ERP selection is one of the highest-stakes technology decisions most organisations make. It shapes supply chain, procurement, finance, and operations for a decade or more. It consumes significant implementation budget — typically $1–10 million for Australian mid-market businesses, substantially more for large enterprises. And the cost of getting it wrong — years of operating on a poorly fitted system, eventually culminating in a costly rip-and-replace — is dramatically higher than the cost of the initial selection process.

Yet most organisations approach ERP selection reactively, under-resourced, and without the operational expertise to evaluate supply chain capability properly.

This article covers what a rigorous ERP selection process looks like for supply chain-intensive Australian organisations — and the most common mistakes to avoid.

Why ERP Selections Go Wrong

Before covering what to do, it's worth understanding the failure modes that are most common.

Buying brand, not fit. The major ERP vendors — SAP, Oracle, Microsoft Dynamics 365, NetSuite — all have strong brand recognition. Organisations often select a vendor based on brand confidence, peer reference, or analyst positioning rather than on a rigorous assessment of fit against their specific supply chain requirements. Brand is not irrelevant — vendor financial stability, ecosystem depth, and implementation partner availability all matter — but it is not a substitute for functional fit analysis.

Evaluating features, not use cases. ERP vendor demonstrations are designed to showcase capability. The demonstration shows the system doing impressive things — but not necessarily the specific things your supply chain needs to do, in your specific operational context, with your specific data. Selecting a system based on a vendor-led demonstration without testing it against your own use cases is the most common source of post-selection regret.

Under-investing in requirements definition. The quality of an ERP selection is directly proportional to the quality of the requirements that drive it. Organisations that invest three to four weeks in rigorous requirements definition — engaging supply chain, procurement, operations, and finance — select better and implement better. Organisations that shortcut this step typically find gaps post-selection that are expensive to close.

Underestimating implementation cost and complexity. ERP vendors quote software licences. Implementation costs — the consulting and integration work to configure, customise, migrate data, integrate systems, train users, and manage change — are typically three to five times the licence cost. Organisations that select a system based on licence cost alone without a realistic total cost of ownership model make commercially poor decisions.

Ignoring supply chain fit. Many ERP selections are led by finance or IT, with supply chain as a secondary stakeholder. This is backwards for supply chain-intensive businesses. The supply chain module of an ERP — demand planning, inventory management, procurement, warehouse management, manufacturing — is typically where the most process-critical functionality lives, and where the gap between different vendors is most material.

The Eight Requirements That Matter for Supply Chain

Not all ERP supply chain requirements are equally important. The following eight are the most critical to evaluate rigorously.

Inventory management. How does the system manage multi-location inventory, serialisation and batch tracking, FIFO/FEFO/LIFO costing, and inventory adjustments? For businesses with complex inventory — multiple warehouses, expiry dating, product traceability requirements — the depth of inventory management capability is a primary selection criterion.

Demand planning and forecasting. Does the system have native demand planning capability, or does it require a best-of-breed planning tool? What statistical methods does it support? How does it handle seasonal patterns, promotions, and new product introductions? For businesses where forecast accuracy has a material impact on inventory and procurement decisions, the quality of planning functionality is critical.

Procurement and purchase order management. How does the system manage the procure-to-pay process — from requisition through supplier management, purchase ordering, goods receipt, and invoice matching? Does it support three-way matching? How does it handle blanket orders, price books, and supplier catalogues?

Warehouse management. Does the system have a native WMS module, or does it integrate with third-party WMS solutions? For businesses with complex warehousing operations — multi-bin, pick-and-pack, cross-docking, returns management — the depth of WMS functionality (or the quality of integration to a best-of-breed WMS) is important.

Manufacturing and production planning. For manufacturers, MRP (Material Requirements Planning) and production scheduling capability is a core requirement. The quality of MRP logic — how it handles multi-level bills of materials, capacity constraints, lead time variability — varies significantly between ERP vendors and even between versions of the same vendor's product.

Integration capability. ERPs rarely operate in isolation. Integration with 3PL systems, carrier platforms, e-commerce platforms, supplier portals, and specialist planning tools is standard. The quality of the ERP's integration framework — API capability, pre-built connectors, data exchange standards — significantly affects implementation cost and operational reliability.

Reporting and analytics. What reporting and analytics capability does the system provide natively? What data model does it expose for external analytics tools? For supply chain teams that depend on data for performance management and decision-making, the reporting capability — or the cost and complexity of building it — is a material evaluation criterion.

