The Complete Guide to Enterprise Invoice Processing in 2023

Processing invoices is one of the biggest operational headaches for large enterprises today. According to research, the average enterprise receives over 250,000 invoices per year.1 With the average cost to manually process a single invoice estimated at $15 to $40, the numbers add up quickly.2 For an enterprise processing 250,000 invoices per year, that could mean $3.75M to $10M spent annually just on labor for invoice processing!

Beyond the hard costs, poor invoice management causes all kinds of business problems – late payments leading to poor supplier relations, accounting errors, missed discounts, and insufficient visibility into spend. As one AP manager put it, "Invoice processing is the bane of any large enterprise‘s existence."

The good news is that new technologies like artificial intelligence, machine learning, and intelligent document processing are revolutionizing enterprise invoice processing. They enable a level of automation that was not possible even a few years ago. Adoption is growing rapidly – the accounts payable automation market is predicted to reach $1.9B by 2025.3

However,automated enterprise invoice processing comes with its own set of challenges. In this comprehensive guide, we‘ll cover:

  • Key guidelines for enterprise invoice processing success
  • Leading technology solutions for AP automation
  • Best practices for implementation based on real-world experience

I‘ll approach each topic not as a technology vendor pitching products, but as an independent advisor sharing hard-earned lessons from numerous enterprise automation initiatives. My goal is to provide you with an unbiased perspective to help your organization modernize invoice processing. Let‘s dive in!

Why Enterprise Invoice Processing Needs to Change

Before exploring solutions, it‘s important to understand why change is needed. What exactly is broken with current invoice processes?

Extreme Volume

As mentioned earlier, large enterprises often receive hundreds of thousands of invoices annually, from purchase orders to one-time suppliers and everything in between. Manually processing this volume is labor-intensive, expensive, and prone to errors. Even large shared service centers struggle with these volumes.

Here‘s a snapshot of the scale for a sample enterprise with $15B in annual revenue:

  • 250,000 invoices per year
  • 15,000 invoices per month
  • 3,000 invoices per week
  • 600 invoices per day

And that‘s just averages – invoice volume can spike at month, quarter or year-end closing periods.

Variety of Invoice Types and Sources

Not only are invoice volumes high, but the variety adds complexity. Invoices can arrive via email, EDI, supplier portals or mail. They may have a corresponding purchase order or none at all.

Suppliers range from large strategic vendors to small one-time suppliers. Receiving quality invoices in a structured format from every supplier is unrealistic.

This variety makes it challenging to rely on simple rules-based automation. Advanced intelligence is needed to handle diverse invoice formats and sources.

Accounting Complexity

Behind the scenes, enterprises have intricate accounting requirements that make invoice processing highly complex.

There may be thousands of general ledger codes to which invoices must be accurately posted. Different business entities, divisions, locations and projects create a complex hierarchy for reporting.

With strict financial reporting requirements, a simple miscoding of an expense type or business entity can have major implications.

Limited Visibility

With such high volumes, accounting complexity and variety of sources, enterprises often have very limited visibility into the status of invoices flowing through the system. AP managers don‘t know exactly how many invoices are stuck in exceptions, errors, or bottlenecks.

This makes it incredibly difficult to identify and address processing delays and pain points. Lack of visibility also reduces financial control and forecasting accuracy.

The High Costs of Inefficiency

Ultimately, the volume, variety, complexity and lack of visibility all drive massive inefficiency into enterprise invoice processing. AP teams are stuck in a mode of fighting fires versus optimizing a streamlined process.

The costs of this inefficiency are staggering, including:

  • Higher processing costs: Repeated manual handling drives up per invoice costs.
  • Payment delays: Bottlenecks lead to late payments, damaging supplier relations.
  • Accounting errors: Misallocations and duplications distort financial reporting.
  • Missed discounts: Delays in approvals cause missed discount opportunities.
  • Insufficient control: Poor visibility impedes the ability to optimize and improve processes.

The solution is clear – enterprises must find ways to radically improve the efficiency of their invoice processes. Doing so will significantly reduce costs while improving relations with suppliers, financial control, and accounting accuracy.

