Choosing quotation automation software is harder than it looks in a demo.
Most tools perform well when shown a clean, well-formatted RFQ from a known customer with a recognisable article number. The demo is designed that way. What the demo rarely shows is how the system handles a free-text email from a customer who doesn't know your product codes, a BoQ in GAEB format with 200 positions and a two-day deadline, or an exception that arises because the ordered product was discontinued three months ago.
Before you score
Before applying any specific criteria, confirm which category of software you are actually evaluating. Two fundamentally different categories are often presented under the same label of "quotation automation," and buying the wrong category is a more expensive mistake than choosing the wrong vendor within the right category.
Category
Quote creation tools (CPQ)
These are tools designed to help sales reps build quotes faster, typically through guided configuration, price calculators, and approval workflows. The input is a sales rep sitting at a keyboard. The tool makes their quoting process faster and more consistent.
Category
Inbound quotation automation
These tools handle the inbound side: a customer sends a request by email, and the system reads it, interprets it, matches it to your product catalog, generates a draft quote, and routes it for review. The input is an email in an inbox. The human reviews and approves rather than builds from scratch.
The scoring framework
The criteria below are designed to predict production performance, not demo performance. There is a significant difference between the two.
The scoring criteria
Criterion
Input Format Coverage
"What percentage of our real inbound quote requests can your system process without manual pre-processing?"
Your customers do not send quote requests in a single consistent format. They send free-text emails, Excel spreadsheets, PDF attachments, GAEB files, forwarded email chains, and occasionally handwritten notes photographed and attached as JPEGs. Within each format, every customer has their own conventions.
A tool that handles only structured PDFs or only GAEB files covers part of your inbound volume. The rest routes to manual. If that remainder is 30% of your requests, you have a partial solution, and the 30% is likely the most time-consuming 30%, because it tends to be the least structured.
The only reliable test
Run the tool against a sample of your actual inbound requests, not the vendor's sample documents. Request this explicitly. Any vendor unwilling to process a real sample before the contract is signed is telling you something.Criterion
Product Matching Capability
"Show me how your system handles a request where the customer doesn't provide a product code, only a description or a reference to a previous order."
Product matching is where most inbound quotation automation tools produce incorrect or incomplete results in production. Three scenarios expose matching limitations.
When there is no article number, only a description, the customer writes "the M12 connector with the push-pull locking mechanism, 4-pin." Your catalog has forty-seven M12 connector variants. A rule-based lookup fails. An ML-based tool requires training data. An LLM-based tool interprets the specification and narrows by matching attributes.
When the customer sends their own internal part number or BOM reference, your system needs a cross-reference to resolve it. The question is what happens when the cross-reference table is incomplete.
Vague contextual references are the hardest case. "Same as what we ordered in June" requires order history access and reasoning, not catalog lookup. Ask directly whether the system can access and reason about order history.
Criterion
ERP Integration Depth
"Does your system write directly to our ERP, or does the output require a manual import step?"
ERP integration is the single most important technical criterion, and it is the one most obscured by vague vendor language. "Integrates with SAP" can mean anything from "we have a CSV export you can import manually" to "we write confirmed quote data directly to the ERP as a sales quotation record, reading customer master, product master, and pricing conditions in real time." Those are not equivalent. Only the second one removes manual steps from the process.
The integration questions that reveal depth:
For mid-market European companies, the most common ERP environments are SAP Business One, SAP S/4HANA, and Microsoft Dynamics 365. Confirm that the vendor has production deployments (not just claimed compatibility) with your specific ERP and version.
Criterion
Exception Handling Quality
"What does my team see when the system can't resolve something? Show me an actual exception card."
Every quotation automation tool routes some requests to human review. The question is whether the exception is presented with enough context for the human to resolve it quickly, or whether it arrives as a notification that something needs attention, leaving the human to reconstruct the context from scratch.
Good exception handling presents: the original request, what the system found, what it couldn't resolve and why, and a suggested resolution or next step. The human confirms or overrides in under two minutes without reopening the original email.