Scalability and configurability. Will the system scale with the business as it grows? Can it be configured to match the organisation's operating model without expensive custom development? How does the vendor manage upgrades — will customisations need to be rebuilt with each major release?

The Selection Process

A rigorous ERP selection for a supply chain-intensive business follows five phases.

Requirements definition (4–6 weeks). Document current state processes, identify pain points and capability gaps in the current system, define future state requirements by process area, and develop a weighted requirements matrix that scores requirements by business criticality. Engage all major stakeholders — supply chain, procurement, operations, finance, IT.

Vendor longlist and RFI (2–3 weeks). Develop a longlist of vendors appropriate to the organisation's scale, sector, and requirements. Issue a Request for Information to the longlist — typically 6–10 vendors — to gather basic capability and commercial information. Score RFI responses against the requirements matrix and develop a shortlist of 3–4 vendors for detailed evaluation.

Detailed evaluation (6–8 weeks). Issue a detailed RFP to the shortlist. Conduct scripted demonstrations — vendor demonstrations conducted against your specific use cases, with your own sample data, rather than vendor-scripted showcases. Reference check with comparable Australian deployments. Conduct an implementation partner assessment alongside the software assessment.

Commercial negotiation (4–6 weeks). Negotiate licence terms, implementation scope, support terms, and total cost of ownership with the preferred vendor. Clarify contract terms — particularly around customisation, upgrade paths, and exit rights — before committing.

Business case and decision (2–3 weeks). Develop a total cost of ownership model across 5–7 years (licence, implementation, support, internal resource, opportunity cost of transition) and a benefits case (efficiency gains, capability improvements, cost savings). Present to the decision-making committee with a clear recommendation and risk assessment.

Implementation Partner Matters as Much as Software

In ERP selection, the implementation partner selection is as important as the software selection. Most ERP implementation problems are not software problems — they are implementation problems. A capable, experienced implementation partner with deep supply chain expertise and a track record of successful deployments in comparable Australian businesses is the most important risk mitigation in an ERP programme.

The major ERP vendors have large implementation partner ecosystems. The quality across these ecosystems varies significantly. Evaluating the specific partner — not just the vendor relationship — including team credentials, supply chain methodology, Australian references, and senior resource commitment, is essential.

How Trace Consultants Can Help

Trace Consultants provides independent ERP selection support for Australian supply chain-intensive organisations — helping define requirements, run vendor selection processes, evaluate supply chain functionality, and build business cases that support well-informed technology decisions.

Requirements definition: We facilitate structured requirements workshops with supply chain, procurement, operations, and finance stakeholders — producing a requirements document that drives a rigorous selection process.

Vendor selection management: We manage the RFI/RFP process, scripted demonstration design, reference checking, and commercial evaluation — providing independent assessment and recommendation.

Business case development: We build total cost of ownership models and benefits cases that give leadership teams the financial framework to make technology investment decisions confidently.

Explore our Technology services →Speak to an expert at Trace →

‍

Artificial intelligence has become the defining topic in supply chain technology conversations in Australia. Every planning software vendor has an AI story. Every major consulting firm has an AI-in-supply-chain white paper. Procurement platforms are embedding AI assistants. Warehouse automation providers are claiming AI-powered everything.

Some of it is transformative. A significant portion of it is rebranded analytics, statistical modelling that has existed for decades, or genuine capability that is so far from production-ready deployment in Australian mid-market businesses that it belongs in a five-year horizon, not a current investment decision.

This article cuts through the noise. What is AI in supply chain actually delivering today? Where is it genuinely transforming operations? And what should Australian businesses do about it?

What AI in Supply Chain Actually Means

"AI" in supply chain covers a range of different technologies that are meaningfully different in their maturity, cost, and applicability.

Machine learning for demand forecasting. This is the most mature and widely deployed AI application in supply chain. Traditional demand forecasting uses statistical methods — moving averages, exponential smoothing, regression — that identify patterns in historical sales data. ML-based forecasting uses more complex models that can incorporate a much wider range of signals — weather, events, pricing, social media sentiment, economic indicators — and update forecasts more frequently in response to real-time signals. For businesses with complex, high-SKU demand patterns and rich data, ML forecasting delivers meaningful accuracy improvements. The commercially available implementations (Blue Yonder, Kinaxis, o9, SAP IBP) are mature and deployable today.

