I keep coming back to that game when I think about enterprise AP.
"When the starting configuration changed, USA adapted in real time and won. When a supplier changes their invoice format, most legacy AP systems don't step up. They step out."
The invoice lands in an exception queue. A human picks it up from there. The game stops. That is the template tax. And I'll be direct: most AP directors have never actually added it up. I think it's the most underreported cost in enterprise finance.
What the template tax is
The template tax is not the price you paid to implement your AP capture platform. It's not the annual licensing fee. Those costs are visible. They're on an invoice somewhere.
The template tax is the cost that accumulates invisibly across three separate budget lines that nobody has ever totaled together: template creation and maintenance labor, exception queue management, and the cash cycle cost of every processing gap, every onboarding delay, and every format-change outage that your DPO calculation never captures.
According to Ardent Partners' AP Metrics That Matter, more than 60% of invoices in legacy AP environments still require a human touch. That number has barely moved in five years. The technology has evolved. The exception rate hasn't. That is not a coincidence. It is an architecture problem. And I've been making this argument in boardrooms for three years.
Cost center 1: Template labor
Every supplier relationship in a template-based AP system requires a configured template. New supplier onboards? Someone builds a template. Supplier changes their invoice format? Someone updates it. Supplier starts sending invoices across multiple formats for different product lines? Multiple templates. International supplier with different format conventions? A new configuration cycle.
This is not a one-time cost. It's a maintenance commitment. Every new edge case the system can't handle gets absorbed as a rule, a configuration block, a line of logic that someone has to write, test, and redeploy. That logic becomes interdependent. A change made for one supplier can cascade. IT tickets follow. Redeployment cycles follow.
This is not a flaw in any particular vendor's implementation. It's the architecture. Template-based systems absorb variation as code. That code accumulates. Unlike the software license, it never appears on a single invoice.
Cost center 2: Exception rate
Industry benchmarks put legacy AP exception rates between 15 and 20 percent on template-OCR. At Auditoria, we see it run higher, particularly in companies operating across multiple geographies and currencies, where supplier format variability is wide and template coverage is perpetually one cycle behind. Compare that to sub-5 percent on agentic AP, where the system identifies the vendor before it reads the full invoice and adapts without a template, without a configuration cycle, without a ticket.
But here is what most people miss: the real cost of each exception is not the reviewer time. It's the cycle time. An invoice sitting in an exception queue is an invoice not posted. An invoice not posted is a liability you can't see, a payment you can't time precisely, and a DPO calculation running on incomplete data. Exceptions are not a processing cost. They are a cash cycle visibility problem. Every single one.
Cost center 3: Cash cycle cost
When a new supplier can't be onboarded for two weeks because no one has built their template yet, that is not just an AP operations delay. That is two weeks of invoices sitting outside your DPO calculation. Your cash position reporting has a gap. Your forecast is missing a liability.
When a format change causes three days of manual processing while the template is rebuilt, that is three days where your touchless rate has collapsed, your exception queue has grown, and your AP cycle time has increased. All of which shows up in your cash cycle latency before it surfaces in your month-end report.
Visibility, control, predictability, accountability. The four pillars of cash cycle ownership. A 60 percent human-touch rate doesn't survive any of them. This is not a theoretical argument. It is what I see in production environments every week.
What this looks like at scale
Here is a real example, with one detail changed.
A publicly traded consumer technology company, operating dozens of entities across its US headquarters, European subsidiaries, and Asia Pacific operations, came to us with an AP environment that had quietly become unmanageable. Every new market entry meant a new supplier set. Every new supplier set meant a new template configuration cycle. The IT team managing AP templates had effectively become a permanent maintenance crew, running continuously just to keep pace with the business.
The exception rate across their global entity stack was running well above benchmark. Not because the team wasn't capable. They were excellent. But because they were running a system designed for a different era: fewer suppliers, lower format variability, single-geography operations. The architecture was 2005. The supplier environment was 2025.
The migration to agentic AP eliminated the template maintenance cycle entirely. The system now identifies each supplier on the first read and adapts per entity, per format, per currency, without a template. Cash visibility across the entity stack is real-time rather than batch-processed. The template tax didn't disappear. They stopped paying it.
The legacy AP OCR generation
I want to be honest about something, because I think it gets lost in vendor conversations.
The legacy AP OCR category — including platforms like Hyland Brainware, ABBYY FlexiCapture, and Kofax — was built for a different AP environment. That environment had manageable supplier variability, lower supplier counts, and higher format standardization. In that context, template-based capture was a genuine advance over fully manual processing. It delivered real value for many organizations over many years. I'm not here to dismiss that.
What I take issue with is the assumption, still common in 2026, that these platforms fit the enterprise AP environment that most of you are actually running. Hundreds of active suppliers. Significant format variability. International invoices. Non-PO documents. Multi-currency entity stacks. And a finance team being asked to produce real-time cash visibility from a batch-processing architecture that was never designed for it.
Here is the question I ask every AP director I meet who is running legacy OCR: when was the last time your system handled a supplier format change without a ticket? If the answer is never, you are paying the template tax every day.
The path forward is not more template investment. It's a different architecture with a different destination. Not just better invoice processing. Cash cycle ownership.
The template tax is calculable
Three inputs. One output.
The number of templates your team currently manages. Annualized maintenance hours per template. The cost of your current exception rate in reviewer hours and cycle time. Plus the supplier onboarding delay cost measured in daily DPO exposure. That is your template tax. For most AP organizations running legacy OCR at scale, the number is larger than the software licensing cost. Often significantly larger. When I show this calculation to CFOs, the room usually goes quiet for a moment.
If you want to run this against your own numbers and see what templateless extraction looks like on your actual invoice set, reach out or book 20 minutes.
I'll bring the calculator.
Reach me at rohit@auditoria.ai
Rohit Gupta is CEO and Co-Founder of Auditoria.AI. The Auditoria platform processes invoices across global enterprise environments using agentic AI, no templates required.
Source: Ardent Partners, AP Metrics That Matter, 2025.