Understanding complexity in manufacturing procurement

The invisible network: Tier 2 and Tier 3 suppliers

Modern supply chains are extremely geographically dispersed. They span multiple continents and involve a large number of players at each stage.

‍

‍

While Tier 1 suppliers are often well known and have solid contracts, the second and third tiers (Tier 2 and Tier 3) are often unclear. Info about delivery routes, risks, or critical dependencies often gets lost here, which can have serious consequences in a crisis.

Even minor errors in traceability can quickly lead to:

• Delivery delays
• Quality problems
• Production downtime

The result: delivery failures or production stoppages, even though the Tier 1 suppliers are still formally able to deliver.

Transparency does not end at Tier 1. If you want to make your supply chain resilient, you also need to keep an eye on the second and third tiers and rely on strategic supplier management.

This is precisely where artificial intelligence offers enormous potential.

AI can connect critical data from different sources. It reveals important connections and helps purchasing teams master complexity.

Particularly relevant industries

• Automotive
• Mechanical
• Medical
• Electronics

These are industries in which production reliability, product diversity, and regulatory requirements are particularly high.

Why parts lists in industry often contain hundreds of items

In the manufacturing industry, purchasing is no longer an isolated function. It is closely linked to product development, production, and supply chain strategy.

This becomes particularly clear when managing bills of materials (BOMs). These lists often contain hundreds to thousands of individual items:

• Standard parts
• Special components
• Materials that need to be certified

Each item can be linked to:

• Own material codes
• Technical drawings
• Certifications
• Your own suppliers

‍

‍

This not only leads to a lot of administrative work, but also to a high risk – especially when changes or failures affect individual parts.

This results in the following problems:

         1. Slow processes for complex BOMs:

• Manual tenders for each individual component slow down time-to-procure.
• The effort involved in comparing and selecting hundreds of components is extremely high.

2. Too little competition per item:

• BOM parts often go to known suppliers without any real price comparison.
• Savings potential remains untapped because there is no broad tendering process.

3. Valuable time instead of quality work:

• Purchasers spend too much time on operational processing (matching, etc.).
• Less room for strategic activities such as supplier development or quality assurance.

The consequences of a lack of transparency in BOM structures

Bills of materials are the backbone of every industrial production process. However, a lack of transparency in these structures poses considerable risks, particularly in terms of traceability.

When time-to-procure becomes a competitive disadvantage

Manual tendering processes for complex BOMs with hundreds of components lead to dramatically longer procurement times. While purchasers tender for each item individually and laboriously compare offers, the clock of market dynamics continues to tick relentlessly. The consequences are measurable:

• Delayed market launch of new products because procurement becomes a bottleneck. Competitors are faster to market
• Production stops occur when critical components are not available on time, even though the demand was known long ago
• Opportunity costs due to lost business while the competition is already delivering

Wasted savings potential due to lack of competition

When BOM items are routinely awarded to known suppliers without organizing a genuine price comparison, significant savings potential remains untapped. The reason is often simple: with hundreds of components, a broad market analysis seems too time-consuming.

• Price opacity leads to excessive procurement costs because alternative suppliers are not even identified
• Dependencies on individual suppliers creep in and become strategic risks
• Negotiating power remains unused because volume effects are not recognized or bundled.

Loss of strategy due to operational overload

While purchasers spend their time manually processing commodities and matching offers, there is no capacity for value-adding strategic procurement tasks. The result: a department that should be laying the foundation for competitive advantages becomes a mere processing unit.

• Supplier development falls by the wayside because time is spent on operational activities.
• Quality assurance is neglected, even though it is crucial for product quality
• Risk management is not proactive, but only reactive when acute problems arise

The result: a strategic success factor becomes an operational cost driver—and value creation remains far below its potential. These challenges require new approaches.

How AI helps with BOM and supplier network management

The complexity of global supply chains and multi-level bill of material structures requires new technological approaches. AI offers three key levers:

Data integration & automation: The technological lever

AI connects fragmented data sources such as ERP, PLM, or external supplier databases. This reduces manual data entry, increases data quality, and enables end-to-end transparency for the first time – down to Tier 3. This creates the basis for informed decisions across the entire network.

Multi-tier visibility through graph databases & NLP

Graph databases make supplier networks visually comprehensible, including all direct and indirect relationships. In combination with natural language processing (NLP), even unstructured information from PDFs, emails, or websites can be evaluated. This creates new visibility for previously hidden dependencies.

Predictive insights instead of reactive crisis management

AI recognizes early indicators of impending disruptions, such as delayed shipments, negative risk assessments, or geopolitical changes. Simulations can be used to run through alternative scenarios and generate concrete recommendations for action, such as automatically suggesting alternative suppliers.

Four steps to an AI-supported purchasing strategy

The introduction of AI in purchasing does not begin with technology, but with structure. If you want to use AI successfully, you should start with a clear, step-by-step approach:

‍

Step 1: Analyze data availability and bill of materials quality

First, you need clarity:
What data is actually available? In which systems? How is it structured?

This step identifies data gaps, redundancies, and inconsistencies. It lays the foundation for everything that follows. Without a reliable database, any AI initiative will remain inefficient.

Step 2: Launch pilots for critical assemblies

Don't start everything at once, but rather start where the leverage is greatest.
Select a product group with high complexity or high risk and integrate initial AI modules, e.g., for supplier evaluation or early risk detection.

This creates a realistic use case with measurable impact.

Step 3: Integrate into existing systems

Connectivity is essential for scaling the pilot.
Connection to ERP, PLM, or SCM systems such as Mercanis is essential. Clear interfaces must be defined and access rights neatly regulated.

The goal is smooth system integration – no additional data silos.

Step 4: Continuous optimization & training

AI does not learn on its own; AI needs feedback.
Regular evaluation of the results and targeted manual feedback continuously improve the algorithm.

At the same time, the application can be extended to other product groups, plants, or regions.

Conclusion: AI as the key to resilient manufacturing supply chains

Artificial intelligence is not a panacea, but it is a powerful tool when used in a targeted manner. It does not replace a sound purchasing strategy, but it makes its implementation possible in an increasingly complex world.

Through the intelligent use of AI, companies can:

• Identify critical dependencies in bills of materials and supplier networks
• Identify risks early on instead of just reacting
• and gain new scope for action – in real time.

In volatile markets in particular, the following applies:
Those who invest in transparency and data expertise today will not only be more stable tomorrow, but will also have a clear advantage.

FAQ