Natural language processing in procurement. NLP-powered tools are being applied to procurement document processing — contract analysis, invoice matching, spend categorisation — and to market intelligence gathering. These applications are increasingly production-ready. AI-assisted contract review (flagging non-standard clauses, extracting key terms, benchmarking against standards) is being deployed by large Australian procurement functions. AI spend categorisation tools are improving the quality and speed of spend analysis.

Computer vision in warehousing. Camera-based AI systems for quality inspection, inventory counting, pick accuracy verification, and safety monitoring are being deployed in Australian distribution centres. These applications are technically mature in specific use cases — automated quality inspection on production lines, for example — but deployment in general warehousing is less widespread.

Generative AI for knowledge work. Large language models (LLMs) are beginning to change how supply chain and procurement professionals do knowledge work — drafting RFPs, analysing supplier contracts, synthesising market intelligence, generating operational reports. This is the fastest-moving area of AI application. The tools are available now (through enterprise platforms from Microsoft, Google, and Salesforce, and through purpose-built procurement and supply chain applications), but organisational readiness to use them effectively varies enormously.

Autonomous planning and decision-making. AI systems that autonomously execute supply chain decisions — placing purchase orders, rebalancing inventory, rerouting shipments — without human approval are technically possible in constrained domains but are at early deployment stages in most Australian businesses. The exception is highly automated, repetitive domains like VMI (vendor-managed inventory) with trusted suppliers, where automated replenishment triggered by AI-assessed demand signals is already running in some large retailers and FMCG businesses.

Where AI Is Delivering Real Value Today

Demand forecasting accuracy. The most consistent and well-documented AI value in supply chain is improved forecast accuracy from ML models. In high-SKU environments with complex demand patterns, ML-based forecasting consistently outperforms traditional statistical methods — with accuracy improvements of 10–25% in MAPE (Mean Absolute Percentage Error) documented across multiple implementations. For businesses where inventory, production, and procurement decisions are driven by demand forecasts, each percentage point of forecast accuracy improvement has a direct bottom-line impact.

Spend analytics and categorisation. AI-powered spend categorisation tools are transforming what used to be a labour-intensive, error-prone process into an automated one. Unstructured procurement transaction data — purchase orders from multiple systems, against multiple suppliers, with inconsistent descriptions — can now be cleaned, categorised, and enriched automatically. This enables procurement teams to see their spend clearly and act on it, rather than spending weeks preparing data before any analysis can begin.

Contract intelligence. For large procurement teams managing complex supplier contract portfolios, AI contract analysis tools are delivering genuine efficiency gains. The ability to extract key terms, flag non-standard clauses, and alert on approaching renewal or expiry dates across hundreds of contracts — without manual review — is transforming contract lifecycle management in large Australian organisations.

Predictive maintenance in operations. Manufacturers and logistics operators with sensor-equipped assets are using ML-based predictive maintenance to reduce unplanned downtime. This is technically mature and well-evidenced — the value case is strong wherever unplanned downtime has significant operational cost.

Route and load optimisation. AI-enhanced transport and route optimisation — incorporating real-time traffic, weather, and vehicle availability — is delivering meaningful freight cost reductions for Australian logistics operators and businesses managing their own fleets.

Where the Hype Outstrips Reality

"AI-powered" planning tools that are still statistical forecasting. Many planning software vendors have rebranded existing statistical forecasting models as "AI" or "ML." A time-series model with an exponential smoothing algorithm is not machine learning in any meaningful sense. Buyers should ask vendors specifically what algorithm is in use, what training data it requires, and what accuracy improvement is documented against a statistical baseline in comparable businesses.

Autonomous supply chain decision-making at scale. The vision of an AI system that autonomously manages end-to-end supply chain decisions — procurement, inventory, logistics — without human involvement is technically distant from production-ready deployment in most businesses. The data infrastructure, process standardisation, and organisational trust required to operate autonomously at scale don't yet exist in most Australian supply chains.

Generative AI replacing supply chain professionals. LLMs are genuinely changing knowledge work in supply chain and procurement — but the productivity impact is an amplification of human expertise, not a replacement of it. The professionals who understand supply chain deeply and use AI tools effectively will produce better work faster. The professionals who don't will be at a disadvantage. Neither group is being replaced by the technology.

