Skip to content

Invoice Automation Rules

Automation rules determine which invoices can be automatically approved without manual review. All active rules must pass for an invoice to be processed automatically - if any rule fails, the invoice is flagged for manual review.

How rules work

The system applies validation rules in sequence. When an invoice is submitted:

  1. Basic validation - Checks for required information and proper formatting
  2. Rule evaluation - Each active automation rule is tested
  3. Decision - If all rules pass, auto-approve; if any fail, flag for review

Rule strategy

Start with conservative rules and gradually adjust based on your comfort level and experience with the system.

Available rules

Maximum invoice amount

Sets the highest total amount that can be auto-approved.

How it works: The system calculates the total of all line items on the invoice. If this exceeds your maximum amount, the invoice requires manual review.

Example scenario

Given: Max invoice amount set to £1,000

When: An invoice for £1,005 is submitted Then: Invoice requires review (exceeds limit)

When: An invoice for £995 is submitted Then: Invoice is auto-approved (within limit)

Best practice: Set this based on your typical invoice values and approval authority levels.

Maximum increase percentage

Prevents auto-approval of invoices that are significantly higher than recent similar invoices.

How it works: The system looks at the last 5 approved invoices from the same supplier to the same cottage owner, calculates the median amount, and compares the current invoice. If the increase exceeds your percentage threshold, it flags for review.

Example scenario

Given: Max increase set to 2% and recent invoices have a median of £100

When: An invoice for £105 is submitted (5% increase) Then: Invoice requires review (exceeds 2% limit)

When: An invoice for £100 is submitted (0% increase) Then: Invoice is auto-approved (within limit)

Best practice: Use 10-20% to catch unusual price jumps while allowing for normal variation.

Rate limiting

Controls how many invoices from the same supplier to the same cottage owner can be auto-approved within a time period.

How it works: The system tracks recent invoices and counts how many have been submitted within your specified number of days. If the current invoice would exceed the limit, it requires review.

Example scenario

Given: Max rate set to 2 invoices per 7 days

When: The 3rd invoice within 7 days is submitted to the same cottage & owner Then: Invoice requires review (exceeds rate limit)

When: The 2nd invoice within 7 days is submitted to the same cottage & owner Then: Invoice is auto-approved (within rate limit)

Best practice: Consider normal service patterns - weekly cleaning might need higher limits than one-off repairs.

Minimum previous invoices from supplier

Ensures new suppliers have a track record before auto-approval begins.

How it works: The system counts approved invoices from this supplier across all properties. Until they reach your minimum threshold, all their invoices require manual review.

Example scenario

Given: Minimum previous invoices set to 3

When: A supplier submits their 3rd invoice ever Then: Invoice requires review (hasn't reached minimum)

When: A supplier submits their 4th invoice ever
Then: Invoice is auto-approved (has reached minimum)

Best practice: Set to 2-5 invoices to establish trust while not being overly restrictive.

Minimum previous invoices to cottage owner

Requires suppliers to have an established relationship with specific property owners before auto-approval.

How it works: The system counts approved invoices from this supplier specifically to this cottage owner. Until they reach your threshold, invoices between them require review.

Example scenario

Given: Minimum previous invoices to cottage owner set to 1

When: A supplier submits their 1st invoice to a cottage owner Then: Invoice requires review (new relationship)

When: A supplier submits their 2nd invoice to the same cottage owner Then: Invoice is auto-approved (relationship established)

Best practice: Set to 1-2 to verify new supplier-property relationships while allowing established ones to flow automatically.

Owner funds protection

Prevents auto-approval of invoices that would consume too large a percentage of an owner's available funds.

How it works: The system calculates how much money the property owner is owed from upcoming guest payments within your specified timeframe. If the invoice amount exceeds your percentage of these funds, it requires review.

Example scenario

Given: Max owner funds set to 10% and owner is owed £1,000 in the next period

When: An invoice for £110 is submitted (11% of funds) Then: Invoice requires review (exceeds percentage limit)

When: An invoice for £90 is submitted (9% of funds) Then: Invoice is auto-approved (within percentage limit)

Best practice: Set the percentage based on typical maintenance costs and the timeframe based on your payment cycles.

AI confidence threshold

Ensures the system is confident about the information it extracted from uploaded invoice files.

