Most finance leads at growing businesses can tell you how much they spend on software. Fewer can tell you what they approved to buy this month before the invoice arrived.
That gap is where purchase order software is supposed to help. The problem is that most of the tools on the market were built for procurement teams in large companies, and the features that matter at a 200-person company without a dedicated buyer are often buried under features it will never use.
This guide explains what purchase order software actually does, what to look for at the 50-200 employee stage, and why many of the names that come up in search results probably aren't the right fit.
What a purchase order actually is (and why it matters before the invoice arrives)
A purchase order is a formal document your business sends to a supplier before goods or services are delivered. It records what you agreed to buy, at what price, and under what terms. When the supplier's invoice arrives, you can match it against the original PO to confirm the charge is correct.
For businesses with three employees, this is unnecessary. For businesses with 50 or more, it starts to matter. Team leads start purchasing things without telling finance. Invoices arrive that no one can trace back to a decision. Suppliers charge differently from what was verbally agreed.
This is not an edge case. Companies without a structured purchase approval process often end up with what procurement teams call maverick spend, purchases made outside any agreed contract or supplier list. The downstream effects compound quickly.
Companies lose between 10-20% of their targeted savings due to maverick buying, as purchases get made outside negotiated agreements.
That figure tends to surprise people. The reason is that maverick spend does not look like a problem from inside any single budget line. It shows up in aggregate, when you compare what you actually paid against what a coordinated approach would have cost.
The scale of the problem for growing companies
Indirect spending, meaning everything your company buys that does not directly go into your product or service delivery, adds up faster than most finance leads expect. At the 100-200 employee stage, this typically includes software subscriptions, agency fees, contractor hours, facilities costs, and a long tail of smaller purchases that nobody formally approved.
In many industries, indirect goods and services run on the order of 10% of revenue, and often more.
Deloitte, CFO Insights: Targeting Procurement, September 2021
For a company doing $10M in revenue, that is $1M or more in indirect spend. Most of it moves through credit cards, informal approvals, or monthly invoices that get paid without anyone checking whether they match an original purchase agreement. Purchase order software is one part of the fix, but only if you choose something appropriate to your actual situation.
What purchase order software actually does
At its core, a PO tool does three things.
First, it creates a digital trail from request to approval to receipt. Someone requests a purchase, a manager approves it, the supplier is notified, and when the goods or invoice arrives it can be matched against the original request.
Second, it enforces spending limits and approval tiers. A $200 purchase from a regular supplier might not need finance sign-off. A $5,000 contract with a new agency should. A good PO tool lets you set those thresholds without the whole system becoming a bottleneck.
Third, it feeds your accounting records. The approved purchase is logged, matched to the invoice, and pushed to your accounting software so there are no surprises at month-end.
What it does not do: it does not replace your accounting system, it does not manage vendor relationships, and it does not help you negotiate better rates. It is a control layer, not a strategy tool.
Where the well-known tools fall short for smaller businesses
When you search for purchase order software, a few names come up repeatedly. Procurify, Tradogram, and Airbase all have genuine strengths. They also share a common trait: they were designed for businesses with procurement teams, larger budgets, and more process sophistication than most 50-200 person companies have.
Procurify is solid if you have someone whose job it is to manage procurement. If you are the person who does procurement on top of three other responsibilities, the complexity will work against you. Tradogram has good pricing tiers but the interface requires setup investment that takes time most ops teams do not have. Airbase does more than POs and is useful if you are consolidating expense management, but it comes with a pricing model that assumes a larger organisation.
None of these are bad products. They are just aimed at a different buyer. The risk is that you implement one, spend months getting it configured, and end up with something your team half-uses because it adds friction to the existing process rather than replacing it.
What actually fits the 50-200 person stage
What you are looking for at this company size is usually simpler than the enterprise tools suggest.
An approval workflow that matches how your business actually makes decisions. Most companies at this stage have two or three people who approve purchases depending on type or amount. The tool needs to reflect that, not a six-tier approval hierarchy built for a 2,000-person procurement function.
Integration with your accounting platform. If you are using Xero, your PO tool should push approved purchases directly to Xero so that reconciliation at month-end does not involve anyone manually re-entering data. This sounds obvious, but it is worth verifying before you commit to a product.
Vendor-facing simplicity. When you send a PO to a supplier, they should be able to confirm it without needing to set up an account or learn a new platform. If the supplier experience is complicated, the tool will create friction you did not have before.
Spend visibility. The primary reason most companies introduce PO software is that they have lost visibility into what is being purchased and why. A dashboard that shows pending approvals, open orders by team, and month-to-date spend by vendor is more useful than advanced analytics features you will never configure.
For context, this connects directly to the broader challenge of vendor spend management, which is about having a clear view of who you are spending money with and whether that spend is under control. Purchase orders are one mechanism for building that visibility, but they work best when they are part of a wider approach to how you manage vendor relationships. You can also read about SaaS spend management if software costs are a particular area of concern.
The approval workflow problem worth naming
The most common reason PO processes fail in smaller companies is not that they lack software. It is that the approval step adds delay, so people work around it.
If getting a purchase approved takes two days, and someone needs to buy something today, they will use a credit card. This is rational behaviour. The system failed, not the person.
The fix is not stricter enforcement. It is reducing the approval time to the point where going through the process is faster than the workaround. That means mobile-accessible approvals, clear delegation rules when the approver is unavailable, and automatic approval for routine purchases below a defined threshold.
Software can support all of this. But the configuration has to match how your team actually works, not how a procurement consultant thinks teams should work.
The bottom line
Purchase order software is worth introducing at the 50-200 person stage, but only if you select something that fits your actual workflow rather than a product that requires you to build new processes to justify its features.
Start by mapping your current purchase approval process, however informal it is. Then find a tool that digitises that process with minimal friction, integrates with your existing accounting platform, and gives you the spend visibility you are currently missing.
The goal is control without added overhead. The right tool makes that achievable. The wrong one just adds a new form of paperwork.