Plug-and-play AI implementations. AI tools require data. Clean, consistent, well-structured data from source systems that are integrated and maintained. Most Australian mid-market businesses don't have this infrastructure in place — which means an AI implementation is frequently preceded by a data infrastructure project that is larger, slower, and more expensive than the AI implementation itself.

What Australian Businesses Should Do Now

Invest in data quality first. The ROI on AI tools is directly proportional to the quality of the data they run on. Organisations that invest in improving data quality in their ERP and operational systems are building the foundation for AI value — regardless of which specific tools they ultimately deploy.

Prioritise high-value, mature applications. Demand forecasting accuracy, spend analytics, and contract intelligence are mature, well-documented AI applications that are deployable in Australian businesses today. Start there before pursuing frontier applications.

Pilot before scaling. AI tools should be piloted in a constrained domain before enterprise rollout. A demand forecasting pilot in one product category, with a clear accuracy benchmark against the current method, provides real evidence of value before committing to a broader implementation.

Build internal capability. The supply chain and procurement functions that are getting the most from AI are the ones investing in the capability of their own people — data literacy, analytical skills, ability to interrogate and challenge AI outputs. Technology without capability investment delivers technology costs, not technology value.

How Trace Consultants Can Help

Trace Consultants helps Australian organisations navigate supply chain technology decisions — including AI — with an evidence-based approach that separates genuine capability from marketing.

AI readiness assessment: We assess your data infrastructure, process maturity, and organisational capability against the requirements for effective AI deployment — and identify where the foundations need strengthening before technology investment.

Technology selection: We run structured technology selection processes that evaluate AI-enabled supply chain tools against your specific requirements — and hold vendors accountable for documented, benchmarked performance claims.

Implementation support: We provide programme management and change management support for technology deployments, ensuring AI tools are embedded in daily operating processes rather than sitting unused after go-live.

Explore our Technology services →Explore our Planning & Operations services →
‍Speak to an expert at Trace →

"Supply chain control tower" has become one of the most used — and most misused — terms in supply chain technology. Vendors apply it to everything from basic dashboards to sophisticated AI-powered orchestration platforms. Consultants recommend them in almost every technology strategy engagement. Executives ask about them because they've read about them in industry publications.

The concept is genuinely valuable. But the gap between what a control tower is marketed as and what it actually delivers in practice — particularly in Australian mid-market businesses — is significant. This article cuts through the noise.

What a Supply Chain Control Tower Is

At its core, a supply chain control tower is a centralised system that provides real-time or near-real-time visibility across supply chain operations — connecting data from multiple sources (suppliers, logistics providers, warehouses, customers) into a single view, identifying exceptions and disruptions, and (in more sophisticated implementations) generating alerts or recommendations that enable faster decision-making.

The term draws on the air traffic control analogy: a tower that can see all the aircraft in its airspace, monitor for conflicts and anomalies, and coordinate responses in real time.

In practice, control tower implementations range considerably in maturity and sophistication.

Visibility-only control towers aggregate data from multiple systems — ERP, WMS, TMS, supplier portals — into a dashboard that gives supply chain operators a single view of inventory, orders, and shipments. They show what is happening, but they don't predict what will happen or recommend what to do about it.

Event-driven control towers add exception management capabilities: the system identifies deviations from plan (a shipment that is late, a supplier that has missed a confirmation, an inventory level that has breached a minimum threshold) and triggers alerts to the relevant operators, often with workflow tools to manage the resolution process.

Predictive control towers use machine learning and advanced analytics to predict disruptions before they occur — identifying a supplier at financial risk, forecasting a stockout based on demand trajectory and inbound supply, or calculating the downstream impact of a freight delay on customer delivery commitments. These systems require significant data maturity and integration depth to function effectively.

Prescriptive or autonomous control towers — the frontier of the category — go beyond prediction to recommendation or autonomous action. They suggest or execute the best response to a disruption: rerouting a shipment, rebalancing inventory between distribution centres, or triggering an emergency purchase order. In most Australian businesses, this level of automation is not yet deployed at scale.

What Control Towers Are Not

There are several common misconceptions worth clearing up.

A control tower is not a supply chain strategy. It is a technology that supports the execution of a supply chain strategy — but it doesn't replace the thinking about what the supply chain should be doing. Organisations that implement a control tower without clear operating processes and decision rights get an expensive dashboard that nobody acts on.

