SaaS Spend Analysis: Find Duplicate and Maverick Spend

Most finance teams already have software spend data, but a transaction export only shows who was paid. It doesn’t show which tools overlap, which contracts are expensive for your company’s size, where teams buy outside policy, or which subscriptions have no owner. That’s where SaaS spend analysis comes in, because it turns raw payments into decisions you can act on.

Spendesk analysed invoice and card-payment data from thousands of European companies and found significant variation in what comparable businesses pay for the same software, alongside widespread tool duplication.

Capturing spend is only the first step, and the value comes from structuring it well enough to act on. This guide walks you through centralising and categorising spend, spotting duplicate tools and price outliers, detecting maverick spend, and prioritising savings. You’ll also see how to turn your findings into approval and renewal controls that keep costs down.

Key takeaways:

• Software spend data becomes useful once you centralise it from every payment route, including invoices, company cards, expense claims, and direct debits.

• Normalising vendor names and grouping vendors into categories makes duplicate and overlapping tools visible.

• Comparing cost per employee or per active user helps you spot price outliers worth investigating.

• An outlier isn’t proof of overpaying, so check the context before acting.

• Maverick spend often hides in small recurring payments with no owner, purchase request, or approved budget.

• Rank findings by financial value and actionability, then turn repeated problems into approval and renewal controls.

What is SaaS spend analysis?

SaaS spend analysis is the process of collecting, cleaning, categorising and analysing software spending to understand:

• what you’re buying;

• who you’re buying it from;

• who owns the cost;

• what each product does;

• where products overlap;

• whether pricing is reasonable;

• and where controls or savings opportunities exist.

The process turns raw transaction data into questions finance can answer. For example:

Transaction data:
€60,000 paid to five AI vendors.

Spend analysis:
Three departments bought overlapping products, including two subscriptions with no named owner. One vendor’s cost has doubled, and an upcoming renewal represents a consolidation opportunity.

That second view is much more useful.

Why invoice data alone is not enough

Software isn’t always bought through accounts payable. It can appear through:

• supplier invoices;

• company cards;

• employee expenses;

• direct debits;

• and other payment methods.

This matters particularly for newer software categories. Spendesk’s research combined invoice and card data because relying on one payment rail alone would miss significant technology spending. For example, card payments represented a large share of spend with some AI vendors in the analysed dataset. A SaaS analysis built only from invoice data can therefore give finance an incomplete picture. The first step is to centralise the data.

1. Create a clean vendor and category taxonomy

Raw supplier data is messy. The same company can appear under:

• different legal entities;

• local subsidiaries;

• abbreviated names;

• merchant descriptors;

• or separate invoice and card records.

Normalise vendors before analysing spend, because records such as OPENAI and OPENAI IRELAND LTD may need to roll into one canonical vendor. Spendesk used the same principle on its European transaction dataset, normalising supplier identities across invoices and card transactions before evaluating software spending. Then categorise each vendor. A useful taxonomy might include:

• AI;

• CRM;

• design;

• collaboration;

• project management;

• finance;

• HR;

• development;

• data;

• cybersecurity;

• marketing;

• customer support.

Without categories, duplication is difficult to see.

2. Find duplicate and overlapping tools

Once vendors are categorised, look at how many products exist within each category. Suppose you find:

AI

• ChatGPT

• Claude

• Gemini

• Dust

Different products may have different purposes, so several products in one category do not automatically indicate a problem. The investigation should therefore ask:

• Who owns each product?

• Which employees use it?

• What problem does it solve?

• What functionality overlaps?

• What would break if we cancelled it?

• Was the overlap deliberate?

• Could licences be consolidated?

The objective is to identify overlap that nobody deliberately chose, rather than reduce every category to one vendor. Spendesk found that 72% of AI buyers and 64% of design-tool buyers in its sample paid for multiple competing products.

3. Identify price outliers by company size or department

A large SaaS bill isn’t necessarily a bad one, because a large company usually spends more than a small company. Spend analysis therefore needs to normalise costs, and these measures help:

• Vendor spend per employee

• Vendor spend per active user

• Category spend per employee

• SaaS spend as a percentage of departmental budget

Spendesk compared companies of similar sizes and still found substantial differences. The report found wide price variation for common tools, with businesses sometimes paying several times more per employee than comparable buyers.

A price outlier isn’t proof you’re overpaying, but it’s a strong reason to investigate:

• plan tier;

• licence count;

• contract date;

• product mix;

• utilisation;

• and negotiating history.

4. Detect maverick and unapproved spend

Maverick spend is purchasing that happens outside the organisation’s agreed process or preferred suppliers.

In software, it can look like:

• an employee buying a tool on a company card without approval;

• a department purchasing software despite an existing company licence;

• a subscription charged through expenses;

• a free trial converting into a paid plan;

• or a team choosing a new supplier outside procurement.

