When a large manufacturer or distributor chooses AI for bills of quantities, quotes and orders, it is not enough to test how well the software matches line items. It matters just as much whether the software fits an enterprise landscape: the IT security requirements, SAP, the master data and a structure with several companies and locations. This checklist sums up the criteria that have proven themselves in evaluations and shows, point by point, how Mercura meets them.

Key takeaways

• Check five areas in the proof of concept: security, identities and roles, ERP integration, master data and group structure.

• Mercura is ISO/IEC 27001 certified, GDPR-compliant and hosted in Europe. Every customer has its own tenant.

• Login with Microsoft Entra ID, user provisioning via SCIM, integration with SAP R/3, ECC and S/4 via IDoc, OData or SAP BTP.

• Large companies such as REHAU (revenue above EUR 4bn) and Bauder (EUR 1.13bn) use Mercura in inside sales.

Why enterprise criteria belong in the proof of concept from day one

Many evaluations run like this: the business team tests the AI with real bills of quantities, likes it, and only then does IT review security and interfaces. If the product fails there, the test was wasted. It is better to include the enterprise questions in the proof of concept from the start and ask the vendor for concrete answers, not intentions.

The following five areas cover what IT, procurement and data protection teams in large companies typically ask.

1. Security and data protection

You will find more detail in our article on security for AI in inside sales.

2. Identities, roles and access

Nobody in a large company wants to maintain yet another user directory. Make sure the software connects to your identity management.

• Single sign-on: Mercura supports login with Microsoft Entra ID (SAML/OIDC).

• User provisioning: users are created via SCIM from Entra ID, Okta or other identity providers. Roles follow your security groups.

• Roles: member, contributor, manager and admin.

• Several organisations: people who work for several companies switch between organisations without logging in again.

3. ERP integration, especially SAP

The AI should not only prepare quotes and orders but hand them over to the ERP. Ask specifically which routes are used in production for your system.

Mercura connects SAP R/3, SAP ECC and SAP S/4 via IDoc (ORDERS05), OData (quotations plus a daily sync), SOAP via SAP BTP or file export. The ERP stays the system of record: prices and conditions come from the ERP. At Sanitär-Heinze, for example, processed requests go to SAP for pricing through an OData interface.

For other systems there is a documented API, signed webhooks and SFTP. Our article on SAP integration for quotes and bills of quantities explains the details. For a general overview, see ERP integration.

4. Master data and synchronisation

The quality of the suggestions depends on the article master. In an enterprise setting, what matters is that the data stays current regularly and without manual work.

• Article master via CSV or Excel up to 2 GB, with a column-mapping wizard

• Data formats such as BMEcat and DATANORM via SFTP

• Master data import and export through the API

• Scheduled pull directly from your system: your admins set up the connection, mapping and schedule themselves

An Integration Center shows the status of every connection. One shared log lists imports, API calls, webhooks and ERP exports, and admins get an email when something fails.

5. Several companies and locations

Corporate groups work with several companies, locations and teams. Check whether the software can reflect this structure without a separate system for each company.

In Mercura, each company can be its own organisation with its own article master and its own rules. Rules route incoming requests automatically to the right location and the responsible person. Exports carry the location of the request, and ERP exports run per location. Read more in Multiple companies, one platform.

How to set up the proof of concept

1. Real data: test with bills of quantities, PDF requests and orders from recent weeks, not with hand-picked examples.

2. Measurable goals: define up front what counts, such as share of line items with a suggestion, hit rate and time per quote.

3. Involve IT early: run the security questionnaire, SSO and ERP route in parallel with the business test.

4. Test the hand-over to the ERP: a quote is only finished when it is in the ERP.

With Mercura, a technical test with your own data usually runs within two working days, and first workflows are live in under four weeks. Danfoss went from proof of concept to go-live in two weeks.

What large companies achieve

• REHAU (revenue above EUR 4bn, 13,000 employees): for up to 98% of bill-of-quantities items the AI proposes a matching product, with a hit rate of up to 90%. Inside sales saves 60% of the time spent on quotes. Next, REHAU plans to roll it out to further departments. Read the customer story

• Bauder (revenue EUR 1.13bn, 1,750 employees): 25% more quotes processed. Read the customer story

In total, more than 80 distributors and manufacturers in Europe and the US work with Mercura. Our overview of vendors shows how Mercura compares.

Frequently asked questions

Is Mercura ISO 27001 certified? Yes. Mercura is ISO/IEC 27001 certified.

Where is the data stored? In Europe, in an Azure data centre in Frankfurt. For customers in the US there is a separate US region.

Does Mercura support single sign-on? Yes, via Microsoft Entra ID (SAML/OIDC). Users can also be managed via SCIM from Entra ID, Okta or other identity providers.

Which SAP systems can Mercura connect to? SAP R/3, SAP ECC and SAP S/4, via IDoc, OData, SOAP via SAP BTP or file export.

Can Mercura handle several companies? Yes. Each company can be its own organisation, and users switch between organisations. Within an organisation, rules route requests to locations and responsible people.

How fast is a proof of concept possible? A technical test with your data usually runs within two working days, and first workflows are live in under four weeks.

If you want to check the points on this list with your own data, book a demo.

Sources

• Customer stories of REHAU, Bauder, Danfoss and Sanitär-Heinze on mercura.ai, success stories section, retrieved 10 November 2026

• Mercura product documentation on security, integrations and API (developer documentation), as of November 2026