Fraud Blocker
Global Logistics Provider | Asia-Pacific

Automating accounts payable across a $5 billion logistics network

Client profile

Toll Group is one of Asia-Pacific's largest logistics providers, a $5 billion subsidiary of Japan Post Holdings operating across 36 countries with three major divisions.

Built through decades of acquisitions, the business runs on four separate ERP and finance systems, including a legacy platform over 27 years old, with decentralised procurement, multiple sub-entities, and finance operations spanning onshore teams in Australia and an offshore Global Business Services centre in Malaysia.

120,000+
invoices processed annually
FTEs
Reduced from 26 to 6
40%
Reduction in invoice cycle time

Business challenges

Toll's Australian business was processing over 120,000 invoices annually, many of them complex, multi-line documents across multiple languages and currencies. But the infrastructure behind that process had not kept pace with the scale of the operation.

  • Invoices were physically copied, scanned, and manually keyed into a 27-year-old finance system with no purchase order matching.
  • Spend data was scattered across multiple ERPs, access databases, and spreadsheets with no consolidated view.
  • The same supplier could appear across six or more sub-entities, each on different payment terms, with no centralised visibility.
  • Invoice queries and approvals were routed through manual email chains with no audit trail or SLA tracking.
  • Processes running outside of systems created audit risks including duplicate payments, missed payments, and unapproved transactions.

The result: 26 full-time employees dedicated to accounts payable, hundreds of hours lost weekly to manual processing, and a finance team unable to shift from transactional work to strategic, value-adding activities.

Why SpendConsole was chosen

Toll evaluated multiple AP automation vendors but found that most offered off-the-shelf solutions that didn't account for the complexity of a multi-entity, multi-currency, multi-language logistics operation. SpendConsole stood apart.

  • Invested time understanding Toll's operating model, from subcontractor invoices to government agency payments, rather than proposing a generic solution.
  • Tested over 10,000 invoices across 11–12 languages, including handwritten annotations, proving the AI extraction was genuine.
  • Successfully integrated with Toll's 27-year-old bespoke finance system
  • Delivered a single, intuitive platform with workflow logic, approval routing, and consolidated visibility, replacing a landscape of disconnected systems.
  • Provided direct access to senior leadership throughout evaluation and implementation.

The solution

SpendConsole was deployed across Toll's Australian business as Phase 1 of a broader global transformation, automating the full accounts payable lifecycle and laying the foundation for consolidated spend intelligence.

  • AI-powered invoice capture and processing across multiple languages, currencies, and document formats, including handwritten annotations.
  • Purchase order matching and three-way matching to eliminate manual invoice-to-PO verification.
  • Workflow automation and approval routing replacing manual email chains with logic-driven workflows, SLA tracking, and full audit trails.
  • Non-PO compliance tracking providing visibility into maverick spend and unapproved purchases.
  • Centralised vendor management consolidating duplicate supplier records and surfacing inconsistent payment terms across sub-entities.
  • Consolidated spend analytics bringing procurement and AP data into a single source of truth.

Implementation was structured as a phased rollout, beginning with a soft go-live on a small subset of vendors and invoices, then gradually expanding to the full Australian operation. The program was delivered cross-border, with SpendConsole working alongside Toll's Australian stakeholders, IT department, and the GBS team in Malaysia.

Results

Within weeks of go-live, Toll began realising measurable outcomes across cost, productivity, compliance, and operational visibility.

  • AP headcount reduced from 26 to 6 FTEs.
  • Invoice cycle time reduced by 30–40%, down from 23 days from receipt to ready-for-payment.
  • 50%+ straight-through processing rate achieved, with a path to 70% as supplier adoption scales.
  • 80% of non-compliant invoices automatically rejected at ingestion, a control that didn't previously exist.
  • Manual exceptions reduced to under 30% of total invoice volume through AI-based validation and workflow routing.
  • Hundreds of man-hours saved weekly by replacing email-based queries and approvals with automated workflows.
  • Consolidated spend visibility across sub-entities for the first time.
best-accounts-payable-automation-software
We truly believe SpendConsole really wanted to understand our business. With the other players in the market, it felt like more of an off-the-shelf solution. SpendConsole took the time, understood our problem statements, and really understood our requirements.
Dion Morgan
Senior Vice President, Global Finance Operations
Transform complex systems into seamless experiences

Moving forward

The Australian rollout represents Phase 1 of Toll's global AP transformation. With the foundation now in place, the next step is extending SpendConsole to Toll's Global Forwarding division, unlocking cross-border vendor analysis, FX exposure visibility, and global supplier optimisation.
From there, consolidated spend data opens the door to working capital improvements, preferred supplier compliance, and category-level benchmarking across the business.

Toll is also building a centralised supplier information hub for vendor onboarding, remittance advices, and query management across all entities, replacing fragmented processes with a single point of control.

Legacy expense tools will give way to SpendConsole's integrated expense management module, bringing every outbound payment onto one platform. And with SpendConsole's AI assistant, finance teams will be able to query spend, vendor, and compliance data using natural language.