Look for transactions that have:

• no purchase request;

• no owner;

• no department;

• no recognised contract;

• no approved budget;

• or no obvious relationship to the company’s existing software stack.

Small recurring payments deserve particular attention. They may never individually trigger an investigation, but dozens of them can create significant shadow SaaS spend.

5. Compare ongoing and one-off spend

Not every transaction should be treated the same way, so it helps to separate software spending into ongoing and one-off groups.

Ongoing spend

Some ongoing costs are contractually committed, such as a £60,000 annual CRM contract, while others are recurring subscriptions that repeat monthly or annually, such as £500 per month for design software.

One-off spend

A one-off cost is a single purchase or temporary project, such as a short-term software licence used for an event.

This distinction helps finance understand what can realistically be changed. You can’t necessarily remove committed spend tomorrow, but an unmanaged monthly card subscription may be easy to cancel.

6. Prioritise savings opportunities

Once the data is clean, you’ll probably find dozens of things worth investigating, but they shouldn’t be treated equally. Rank opportunities by financial value × actionability, where actionability reflects confidence in the finding and ease of action. For example:

This ranking stops finance spending hours chasing tiny expenses while large contracts go unquestioned. In our analysis, comparable businesses often paid multiples of what more disciplined buyers paid for the same products, which makes price variance particularly important to review.

7. Turn analysis into approval and renewal controls

Spend analysis shouldn’t end with a spreadsheet of savings opportunities, because every repeated problem should create a control.

The following pairings turn each common finding into a control that stops the same problem recurring:

• Duplicate purchases: Introduce software requests and category checks

• Unused licences: Introduce pre-renewal usage reviews

• Unexpected auto-renewals: Centralise renewal dates

• Maverick spending: Tighten approval or payment controls

• Price outliers: Add benchmarking to major renewal negotiations

• Unpredictable AI bills: Add thresholds and alerts

Analysis finds the leak, and process changes stop it returning.

A SaaS spend analysis checklist

Use this process when reviewing your software portfolio.

Collect

• ☐ Invoice transactions

• ☐ Card transactions

• ☐ Recurring payments

• ☐ Expense data

• ☐ Contract information

Clean

• ☐ Standardise vendor names

• ☐ Merge duplicate supplier records

• ☐ Remove non-software spend

• ☐ Categorise every vendor

Enrich

• ☐ Add department

• ☐ Add cost owner

• ☐ Add licence count

• ☐ Add active-user information

• ☐ Add renewal date

• ☐ Add contract value

Analyse

• ☐ Calculate spend by vendor

• ☐ Calculate spend by category

• ☐ Calculate spend per employee

• ☐ Identify duplicate tools

• ☐ Find pricing outliers

• ☐ Flag unapproved vendors

• ☐ Review rapid spend increases

Act

• ☐ Cancel unnecessary products

• ☐ Consolidate overlapping tools

• ☐ Reduce licences

• ☐ Renegotiate contracts

• ☐ Assign missing owners

• ☐ Add renewal reminders

• ☐ Introduce approval controls

Spend analysis is valuable when it changes behaviour

Finance doesn’t need another dashboard showing that software costs money. It needs information that leads to a decision. In the European Spend Report, it is estimated that pricing differences among comparable SaaS buyers could represent roughly €7 billion in annual overpayment, and tool duplication another €3 billion opportunity, across European businesses with 50 to 250 employees.

Not every company can cut software spending by a fixed percentage, so teams shouldn’t assume that. The more useful conclusion is that significant differences can go unnoticed when software spend isn’t categorised or reviewed against benchmarks. A good SaaS spend analysis process exposes those differences, and updating spend management processes stops the same problems from quietly returning.

Read the European Spend Report 2026 for the full analysis behind these findings.

For greater oversight of company spending, explore Spend controls and approval workflows. If you’re evaluating a more centralised approach to managing business spend, book a Spendesk demo.

Frequently asked questions about SaaS spend analysis

How often should finance teams run a SaaS spend analysis?

A full review is usually most useful ahead of budget planning and before major renewals. In between, lighter checks on new vendors and recurring card payments often keep surprises small. The right rhythm depends on how quickly your software stack and headcount change.

Who should be involved in a SaaS spend analysis?

Finance usually leads because it holds the payment data. Department cost owners and IT or security colleagues who know the tools well can then help confirm who uses each one, why it was bought, and what would break if it were cancelled.

Can you run a SaaS spend analysis in a spreadsheet?

Yes, a spreadsheet can work well for a first review, especially when your software stack is small. The trade-off is manual effort, because it gets harder to keep current as vendors, payment routes, and renewal dates multiply. Many teams start here, and that’s a sensible place to begin.

What should you do when a subscription has no clear owner?

Start by identifying who’s paying for or using it, for example the cardholder or the person who submitted the expense claim. Then assign an owner, or agree to pause or cancel it before the next renewal, so it doesn’t quietly renew without anyone deciding to keep it.