to 65% year on year | Spendesk, within the same window |
| Provisioned licenses actually used | 49% | Zylo 2024 SaaS Management Index |
One caution on Zylo's data. Its dataset spans mid-size companies to large enterprises, including its reported average of $18M in annual wasted spend. Do not read that as a typical figure for a $5-100M ARR company. Use it as evidence that unused licenses are common, then measure your own.
Spendesk announced the report and its headline findings on LinkedIn, which is a useful reference for how finance teams frame the duplication story:
"Set the review date ahead of the notice deadline, not on the renewal date. A decision made after the deadline is a decision made by default."
— Aleksandar Stojanovic, CEO & Founder at FiscallionThe practical lesson is that this is a category problem, not a tool problem. Cutting one design tool is easy. Deciding which design tool the company standardizes on is the real decision.
What to do next: owners, rules and controls
Assign two owners per tool. Fiscallion's board-reporting approach separates the metric data owner from the decision owner. For software, that means one owner for the data, one for the decision.
- Data owner: usually finance or operations. Keeps spend, seat counts and notice dates current.
- Decision owner: the budget holder, or the CEO for company-wide tools. Decides whether the tool renews.
Then apply a renewal rule to every line, using four outcomes:
| Decision | Use it when |
|---|---|
| Renew | Usage evidence supports the seat count and the tool has a clear job |
| Consolidate | Another tool in the same category already covers the need |
| Reduce seats | Utilization is low but the tool is still needed |
| Renegotiate | Usage is fine, but your price per seat sits well above comparable buyers |
"If they are scattered across cards, invoices and team budgets, your burn and runway numbers are softer than they look."
— Aleksandar Stojanovic, CEO & Founder at FiscallionFor usage-based AI spend, set a spending limit and an alert at the point of spend, and track cost per output rather than the total bill. Feed the monthly figure into your forecast so runway reflects what you actually spend. If you need a starting structure, use the startup budget template as your baseline.
Common mistakes and the replacement move
- Tool first, ownership second. A spend platform will not decide anything for you. Replace it with an owner and a renewal rule for each line, then evaluate tooling.
- Treating renewals as finance admin. Finance can send the reminder, but the budget holder must decide. Replace the admin task with a dated decision.
- Judging a tool by its list price. The same tool can cost two to four times more at a typical buyer. Replace list price with price per employee and utilization.
- Running usage-based spend on an annual assumption. Replace it with monthly monitoring and alerts.
- Booking delivery-critical software in the wrong line. Replace it with a cost classification review so gross margin is accurate.
The software inventory and ownership checklist
Copy this table into a spreadsheet and fill one row per vendor. It is ungated, and nothing here requires an email address.
| Vendor | Owner (data / decision) | Monthly spend | Usage evidence | Notice deadline | Renewal decision |
|---|---|---|---|---|---|
| Example: design tool | Ops / Head of Design | Amount from invoice or card | Active users vs. provisioned seats, last login | Date contract requires cancellation notice | Renew, consolidate, reduce seats, or renegotiate |
Complete these checks before you call the inventory finished:
- Every vendor from accounting, cards and bank feeds appears in the sheet.
- Every row has a named data owner and a named decision owner.
- Every row has usage evidence, not an opinion about usage.
- Every notice deadline is in a shared calendar with a review date ahead of it.
- Every usage-based tool has a limit and an alert.
- Total monthly software spend feeds your burn and runway forecast.
Fiscallion is a fractional CFO and FP&A provider for B2B SaaS, working above the ledger. It does not sell spend management software and is not a bookkeeping firm. When software spend needs to be tied to burn, runway and hiring trade-offs, that is the fractional CFO layer. Every Fiscallion client works directly with Aleksandar Stojanovic at the CFO layer, a senior partner on every engagement rather than an account-manager or junior-team handoff. His background is FP&A leadership experience gained through the growth of a SaaS company to €100M ARR.
FAQ
What is SaaS spend management and why does it matter for startups?
It is the practice of finding every software subscription, assigning an owner, and deciding each renewal on purpose. It matters because software sits inside gross burn, so unmanaged subscriptions weaken your cash and runway picture. Zylo found companies use only 49% of provisioned licenses, though its dataset skews toward larger organizations.
How do I find out which SaaS subscriptions my company is actually paying for?
Start with payment data, not surveys. Export vendors from your accounting system, card statements and bank feeds, then add expensed subscriptions from expense reports. Any vendor without an owner is a likely shadow IT item, which Zylo reports makes up more than one-third of company applications.
What should I look for in a SaaS spend management tool?
Decide who owns the decision before you buy anything, because a tool shows spend but cannot make the call for you. Accountable senior judgment is still needed for renewals and trade-offs. Look for discovery across invoices, cards and bank feeds, an owner field per tool, a renewal calendar with notice dates, usage data, limits and alerts for usage-based spend, and an export that feeds your forecast.
How do I decide which software subscriptions to keep, renew, or cancel?
Use four outcomes: renew, consolidate, reduce seats, or renegotiate. Base each on usage evidence, overlap with another tool in the same category, and price per employee. Make the decision before the notice deadline, with the decision owner named in the inventory.
Decide the owner first
Software spend gets under control when each line has a person, a date and a rule. The tool comes after that, if you need one.
Build the inventory from the checklist above, assign the two owners, and set review dates ahead of every notice deadline. Then connect the total to your burn and runway model.