4 Guidelines for Enterprise Invoice Processing

Given the challenges outlined above, how should enterprises approach invoice processing? Here are four key guidelines:

1. Require Intelligent Automation for All Invoice Types

The first step is accepting that rules-based automation is not sufficient for enterprise invoice volumes and variety. Instead, advanced artificial intelligence and machine learning are needed.

Unlike rules-based systems, intelligent automation tools can handle the diversity of invoice types and exceptions at enterprise scale:

Purchase Order Invoices

  • Match PO details to invoice line items
  • Resolve price/quantity discrepancies

Non-Purchase Order Invoices

  • Digitally extract unstructured data locked in scanned images and PDFs
  • Classify suppliers and spending categories without PO data

Foreign Language Invoices

  • Leverage multilanguage machine learning models
  • Consistent automation rates globally

The key is having integrated intelligent automation across the entire process – from extracting invoice data all the way through integration into accounting systems. This provides the end-to-end automation required to eliminate manual steps.

2. Centralize Invoice Data and Systems

To achieve efficiency at scale, enterprises should centralize storage of invoice data in an enterprise data lake, instead of siloed systems across business units and geographic regions.

Consolidating data provides a "single source of truth" enabling real-time visibility into invoices enterprise-wide using analytics and dashboards. It also improves Machine Learning model accuracy by training on more data.

Centralizing also allows for a shared services model for processing invoices, concentrating specialized expertise in one location. According to McKinsey, centralizing procurement can reduce process costs 15-25%.4

3. Optimize Workflows for Productivity

In addition to centralizing data and systems, workflows should be analyzed and optimized:

  • Eliminate redundant steps: Remove unnecessary approvals or status changes that complicate workflows.
  • Prioritize touchless processing: Configure system to automatically approve/post compliant invoices without human intervention.
  • Expedite exceptions: Flag exceptions early for rapid resolution to prevent downstream delays.
  • Configure smart rules: Set approval rules based on invoice attributes like amount, supplier, GL code.

Optimized workflows eliminate bottlenecks and speed throughput for maximum productivity. They balance automation with smart, risk-based approval rules requiring human review only when truly necessary.

4. Take a Phased, Rapid Implementation Approach

Don‘t attempt to overhaul enterprise invoice processing overnight. This high-risk "boil the ocean" approach is prone to fail. Instead, take an iterative, phased approach focusing first on automation of high-impact areas.

For example, early phases may focus on automating high-volume purchase order invoice processing. Later phases tackle different invoice types or business units.

The key is to show quick wins and incremental value. Starting small also allows time to adopt change management best practices covered later.

Leading Invoice Processing Solutions

Now that we‘ve covered key guidelines, let‘s examine leading solutions for enterprise invoice automation:

Specialized AI-Powered Accounts Payable Automation Platforms

Purpose-built accounts payable automation platforms take an integrated, end-to-end approach powered by artificial intelligence and machine learning algorithms. Key capabilities:

Intelligent Document Processing

  • Advanced OCR and machine learning extract unstructured invoice data locked in images and PDFs

3-Way Invoice Matching

  • Match invoice line items against POs down to the penny

Invoice Workflow

  • Configure approval workflows and business rules for automated routing

Accounting Integration

  • Seamlessly integrate with ERPs like SAP and Oracle for booking

Procure-to-Pay Integration

  • Connect with procurement systems for seamless P2P automation

Analytics

  • embedded analytics provide real-time invoice processing visibility

Leading providers in this space include Rossum, Hypatos, and Medius.

RossumHypatosMedius
Key StrengthsAdvanced AI, rapid implementationself-configuring ML, easy integrationend-to-end P2P suite, global delivery model
Ideal CustomerMid-market and enterpriseEnterpriseLarge multi-national enterprises
PricingAnnual subscription, volume-basedAnnual subscription, volume-basedAnnual subscription, user-based

When It Works Best

Integrated AP automation platforms excel at providing rapid, scalable automation across the entire invoice process. They are ideal for enterprise-wide initiatives.

Point Solutions for AP Automation

Enterprises often combine specialized point solutions:

Intelligent Document Processing (IDP)

IDP software focuses specifically on digitally extracting unstructured invoice data locked in scanned documents and PDFs using advanced OCR and machine learning algorithms.