A control tower is not an ERP. It sits alongside the ERP and draws data from it, but it doesn't replace the system of record for transactions. The relationship between a control tower and the organisation's ERP, WMS, and TMS needs to be clearly designed — which system is the source of truth for what type of data, and how do they interact?

A control tower is not a magic solution to poor data quality. Control towers aggregate data from multiple source systems — and if those source systems have incomplete, inaccurate, or inconsistent data, the control tower will surface and amplify those data problems, not solve them. Organisations with immature underlying data often find that a control tower implementation is primarily a data quality improvement project in disguise.

The Business Case

The value of a supply chain control tower is captured in four areas.

Reduced disruption impact. Faster identification of disruptions — supplier delays, logistics failures, demand spikes — allows faster response, reducing the cost and service impact of events that would previously have gone undetected until they became crises. In supply chains with complex international sourcing, where lead times are long and disruption events are frequent, this has demonstrable financial value.

Lower operational overhead. Manual monitoring of supply chain performance — chasing supplier confirmations, tracking inbound shipments, managing exception reports — consumes significant planner and analyst time. A control tower that automates exception detection and alerts frees that time for higher-value activity.

Improved customer service. Earlier visibility into supply disruptions allows customer service teams to proactively manage customer expectations — communicating delays before they materialise rather than after. This is a competitive differentiator in sectors where reliability is a key purchase criterion.

Better inventory management. Supply chain visibility enables tighter inventory management — reducing buffer stock that exists because of uncertainty rather than genuine demand variability, and reducing the frequency and cost of expediting activity triggered by stockouts.

Quantifying these benefits in the Australian context: organisations with mature supply chain visibility capabilities typically report 10–20% reductions in stock-out frequency, 15–25% reductions in expediting costs, and meaningful reductions in the management overhead associated with exception handling.

When You Need One — and When You Don't

Control towers deliver the most value in specific supply chain contexts.

You probably benefit from a control tower if: your supply chain involves multiple suppliers across different geographies, with lead times of several weeks or more; you experience frequent supply disruptions that are currently identified late and managed reactively; you have multiple distribution nodes that need to be coordinated in real time; and your planning team is spending significant time on manual tracking and exception management rather than on analysis and decision-making.

You probably don't need a control tower if: your supply chain is simple — a small number of domestic suppliers, one or two distribution points, and a limited SKU range; your current ERP and operational systems already provide adequate visibility for your planning team's needs; your primary supply chain challenge is process discipline rather than visibility; or your organisation doesn't have the data infrastructure to feed a control tower with reliable, consistent data.

For many Australian mid-market businesses — particularly those under $500 million in revenue with relatively straightforward supply chains — the right investment is better use of existing ERP visibility capabilities, improved exception reporting within current systems, and more disciplined operating processes — not a six-figure control tower implementation.

Selecting and Implementing a Control Tower

For organisations where the business case is clear, the implementation approach matters significantly.

Define the use cases first. What specific decisions do you want the control tower to support? Which exceptions do you most need to see faster? Which supply chain processes currently involve too much manual monitoring? Answering these questions before evaluating technology prevents the common failure mode of selecting a technology and then trying to find a use case.

Assess data readiness. Map the data sources the control tower will need to consume and assess their quality and accessibility. Plan for data integration investment as a significant component of total implementation cost — it typically represents 30–50% of the total project.

Start narrow and expand. A phased implementation starting with one supply chain segment or one data domain is significantly more likely to deliver value than a broad implementation that attempts to connect everything simultaneously.

Invest in adoption. A control tower that is technically implemented but not embedded in daily operating processes delivers no value. Change management — training, process redesign, performance management integration — is as important as the technology itself.

How Trace Consultants Can Help

Trace Consultants helps Australian organisations assess, select, and implement supply chain technology — including control tower solutions — in a way that is grounded in operational reality rather than vendor marketing.

Technology needs assessment: We assess whether a control tower is genuinely the right investment for your supply chain context, and if so, what capabilities and data infrastructure are prerequisites.

Vendor selection: We run structured selection processes across the control tower vendor landscape — including platforms from major ERP vendors (SAP IBP, Oracle) and specialist providers — to find the right fit for your requirements and budget.

Implementation support: We provide programme management and change management support for technology implementations, ensuring the business case is realised in practice.

Explore our Technology services →Explore our Strategy & Network Design services →Speak to an expert at Trace →

‍