AP automation delivers transformative results when implemented well. But it can also disappoint when common mistakes undermine the initiative. After helping dozens of organisations implement AP automation on Oracle Fusion Cloud, Sharpe Project Consulting has identified the most frequent mistakes — and, more importantly, how to avoid them.
If you are planning, implementing, or optimising AP automation, this guide will help you sidestep the pitfalls that trip up others.
Mistake 1: Automating a Broken Process
The mistake: Taking your existing manual AP process and automating it as-is, without addressing underlying inefficiencies, data quality issues, or unnecessary steps.
Why it happens: Teams want quick results and assume that simply adding technology will solve the problem. Process redesign feels risky and time-consuming.
The consequence: You end up with an automated version of a bad process. Automation amplifies both efficiency and inefficiency — if your process includes unnecessary approval steps, redundant checks, or poorly designed exception handling, automation will execute these flawed steps faster but will not eliminate them.
How to avoid it: Before implementing automation, map your current process and challenge every step:
- Does this approval step add value, or is it a legacy requirement?
- Are these tolerances based on data, or are they arbitrary?
- Are these exception categories still relevant?
- Is this report actually used?
SPC3's implementation methodology includes a process optimisation phase specifically designed to address this issue before technology is deployed.
Mistake 2: Neglecting Master Data Quality
The mistake: Implementing automation without first cleaning up supplier master data, chart of accounts, and PO data in Oracle Fusion.
Why it happens: Data cleansing is unglamorous work that teams defer in favour of moving quickly to "the good stuff" — technology deployment.
The consequence: Automation depends on clean master data for validation, matching, and coding. Dirty data generates false exceptions, failed matches, and incorrect account coding. Your auto-match rate stays stubbornly low, and the automation appears to underperform.
How to avoid it: Invest in data cleansing before or in parallel with automation implementation:
- Merge duplicate supplier records.
- Validate supplier tax registrations and bank details.
- Review and correct PO pricing and quantities.
- Disable expired GL account combinations.
- Standardise tax code usage.
Mistake 3: Setting Unrealistic Tolerances
The mistake: Setting match tolerances either too tight (generating excessive exceptions) or too loose (letting errors through).
Why it happens: Without data analysis, tolerance settings are often based on guesswork, inherited from legacy systems, or set by someone who left the organisation years ago.
The consequence: Tight tolerances destroy your auto-match rate. If a $0.01 rounding difference triggers a hold, you are creating manual work for commercially insignificant variances. Loose tolerances allow genuine errors through, creating overpayments and audit risk.
How to avoid it: Analyse your historical matching data before setting tolerances:
- What is the distribution of price variances by supplier and item category?
- What percentage of quantity variances are within 2%? 5%? 10%?
- What is the average tax rounding difference?
Set tolerances at a level that captures 90-95% of invoices on first pass while maintaining meaningful financial controls. Review and adjust tolerances quarterly based on actual performance data.
Mistake 4: Ignoring Change Management
The mistake: Treating AP automation as a technology project rather than a people-and-process change initiative.
Why it happens: The technology is the visible, tangible part of the project. Change management is less visible and harder to measure.
The consequence: AP staff resist the new system because they were not involved in its design, do not understand how their roles change, or fear that automation threatens their jobs. Approvers ignore new workflow notifications because they were not trained. Suppliers continue submitting paper invoices because no one communicated the new process.
How to avoid it:
- Involve AP staff early. They understand the current pain points better than anyone and will champion automation if they help design it.
- Communicate role changes. Be explicit about how jobs will change — from data entry to exception management, from processing to analysis. Emphasise that automation frees them for more valuable work, not that it replaces them.
- Train thoroughly. Do not assume the new system is self-explanatory. Provide structured training and ongoing support.
- Engage suppliers. Communicate new invoicing channels, PO reference requirements, and payment process changes to your supplier base.
Mistake 5: Going Big Bang Instead of Phased
The mistake: Trying to automate everything at once — all invoice types, all suppliers, all business units, all capabilities — in a single deployment.
Why it happens: Enthusiasm, executive pressure for rapid results, or a desire to minimise the transition period.
The consequence: Complexity overwhelms the project team. Issues that would be manageable in a limited rollout become crises when multiplied across the entire organisation. The team loses confidence, and the project stalls.
How to avoid it: Implement in phases:
- Start with automated capture for your highest-volume invoice channel and top 50 suppliers.
- Enable automated matching for PO-based invoices from these suppliers.
- Expand to additional suppliers and invoice types.
- Add approval automation and duplicate detection.
- Optimise and expand to remaining scope.
Each phase delivers measurable value and builds confidence for the next.
Mistake 6: Not Measuring Baselines
The mistake: Implementing automation without first measuring your current state metrics — cost per invoice, cycle time, auto-match rate, exception rate.
Why it happens: Baseline measurement requires effort, and teams are eager to start the project.
The consequence: You cannot prove the ROI of your automation investment. When executives ask "what did we get for our money?", you cannot provide a credible answer. This makes it harder to justify further investment and can undermine confidence in the project.
How to avoid it: Measure baselines before you start. At minimum, capture:
- Cost per invoice.
- Average cycle time.
- Exception rate.
- Duplicate payment rate.
- Early payment discount capture rate.
Three months of baseline data is ideal. Even one month is better than nothing.
Mistake 7: Choosing the Wrong Solution
The mistake: Selecting an AP automation solution that is not designed for Oracle Fusion Cloud, requiring complex integration, middleware, and workarounds.
Why it happens: Organisations evaluate solutions based on feature lists without considering integration depth. Generic AP automation tools may offer impressive demos but lack native Oracle Fusion integration.
The consequence: Integration issues consume implementation time and budget. Data synchronisation problems create reconciliation headaches. The solution sits beside Oracle Fusion rather than inside it, creating a fragmented experience for AP staff.
How to avoid it: Choose a solution purpose-built for Oracle Fusion Cloud. AP Automation from SPC3 is designed specifically for the Oracle Fusion Cloud platform, using native APIs, referencing Oracle master data, and operating within Oracle Fusion's security model.
Mistake 8: Declaring Victory Too Early
The mistake: Celebrating the go-live date as the finish line and disbanding the project team.
Why it happens: Projects have defined start and end dates. Go-live feels like the natural conclusion.
The consequence: Without ongoing optimisation, automation performance plateaus. New suppliers are not properly onboarded. Tolerance settings are not adjusted. Exception root causes are not investigated. Over time, the auto-match rate degrades and manual effort creeps back in.
How to avoid it: Plan for a 3-6 month optimisation phase after go-live. Assign ongoing responsibility for:
- Monthly metrics review and reporting.
- Tolerance and rule optimisation.
- Supplier exception root cause analysis.
- New supplier onboarding into the automated process.
- Coordination with procurement and receiving teams on upstream data quality.
Learn from Others' Mistakes
Every mistake on this list has been made by real organisations. The good news is that they are all avoidable with proper planning, experienced guidance, and a realistic approach.
Sharpe Project Consulting has helped organisations across Australia navigate these challenges successfully. Our consulting and implementation services include the process optimisation, change management, and ongoing support that turn AP automation projects into AP automation successes.
Get in touch with the SPC3 team to discuss your AP automation plans and learn how we can help you avoid these common pitfalls.