Top solutions like Rossum, Hypatos, and IBM Datacap provide pre-built templates to identify and capture key fields from invoices of any format. Some integrate directly with Procure-to-Pay suites.

When invoices arrive primarily as unstructured image files, IDP should be part of the automation stack.

Buyer-Supplier E-Invoicing Networks

E-invoicing networks like Taulia, Tradeshift and Basware provide a digital platform enabling buyers and suppliers to exchange structured e-invoice data. This eliminates document processing needs.

However, they require supplier enrollment and adoption. Large global enterprises may struggle getting small suppliers onto new technology platforms.

E-invoicing networks make sense for mid-market companies or enterprises with highly sophisticated suppliers.

Generic Automation Tools

Sometimes enterprises try using generic automation tools for invoice processing:

Robotic Process Automation (RPA)

RPA tools like UiPath and Automation Anywhere are designed for automating repetitive, rules-based processes. Historically RPA has lacked intelligence for unstructured data. RPA bots can help but they still require manual configuration.

Integration Platforms

Cloud integration platforms like Mulesoft, Workato, and Microsoft Power Automate can help connect systems. However, this just pipes data – someone still needs to configure the workflows.

Business Process Management (BPM) Tools

BPM platforms like Pega help model processes and workflows. But again, they require extensive configuration for invoice handling logic.

When They Work Best

Generic automation tools can complement specialized AI solutions by connecting surrounding systems and processes. But for core invoice processing, opt for purpose-built automation.

Implementation Best Practices

Let‘s shift gears and talk implementation. What are some real-world lessons learned from successful enterprise automation rollouts?

Secure Executive Sponsorship

Any major business transformation requires strong executive sponsorship. Invoice automation impacts people and processes across the organization. An influential executive sponsor can ensure collaboration across business units. They spearhead culture change and drive continued focus.

Take a Phased Approach

As mentioned earlier, don‘t try to boil the ocean. Maintain focus by rolling out capabilities in phases:

  • Phase 1: Automate high-volume areas like PO invoices
  • Phase 2: Expand to additional invoice types, business units, or geographies
  • Phase 3: Optimize and enhance performance. Add emerging technologies like RPA.

Phase the rollout to show incremental value vs. aiming for perfection upfront. Be flexible – the solution can evolve over time.

Engineer for Scale from the Start

Even if you start small, architect the solution to support massive volumes from day one. Here are two examples:

  • Central data pipeline: Choose technology that can handle millions of invoices for unified data ingestion/processing
  • Elastic computing: Cloud computing architecture that auto-scales to handle spikes in invoice volume

Bolting on scalability after the fact can be very difficult and expensive.

Provide Ongoing User Training

User adoption requires education. Many AP staffers may be unfamiliar working in automated environments. Provide training materials and forums to share best practices. Cycle training as new features roll out.

Foster Open Communication

Keep stakeholders informed through frequent project updates and celebrating wins. Be transparent about challenges and invite input on improving adoption. Develop supporter networks to drive engagement.

Monitor Performance Metrics Closely

Tracking KPIs helps spot issues early before they become problems. Some metrics to closely monitor include:

  • Invoice cycle time
  • Automation rates
  • Exception rates
  • Productivity per AP staffer

Perform root cause analysis on metrics trending the wrong way. Don‘t let small process breaks become major bottlenecks.

The Bottom Line

Let‘s recap the key points:

  • Due to extreme invoice volumes and complexity, manual enterprise invoice processing is painful and costly. Intelligent automation is critical.
  • Follow guidelines like centralizing data, optimizing workflows, taking a phased approach and requiring automation across all invoice types.
  • Specialized AI-powered accounts payable automation platforms provide an integrated solution. Point solutions or generic tools have limitations.
  • Rollouts require executive sponsorship, training, communication and monitoring to drive success.

With the right strategy and solutions, enterprises can transform invoice processing from a costly headache into a streamlined, efficient driver of financial health and savings. The technology now exists to make this vision a reality!

What questions do you still have? What are your biggest invoice processing pain points? I‘m happy to provide any additional details. Feel free to reach out!

Similar Posts