Bad exception handling presents: a flag in a queue. The human opens the queue, opens the original email, opens the ERP, and rebuilds the context manually. The automation has added a step, not removed one.
In a quoting environment
Exception quality matters more than exception rate. A system that generates exceptions for 20% of requests but resolves each one in 90 seconds is more useful than a system with a 5% exception rate where each exception takes 15 minutes because the context is absent.Ask to see a live demonstration of the exception handling interface, not a screenshot: a live session where the vendor walks through resolving an actual exception.
Criteria continued
Criterion
Implementation Timeline and Requirements
"What does implementation require from our side, how many hours per week, and what is the realistic go-live date from contract signature?"
Mid-market companies do not have dedicated AI implementation teams. An implementation that requires months of discovery, data labeling, model training, and integration consultancy is an enterprise solution presented to a mid-market buyer, and the mismatch will show up in timeline slippage and cost overruns.
The specific questions:
LLM-based advantage
LLM-based systems do not require training data. A system that can process your documents on day one, without a labeled dataset, compresses the implementation timeline significantly and removes a common source of project risk.Criterion
GDPR and Data Residency
"Where is our data processed and stored, and can you guarantee EU data residency by contract?"
For European mid-market companies, this is not an optional compliance checkbox. Quote data contains customer information, commercial terms, and pricing that is commercially sensitive. Many DACH-region companies have internal IT policies requiring EU data residency for all cloud-processed business data.
The specific questions:
For DACH-region companies
Vendors who cannot answer these questions directly in the evaluation process are vendors whose legal and compliance posture has not been designed for the European market. This is a disqualifying gap.Criterion
Pricing Model and Total Cost of Ownership
"Walk me through the total cost including implementation, integration, and ongoing fees at our current and projected quote volume."
Per-quote or per-document pricing is common in this category. It appears simple but can produce surprises at volume. The questions that reveal the real cost:
Before comparing on headline price
Build a total cost model first. A tool priced at €0.20 per quote that requires six months of paid implementation and annual integration maintenance may be more expensive over three years than a tool priced at €0.50 per quote with implementation included.The scoring worksheet
| Criterion | Weight | Vendor A | Vendor B | Vendor C |
|---|---|---|---|---|
| Input format coverage | ||||
| Product matching capability | ||||
| ERP integration depth | ||||
| Exception handling quality | ||||
| Implementation timeline | ||||
| GDPR and data residency | ||||
| Pricing / TCO | ||||
| Weighted total |
How to weight the criteria
Weight the criteria by what matters most for your operation. For most mid-market B2B companies in Europe, ERP integration depth and GDPR/data residency carry the most weight: they are the criteria most likely to create downstream problems if underspecified. Exception handling quality and product matching capability determine day-to-day operational performance.The one question that surfaces capability fastest
If you have time for only one question in a vendor evaluation, make it this:
"Send me a real email from one of our customers: one that is vague, has no article number, and references a previous order, and show me live what your system does with it, from inbox to draft quote."
A vendor who can do this with your actual email, using your actual product catalog, producing a draft quote that a human could review and approve in under two minutes, is a vendor whose system works as claimed. Everything else is a demo environment.
Where turian sits
Turian's RFQ Intake agent is an inbound quotation automation tool (Category 2, not CPQ). Against the criteria in this guide:
Criterion 01
Input formats
Criterion 02
Product matching
Criterion 03
ERP integration
Criterion 04
Exception handling
Criterion 05
Implementation
Criterion 06
GDPR
Criterion 07
Pricing
Book a 30-minute quotation workflow review
If you are mid-way through an evaluation and want a second opinion on your scoring before you finalise, this is the fastest way to test the criteria above against your actual quote mix.
By clicking "Accept", you agree to the storing of cookies on your device to enhance site navigation, analyze site usage, and assist in our marketing efforts. View our Privacy Policy for more information.
