Predictive Procurement: How asset data makes purchasing smarter

by Michael Lehnert · CEO & Co-Founder, seventhings

A facility manager orders three new projectors for the conference rooms because two colleagues complained about a lack of equipment. Two weeks later, an old projector turns up in the storage room. No one knew it was there. This is what purchasing looks like without a data foundation: reactive, on-demand, and blind to what is already in the building.

This is exactly where predictive procurement comes in. The term originates from procurement research and describes purchasing decisions based on historical usage, failure, and lifecycle data rather than gut feeling or order history alone. 73% of companies with an ERP system continue to maintain parallel Excel lists for their physical assets (Forrester, 2023) – a data foundation from which reliable purchasing decisions can hardly be derived.

Key Takeaways

• 73% of ERP users maintain parallel Excel lists for assets – a weak foundation for purchasing decisions (Forrester, 2023)
• According to McKinsey, medium-sized companies have €50,000–€200,000 worth of unused equipment in their inventory, which is often repurchased anyway (McKinsey, 2024)
• Without tracking, 12–18% of tools are lost annually – and are usually just reordered (Nexess Solutions, 2024)
• 67% of CFOs count tied-up capital among their top three concerns – unused equipment is part of that (Deloitte CFO Signals, 2024)

‍

What does predictive procurement mean for physical assets?

Predictive procurement refers to purchasing decisions based on structured historical data: usage intensity, location distribution, damage and loss rates, maintenance costs, and remaining service life. Procurement no longer asks, "What do we likely need?" but rather, "What does the history actually show?" For physical assets without IoT sensors—tools, furniture, testing equipment, conference technology—this history comes from asset management itself: handovers, location changes, reports, and disposals.

It is important to distinguish this from pure requirements planning in an ERP. An ERP system knows order quantities and lead times. However, it rarely knows if a projector has been sitting unused in a cupboard on the third floor for ten months. A structured asset history closes this gap—not as a replacement for the ERP, but as an additional data layer for decision-making before placing an order.

‍

‍

What is the difference between predictive procurement and predictive maintenance? Predictive maintenance attempts to predict the failure time of an asset using sensor data. Predictive procurement does not require sensors for this—it evaluates existing usage and lifecycle data. seventhings provides exactly this history without calculating failure time forecasts itself.

‍

Why do companies buy too much—and still miss the actual need?

Mid-sized companies have an average of €50,000 to €200,000 worth of unused equipment in their inventory (McKinsey, 2024). The pattern behind this is almost always the same: a location reports a need, procurement places an order, and no one checks the company-wide inventory beforehand. With furniture or conference technology—classic workplace assets without their own network connection —this happens particularly often because no one has a central view of what is available at other locations.

At the same time, without tracking, 12–18% of mobile tools are lost every year (Nexess Solutions, 2024). A piece of testing equipment that can no longer be found is usually reordered rather than searched for—replacing it is faster than looking for it. Together, these two effects create a procurement pattern that is structurally too high, without anyone consciously over-ordering.

‍

Which key performance indicators are needed for data-driven procurement planning?

Not every metric is equally helpful for procurement. In practice, four data points are most relevant before a new order is triggered:

• Current inventory by location – before placing a new order, the central asset list shows what is already available and unused
• Useful life and age – an asset nearing the end of its economic life justifies a replacement, whereas a newer asset typically does not
• Damage and loss history – recurring losses in a specific asset class point to a process issue, not necessarily a genuine need for more
• Maintenance costs over time – rising maintenance costs are a more objective signal for replacement than a single complaint

These four points can be derived from a structured asset history without anyone having to manually search through Excel spreadsheets. How an organization moves from unstructured lists to reliable asset data is described in the maturity model from issue 4 of this series.

‍

‍

How much capital do companies tie up through unnecessary purchases? 67% of CFOs cite tied-up capital as one of their top three concerns (Deloitte CFO Signals, 2024). Unused equipment in inventory is a direct contributor to this. seventhings makes this inventory visible before a new order unnecessarily increases it.

‍

What does data-driven procurement look like in practice?

In practice, predictive procurement does not mean that software automatically triggers orders. It means that before any major acquisition, the procurement team receives a reliable answer to three questions: Is the device already sitting unused somewhere? How much longer is the existing stock usable? And is a repair more worthwhile than a new purchase? These questions can be answered when asset data is maintained centrally, across all locations, and kept up to date – rather than scattered across individual departments' Excel lists.

This is particularly relevant for high-volume capital goods: conference technology, furniture, testing equipment, PPE, and mobile phones without MDM. For machines with their own control systems or networked building technology, ERP and CAFM systems already handle this role—predictive procurement for physical assets fills the gap for everything in between.

A CFO who sees their assets in their entirety for the first time almost always asks the same question: Why didn't we order this sooner? The answer: Because the data wasn't in one place before. This and other questions are covered in the article on the three key CFO questions for asset management.

‍

What does this mean for your next investment decision?

Predictive procurement is not a new tool or an algorithm that automates orders. It is a shift: from "What are the locations reporting?" to "What does the history show?". The technical foundation for this is unspectacular—a well-maintained, central asset database instead of scattered Excel lists. The effect, however, is measurable as soon as bad purchases become less frequent and replacement procurement is based on actual service life rather than individual reports.

‍

What now?

1. Review your inventory – before your next major purchase, check whether comparable equipment is already sitting unused at another location
2. Analyze loss and damage patterns – identify asset classes with recurring losses before placing new orders for them
3. Conduct an asset potential analysis – a structured look at your current inventory shows how much capital is already tied up before you commit more

With seventhings inventory software , you can build the data foundation for this in under 14 days, with your first assets live in the system.

‍

‍

Frequently Asked Questions

What is Predictive Procurement?

Predictive procurement refers to purchasing decisions based on historical asset data rather than individual requests. Key data points include service life, location distribution, and loss history. 73% of ERP users still rely on parallel Excel spreadsheets (Forrester, 2023) – a weak foundation for reliable decision-making.

How does Asset Intelligence help with procurement planning?

Asset Intelligence makes it visible what is already available within the company before a new purchase is triggered. This primarily concerns furniture, conference technology, and tools without network connectivity—assets that are often only incompletely recorded in ERP systems.

Does seventhings replace an ERP system for procurement?

No. seventhings complements the ERP with a data layer that is usually missing: the actual condition, location, and usage history of physical assets without IoT sensors. Ordering processes and supplier management remain in the ERP.

What data does procurement need to avoid bad purchases?

Four data points are crucial: current inventory per location, service life, damage and loss history, and maintenance cost trends. Together, they show whether a new purchase is necessary or if existing equipment is sufficient.

For which asset classes is data-driven procurement planning particularly suitable?

It is most effective for high-volume assets without built-in sensors: furniture, conference technology, testing equipment, PPE, and mobile phones without MDM. For networked machinery or building technology, ERP and CAFM systems already perform this function.

‍

Conclusion

Bad purchases rarely happen due to carelessness. They happen because no one sees the full inventory before placing an order. €50,000 to €200,000 worth of unused equipment per mid-sized company (McKinsey, 2024) is not an outlier, but the result of a missing data foundation. Those who establish this foundation don't buy less—they buy more strategically.

‍