Invoice Exception Handling: Turning the 20% That Breaks Automation Into a Manageable Queue
Every AP automation system has a straight-through processing rate — the percent of invoices that flow from receipt to approval to payment without human intervention. Modern systems can push this to 80% or higher with disciplined PO usage, clean vendor masters, and well-trained matching rules. But 100% is a fantasy. Some invoices will always require human touch — genuine disputes, unusual situations, errors from vendors, edge cases the rules don't cover.
The question isn't whether exceptions exist but how they're handled. AP teams that treat exceptions as an unstructured inbox get buried in backlog. AP teams with a structured exception workflow — clear categorization, routing rules, SLAs, and resolution patterns — process exceptions steadily alongside routine work and maintain cycle times that keep vendors happy.
Exceptions fall into categories with different resolution patterns:
Common invoice exception categories
- Missing PO — invoice received for purchase with no corresponding PO
- PO mismatch — invoice doesn't match PO on quantity, price, or line items
- Receipt mismatch — quantity received doesn't match quantity invoiced
- Price variance outside tolerance — unit price higher than PO within configured tolerance
- Missing documentation — lien waiver, COI, or other required attachment missing
- Coding ambiguity — system can't determine correct job/cost code
- Vendor not in master — invoice from unknown vendor
- Duplicate suspicion — invoice looks like a duplicate of another recent one
- Dispute — known dispute about amount, scope, or quality
- Tax issues — tax calculation differs from expected
Each category has a different typical cause, different resolution path, and different optimal routing. Treating all exceptions as one generic queue loses the specificity that enables fast resolution.
Different exception types should route to different resolvers:
Exception routing by type
- Missing PO — to the buyer's manager to create/approve retroactive PO or deny
- PO mismatch — to the PO owner (often PM or buyer) for reconciliation
- Receipt mismatch — to superintendent or field staff who would know what was actually received
- Price variance — to buyer or PM for approval or rejection
- Missing documentation — to AP clerk to request from vendor
- Coding ambiguity — to PM for clarification
- Vendor not in master — to AP lead to verify and add or reject
- Duplicate suspicion — to AP specialist for investigation
- Dispute — to PM and potentially commercial/legal team
- Tax issues — to tax specialist or AP lead
Routing matters because the wrong router can't resolve the issue efficiently. Sending a missing-PO exception to AP is less effective than sending it to the PM who should have created the PO — the PM has the context; AP doesn't.
Exceptions with no time pressure become a backlog. SLAs create pressure:
Typical exception resolution SLAs
- Missing PO — 2 business days
- PO mismatch — 3 business days
- Receipt mismatch — 3 business days
- Price variance — 2 business days
- Missing documentation — 5 business days
- Coding ambiguity — 2 business days
- Vendor not in master — 2 business days
- Duplicate suspicion — 1 business day
- Dispute — 5-10 business days depending on complexity
- Tax issues — 5 business days
SLAs are measured and tracked. Exceptions approaching or exceeding SLA are escalated. The goal is predictable timing — vendors and internal stakeholders know roughly how long resolution takes for different issue types.
When SLAs slip, escalation is automatic:
Exception escalation mechanics
- First escalation — SLA breached by 1 day, notification to the resolver's manager
- Second escalation — SLA breached by 3 days, escalation to department head
- Third escalation — SLA breached by a week, exception becomes VP-level attention item
- Automatic breach notifications to stakeholders — PM aware that their exception is overdue
- Weekly aged exception review — management looks at anything beyond SLA
Without escalation, exceptions stuck in someone's queue stay stuck. Escalation creates accountability and ensures attention flows to the longest-stuck items.
Exception aging tells you more about your AP operations than straight-through rate does. High STP rate with a growing backlog of aged exceptions isn't success — it's partial success masking a growing problem. Track both.
Some exceptions require vendor communication — incorrect invoices, missing documentation, disputed items:
Vendor communication for exceptions
- Template responses for common issues — consistent, professional wording
- Clear description of the issue and what's needed to resolve
- Specific response deadline to the vendor
- Contact person identified for follow-up
- Secondary escalation if vendor doesn't respond
Vendor communication is a differentiator. Vendors dealing with a professional, systematic exception response appreciate the clarity even when the news isn't what they wanted. Vendors dealing with ad-hoc inconsistent communication get frustrated and may apply pressure through other channels.
Get AP insights in your inbox
A short monthly roundup of construction AP + accounting posts. No spam, ever.
No spam. Unsubscribe anytime.
Persistent exception patterns reveal systemic issues that should be fixed:
Root cause patterns
- Specific vendor consistently generates exceptions — may need vendor-side changes or relationship review
- Specific PM consistently generates missing POs — training or process gap
- Specific project generates many receipt mismatches — site process issue
- Specific exception type dominates — automation rules may need tuning
- Exceptions concentrated in specific time period — may reflect volume surge or system issue
Exception data is operational intelligence. Regular pattern analysis (monthly is reasonable) identifies improvement targets. A PM generating 40 missing-PO exceptions a month is a training opportunity; a vendor generating 20 PO mismatches is a vendor-side conversation.
The cheapest exception is the one that never occurs. Prevention strategies:
Exception prevention strategies
- PO discipline improvements — reduces missing-PO exceptions
- Vendor master hygiene — reduces vendor-not-in-master and duplicate suspicion
- Cost code training for PMs — reduces coding ambiguity
- Receipt process improvements — reduces receipt mismatches
- Price tolerance tuning — fewer legitimate invoices flagged as variance
- Vendor communication about invoice requirements — reduces missing-doc exceptions
- Automation rule refinement — removes false positives that create unnecessary exceptions
Prevention typically has higher ROI than resolution speed. Moving 100 exceptions a month to zero through process improvements beats making 100 exceptions resolve faster.
An exception management dashboard provides operational visibility:
Exception dashboard elements
- Current open exception count by category
- Aging distribution — within SLA, breached, severely aged
- Exception count by resolver — who has what load
- Exception trend month over month — are we getting better or worse
- Top exception sources — vendors, projects, and categories driving volume
- Resolution time trend — are we resolving faster or slower
Dashboards enable daily and weekly operational decisions. "This resolver has 40 open exceptions and 15 are overdue" is actionable in a way that end-of-month aggregated reporting isn't.
How the AP team is structured affects exception handling:
AP team structure considerations
- Generalist AP clerks — everyone handles everything. Works at smaller scale
- Specialists by exception type — one person focuses on PO issues, another on documentation, etc. Better at scale
- Dedicated exception specialist — one role focused specifically on exceptions while others handle routine
- PM-side partner — someone in project management designated to respond to AP exceptions quickly
As volume grows, specialization usually pays off. A specialist handling 100 PO exceptions develops pattern recognition and efficiency a generalist handling 10 of each type never achieves.
Invoice exception handling in construction AP is where most AP operations struggle to scale. The answer is structure: categorize exceptions by type, route them to the right resolver, set SLAs, escalate breaches, communicate clearly with vendors, analyze root causes, prevent the preventable, and monitor through a dashboard. Done well, exception handling becomes a managed operational process with predictable throughput rather than a source of constant firefighting. The exception queue is also a goldmine of operational intelligence — the patterns in what breaks tell you where to improve processes upstream. Companies that treat exceptions as a managed workflow rather than an unstructured problem consistently report better AP cycle times, better vendor relationships, and better visibility into what's actually happening in their operations.
Written by
Sarah Blake
Head of Product
Former AP Manager at a $200M construction firm, now leads product at Covinly. Writes about what AP teams actually need from automation — beyond the marketing promises.
View all posts