From "just purchase this" to structure: How AI is revolutionizing the specification of requirements
Monday morning, 8:17 a.m. The next e-mail reaches the operational purchaser's inbox: “Hello, we need the same thing as last time. See attachment. Thank you!”
The attachment? A scanned PDF invoice from a supplier from last year or – if things go well – a current PDF sample offer. The purchaser knows that this means queries, follow-up research, copy and paste. What should actually be a quick price request process drags on – or is abandoned altogether.
Instead of an automated, digital process, a time-consuming e-mail ping-pong begins. This is not an isolated case – it is everyday life. And it is precisely here that the great potential for intelligent automation in purchasing can be seen.
- AI-supported requirement specification creates structure before purchasing begins
- This is precisely where FUTURA Smart comes in: Specialized AI agents automatically analyze, check and structure all input. Even unstructured e-mails or PDFs are converted into an RFQ-capable format – without manual rework.
- Purchasing only gets involved once the Requirement has been fully prepared. This means: fewer queries, more tenders, greater transparency
The pain of “just purchase this”
Unstructured or incomplete requirements are among the biggest time wasters. Many are based on old supplier offers, are not comparable, contain gaps – and end up sent to the purchasing department by e-mail as a PDF.
Many companies lack a standardized purchase requisition process – often deliberately. Why? Because the initial hurdles for consumers are too high: SAP is complex, the transactions are not self-explanatory, and the form-based thinking is off-putting. The result: workarounds arise via e-mail or telephone that are neither systemic nor evaluable.
👉 The result: Queries, delays, additional work.
👉 The reaction: Value limits are raised in the organization to avoid the extra work.
👉 The result: Potential savings remain untapped.
Structure instead of queries: AI-supported requirement specification
With the right use of AI, this process gap can be closed. FUTURA Smart puts AI exactly where it has the greatest impact: at the beginning.
What the AI agents do:
- Recognize gaps or contradictions in demand reports
- Ask specific questions for clarification
- Neutralize PDF offers and convert them into neutral, processable tenders or requests for quotation
- Automatically assign product groups, EKORG etc.
For the purchaser, this means no more manual rework. The request for quotation is prepared – the RFQ process can start immediately.
How AI is changing the process
| Formerly | Today with AI agents | |
| Notification of requirements | "Can you just purchase this?" | Intelligent input with queries from the AI |
| Basis for quotation | Old PDFs, e-mails from suppliers | Neutralized, structured free-text requests |
| Information gaps | Questions, loops, additional work | Complete specification with the first step |
| Classification | Manual assignment of material groups & purchasing organization | Automatic classification by the AI |
| Sourcing rate | Many requirements above the value limit | More requests for quotation being issued, more competition |
What is possible? – How to recognize untapped potential
The greatest potential in operational purchasing lies where it is often underestimated: At the beginning. Companies that analyze how many requests for quotation are not issued – and why – quickly realize that it is rarely due to the volume.
If you want to know how much efficiency, transparency and savings you can make in your purchasing, you should ask the following questions:
1. Analyze proportion of manual requisitions
- How many requirements come in by e-mail or as a PDF?
- How many of them need to be reworked manually?
👉 The difference between incoming demands and the RFQ issued in a system-compatible manner is a good indicator of efficiency losses.
2. Check the rate of direct awardings below the value limit
- How many purchase orders are placed without a request for quotation – and why?
- Is there a noticeable increase in certain material groups or areas?
👉 Often, these workarounds are a reaction to a lack of structure at the outset.
3. Measure the processing time and loops in the purchasing process
- How many queries are needed on average before a request for quotation can be issued?
- How long is the lead time from initial contact to RFQ release?
👉 This is where the operational effort – and the leverage of an automated requirement specification – becomes apparent.
4. Compare sourcing rate vs. total demand
- How high is the proportion of procured volumes that are actually the subject of a request for quotations?
- What would be possible if more requirements were available in a structured way?
👉 The economic potential of increased competition can be directly derived from this.
Conclusion: Automation without data quality is incomplete
What used to be a pragmatic practice – “just purchase this as you did last time” – now measurably costs efficiency, transparency and competitiveness. The requirement specification is a crucial lever. Because only structured purchase requisitions make an automated process possible at all.
This is precisely why FUTURA Smart, with its AI agents, comes into play at the first and crucial point in the purchasing process: The requirement specification. Here, unclear e-mails are converted into a standardized, system-compatible RFQ ‒ automatically, comprehensibly, integrated into SAP.
More on the topic:
The RFQ Maker is part of our trade fair motto "RFX meets AI", with which we will be on site at this year's eLÖSUNGSTAGEN in Düsseldorf. Want to experience it live? Then visit us at our stand!