How it works: When processing uploaded invoices, AI document analysis assigns confidence scores to extracted data. If the confidence for key information (amounts, dates, etc.) falls below your threshold, the invoice requires human verification.

Example scenario

Given: Minimum confidence set to 80%

When: Document analysis has 75% confidence in extracted data Then: Invoice requires review (below confidence threshold)

When: Document analysis has 85% confidence in extracted data Then: Invoice is auto-approved (above confidence threshold)

Best practice: Start with 70-80% confidence. Lower values allow more automation but risk errors; higher values ensure accuracy but require more manual review.

Enforce supplier relationship

Requires suppliers to have an established business relationship with the property before auto-approval.

How it works: The system checks if the supplier has been formally linked to the cottage or cottage owner in your supplier database. This prevents unknown vendors from having invoices automatically approved.

Example scenario

Given: Enforce supplier relationship is enabled

When: An invoice arrives from a supplier not linked to the property Then: Invoice requires review (no established relationship)

When: An invoice arrives from a supplier linked to the property
Then: Invoice proceeds to other rule checks (relationship verified)

Best practice: Enable this to prevent unauthorized or fraudulent invoices from unknown suppliers.

Supplier daily rate limit

Prevents any single supplier from overwhelming the auto-approval system with too many invoices per day.

How it works: The system tracks how many invoices each supplier has submitted in the current day. If a supplier exceeds 10 invoices in one day, additional invoices require manual review regardless of other rules.

Example scenario

Given: Daily rate limit is 10 invoices per supplier

When: A supplier submits their 11th invoice of the day Then: Invoice requires review (exceeded daily limit)

When: A supplier submits their 5th invoice of the day Then: Invoice proceeds to other rule checks (within daily limit)

Best practice: This is a system-level protection that cannot be disabled. It prevents abuse and ensures human oversight for high-volume scenarios.

Duplicate prevention

Blocks auto-approval of invoices that appear to be duplicates based on external invoice numbers.

How it works: When an invoice includes an external reference number (supplier's invoice ID), the system checks if that same number has been used before by the same supplier. If found, it prevents duplicate processing.

Example scenario

Given: Supplier has previously submitted invoice #INV-001

When: The same supplier submits another invoice #INV-001
Then: Invoice requires review (duplicate reference detected)

When: The supplier submits invoice #INV-002 Then: Invoice proceeds to other rule checks (unique reference)

Best practice: This automatic protection helps prevent paying the same invoice twice, which is especially important with manual invoice entry.

Basic data validation

Ensures invoices have the minimum required information for processing.

How it works: The system verifies that invoices have: - At least one line item with a description and amount - Non-negative total amounts - Proper cottage and supplier assignment - Valid data extraction from uploaded files

Example scenario

Given: An invoice is submitted

When: The invoice has no line items or amounts Then: Invoice requires review (incomplete data)

When: The invoice has negative amounts Then: Invoice requires review (invalid amounts)

Best practice: This fundamental validation cannot be disabled and ensures data integrity before any business rules are applied.

Rule levels

Global rules

Apply to all suppliers by default. Set these as your baseline automation policies.

Access: Navigate to Finance > Invoices, then click the automation rules button (gear icon)

Supplier-specific rules

Override global rules for specific suppliers. Use these for trusted vendors or those requiring special handling.

Access: Go to People > Suppliers > select supplier > Invoice Rules tab

Rule hierarchy

Supplier-specific rules override global rules where defined. If a supplier rule is not set, the global rule applies.

Setting up automation

Initial setup

  1. Start conservative - Set strict rules initially
  2. Monitor results - Watch approval/review ratios
  3. Adjust gradually - Loosen rules as you gain confidence

Monitoring and adjustment

  • Review flagged invoices to identify patterns
  • Adjust rules if too many legitimate invoices are flagged
  • Tighten rules if questionable invoices are auto-approved
  • Consider seasonal variations in service frequency and costs

Troubleshooting

Too many invoices flagged for review

  • Check if maximum amounts are too low for current costs
  • Review if increase percentages are too strict
  • Consider if rate limits match actual service patterns
  • Verify AI confidence threshold isn't too high

Inappropriate auto-approvals

  • Lower maximum amounts temporarily
  • Tighten increase percentage limits
  • Add or strengthen relationship requirements
  • Increase AI confidence requirements

Next steps