For decades, enterprises have invested billions of dollars in software to make procurement more efficient. They've implemented sophisticated ERP systems, digitized purchasing and payments, and built increasingly complex technology stacks to manage their operations.
Yet the average procurement team still spends 43% of its time on tactical work, including entering data, validating documents and chasing approvals.
That's one of the findings from a new report by Ardent Partners, The S2P Exception Tax: Exposing the Hidden Cost of Operational Rework. The research quantifies a problem I’ve seen firsthand at some of the world's largest enterprises, where despite decades of investment in automation, people are still doing much of the work required to keep business operations moving. What’s more, the cost goes far beyond employee time.
The hidden cost of procurement exceptions
Enterprise software has become very good at managing transactions when everything goes according to plan. But in the real world, things rarely follow the “happy path.”
Suppliers miss deliveries, purchase orders change as production requirements shift and invoices don’t always match the agreed-upon price. Each of these situations creates an exception that someone has to investigate and resolve. For large manufacturers managing thousands of suppliers and millions of transactions, these problems arise constantly.
Ardent Partners' research found that nearly one in five invoices (19.9%) requires exception handling. And, when procurement transactions require manual intervention, processing costs are 60–80% higher than for fully automated transactions.
We define the cumulative burden of this manual work as the “Exception Tax.” This tax is felt across five interconnected costs that enterprises often struggle to see in full:
- Labor and rework: Employees spend hours investigating discrepancies and coordinating fixes across departments and systems.
- Delays: Unresolved exceptions can hold up orders and shipments, disrupting production schedules and supplier relationships.
- Value leakage: Pricing discrepancies, missed discounts and payment errors can erode savings and increase costs.
- Operational and compliance risk: Manual processes increase the likelihood of errors, inconsistent decisions and compliance issues.
- Diverted capacity: Skilled employees spend time resolving routine problems instead of strengthening supplier relationships and managing risk.
These costs rarely show up together on a single dashboard. Procurement might see the time spent resolving a purchase order change, while finance sees the delayed invoice and operations feels the impact of a late delivery. No single team sees the full picture, making it difficult to understand the cumulative impact of thousands of small disruptions.
Why we chose the direct procurement challenges others left behind
While much of the procurement software industry has historically focused on indirect spend, Naveen and I saw an opportunity to tackle the complex, exception-heavy workflows in direct procurement that traditional solutions have struggled to automate. That's why we started Fragment with a focus on the world's largest manufacturers, where even small disruptions can have significant consequences for production and supply chains. Ardent's research underscores the scale of the challenge, finding that 68% of direct procurement teams identify purchase order processing as a significant source of buyer workload.
For years, companies have added people and software to manage procurement complexity, yet buyers still spend their days investigating discrepancies and coordinating fixes. Their decisions shape both the cost of materials and whether those materials arrive in time to keep production running. That expertise has enormous business value, but too much of their time goes toward tactical firefighting.
We saw an opportunity to take on this unresolved work, freeing buyers to apply their judgment where it matters most. We built our Autonomous Context Engine (ACE) to give agents the operating knowledge that work requires. The ACE connects company data, policies and institutional knowledge so agents can investigate issues and carry out authorized resolutions across existing systems, escalating decisions that require human judgment.
By reducing the burden of investigation and coordination, Fragment gives procurement professionals more time to anticipate shortages, negotiate better terms and strengthen supplier relationships. For manufacturers operating on tight margins, those decisions can protect profitability, keep production moving and support growth.
It's time to change what we expect from enterprise software
For the first time, advances in AI, combined with the company-specific context provided by Fragment's Autonomous Context Engine, are making it possible for agents to take on complex operational work that traditional enterprise software has left to people.
But access to data alone isn't enough. Agents need to understand how a company operates, apply its specific policies and business rules, and take action across systems to resolve issues from start to finish. The ACE makes that possible by turning fragmented enterprise information into the operational context required to do the work.
The Ardent research makes clear just how much enterprises are paying for work that their software was supposed to make easier. And as operations become more complex, that burden will only grow unless we fundamentally change how the work gets done.
We've spent decades building software that helps people manage procurement. Now we have an opportunity to build software that actually does the work, freeing enterprises to operate more efficiently and their people to focus on what matters most.
That's the future we're building toward at Fragment. If you like solving hard, complex problems and want to join us, we’re hiring.
Check out the full Ardent Partners report here.