“Automated” is one of those words that sounds great in a meeting and gets a little murkier once you look under the hood.
A team may capture invoices digitally and move files faster and still have approvals buried in inboxes and exceptions piling up. With a tool in place, it’s still hard to know whether costs are down, turnaround is better, or control has actually improved.
What matters in 2026 isn’t whether AP looks more digital. It’s whether the process is actually better. Only 4% of surveyed middle-market organizations said they had fully automated AP, while 89% were still running partial automation. Nearly half said they’d seen little to no cost savings. This leaves most organizations with a more modern-looking process that still isn’t delivering.
For controllers, the question is more specific: is the process actually improving? Are approvals cleaner? Are exceptions shrinking? Is the team spending less time chasing work and more time controlling it?
That’s where the right AP metrics come in. The right KPIs won’t just tell you whether automation exists. They’ll tell you whether it’s actually fixing the process or simply helping old problems move faster.
Finance teams are being asked to move faster while managing cost pressure, risk, and tighter scrutiny. Fraud is still part of that reality. In 2024, 79% of organizations experienced attempted or actual payments fraud activity, and 63% said business email compromise was the leading avenue for fraud attempts.
This makes AP performance about more than speed. Moving work faster doesn’t help much if the process is still inconsistent or exposed. A faster mess is still a mess, and a faster weak control is still a weak control.
cntrollers need a way to measure whether AP is becoming cleaner, tighter, and easier to manage, not just more digital. A useful measurement approach should do more than count transactions or confirm that invoices entered a system. It should show whether manual effort is falling, handoffs are getting cleaner, approvals are moving more smoothly, and payments are staying inside policy.
At the highest level, that means looking at AP from four angles: speed, workflow quality, control, and financial impact. Looking at all four together gives controllers a much more honest view of performance than any one metric can on its own.
Cycle time and approval turnaround still matter. Benchmarks like invoice-to-payment cycle time and approval-to-schedule cycle time give finance teams a concrete view of throughput.
Speed alone tells an incomplete story.
An invoice that gets routed quickly and then sits in exception handling for three days isn’t really a win. A payment that goes out on time after several manual follow-ups doesn’t point to a healthy process either. controllers need to look at cycle time alongside approval bottlenecks, exception volume, and manual touches per invoice.
A simple gut check helps here: are invoices moving faster because the process is better designed, or because the team is working around the system harder? When it’s the second one, the metrics give it away through rework, stalled approvals, and exception queues that never seem to shrink.
One of the clearest signs that AP automation is working is that certain types of work begin to fade into the background.
Manual routing should drop. Re-keying should drop. Chasing approvers should drop. Exceptions tied to inconsistent intake, poor coding, or weak matching logic should start shrinking too. When those things stay stubbornly high, the process hasn’t improved nearly as much as it looks on paper.
Exception rate should be near the top of the list. The percentage of invoices requiring rework matters too, along with the share of invoices that still need manual handoffs to keep moving. Metrics like invoices processed per FTE help controllers see whether capacity is truly improving or whether the same amount of effort is simply being buried in different places.
This is also where partial automation tends to fool people. A team may capture invoices digitally and automate routing, yet still rely on email, spreadsheets, and side conversations to resolve exceptions. The process looks more modern, but the old work is still there. The right indicators expose that instead of letting it hide behind cleaner screens and better-looking dashboards.
Speed without control isn’t improvement.
Start with first-time error-free disbursement rate. It shows whether the process produces clean outcomes, not just quick movement.
Other useful control metrics include duplicate payment rate, percentage of approvals completed within policy, vendor-change verification compliance, and the share of payments still executed outside the primary workflow. These aren’t optional reporting details. They show whether the workflow is staying inside policy or constantly making room for exceptions.
Fraud risk raises the stakes here. Recent data shows that business email compromise and vendor impersonation remain common threats. A process with weak approval discipline, poor visibility, or too many off-system handoffs isn’t just inefficient. It leaves the business more exposed.
Controllers also need a measurement approach that resonates beyond the AP team. It should connect workflow performance to cost discipline, staffing pressure, and finance outcomes leadership actually cares about.
Benchmarking data shows a major cost gap between high and low performers in AP. Top-performing organizations spend about $0.38 per $1,000 in revenue to process accounts payable, while bottom performers spend about $0.92. For a $1 billion company, that gap can translate into more than $500,000 in annual savings potential.
That’s why cost-to-process, on-time payment rate, discount capture, and invoices processed per FTE deserve a place in the approach. These measures connect workflow performance to something bigger than day-to-day activity. They show whether AP is becoming easier to scale, easier to manage, and less expensive to run.
They can also uncover a common trap. Some teams move more volume through the system, but the labor pressure doesn’t really change. If invoice volume climbs and headcount strain stays the same, some part of the process is still eating time. It may be intake quality, approval routing, or exception handling. Whatever the cause, these metrics should make the drag visible instead of letting it hide behind higher throughput numbers.
The most useful AP metrics don’t just sit on a dashboard. Together, they create a diagnostic tool.
When cycle time improves but exception rates stay flat, the process may be moving faster without getting cleaner. When invoices per FTE go up but supplier inquiries rise too, the team may be pushing more volume through without improving visibility. When more invoices enter the system digitally but too many payments still happen outside the workflow, automation isn’t actually connected end-to-end.
This is where the broader change-management point comes into focus. Technology on its own doesn’t fix weak handoffs, vague ownership, inconsistent policy, or approval paths that are hard to follow. In some cases, automation simply makes those issues move faster.
A strong measurement framework helps bring that into view. It gives controllers a way to tell the difference between meaningful improvement and a process that only looks more modern on the surface.
A healthier AP process shouldn’t be hard to spot. Approvals move with fewer follow-ups, exceptions shrink, payment activity stays visible, and policy holds up under pressure. The team spends less time reconstructing what happened and more time controlling what’s next.
Many teams still can’t prove that. They may have automation in place, but they can’t clearly show the process is better. And in 2026, that gap matters. Costs are under pressure, decisions have to move faster, and fraud risk isn’t easing up.
This is why these AP metrics matter. They show whether the workflow is actually improving or just looking more digital. When the metrics point to fewer touches, lower exception volume, stronger approval discipline, cleaner payments, and better operating leverage, progress is real. When they point to delays, workarounds, and off-system activity, they’re showing you where the process still needs work.
Here’s the bigger point: automation isn’t the win. Better process performance is.
The same principle applies to ERP adoption: turning on technology isn’t the same as improving how work moves. That’s the problem onPhase is built to solve, connecting capture, approvals, documents, workflow, and payments in a way that’s easier to measure and easier to trust.