DMS procurement
How to Select a DMS: A European Dealer RFP and Scorecard
The strongest RFP compares the exact proposed product, market and implementation through real workflows, contractual evidence and predeclared scoring.

Key takeaways
- Score the named product, deployment and country, not vendor-level marketing.
- Use eligibility gates before weighted scoring.
- Make vendors demonstrate both normal and exception workflows.
- Require evidence for APIs, security, localisation, migration and outcomes.
- Keep public evidence status separate from final procurement verification.
1. Form the decision team and scope
DMS selection affects sales, used vehicles, workshop, parts, finance, IT, privacy and group reporting. Create a decision team with accountable operational owners, not only representatives. Name one executive sponsor, product owner, data lead, integration lead, security/privacy lead, finance controller and change lead. Define who recommends, who approves and who can reject on a mandatory requirement.
Document legal entities, rooftops, brands, countries, languages, users, transaction volumes and critical periods. Separate current scope from a plausible three-year roadmap. A requirement for every hypothetical future market can distort the decision, while ignoring likely expansion can create another replacement.
2. Convert needs into testable requirements
A requirement should name actor, trigger, data, action, output and acceptance. Replace “strong CRM” with “a web, OEM or phone enquiry is matched or created, consent is recorded, the lead is routed by brand and geography, an owner and SLA are visible, communication is captured, and a completed deal returns status without a duplicate customer.”
Build a country and OEM matrix. For every cell, record accounting, tax, invoice, payment, registration, warranty, parts, campaign, reporting, identity and language requirements. The European context is materially varied. Eurostat's passenger-car data show large differences in fleet age and powertrain by country, while EU rules on privacy, data access and e-invoicing still require local implementation.[1]
3. Use gates, weighted criteria and evidence levels
Eligibility gates prevent a high total score from hiding a fatal gap. Examples include production support for a required country, named OEM interface, statutory accounting output, data-residency boundary or migration deadline. A failed gate can be resolved only by approved remediation with a date, owner, cost and contractual commitment.
| Dimension | Illustrative weight | Required evidence |
|---|---|---|
| End-to-end functional workflows | 25% | Scripted demonstration in proposed product |
| Country and OEM fit | 15% | Named production references and specifications |
| Data, API and ecosystem | 15% | Catalogue, sandbox, limits, ownership, change policy |
| Migration and implementation | 15% | Plan, resources, acceptance, rollback, references |
| Security, privacy and resilience | 10% | Reports, architecture, DPA, DR test and controls |
| User experience and adoption | 10% | Role-based task testing and training plan |
| Five-year TCO and contract | 10% | Price model, indexation, change, support and exit |
4. Script demonstrations instead of accepting product tours
Provide representative but safe data and fixed scripts. Ask the vendor to show a lead through quote, trade-in and order; a used car through appraisal, inspection, preparation, media, publication, pricing and sale; a repair order through booking, technician work, parts, additional approval and invoice; and a period-close or management-report path.
Add exceptions: duplicate customer, wrong VIN, cancelled deal, unavailable part, failed interface, offline inspection, reversed invoice and user leaving mid-process. Count systems, clicks, rekeyed values, manual exports and invisible background dependencies. Record the version and market demonstrated.
Standards can improve interoperability, but do not substitute for a demonstration. STAR publishes automotive lead, deal and retail-delivery APIs and a retail domain model.[2] Ask whether and how a vendor implements relevant standards, then test the actual proposed interface.
5. Validate cloud, API, security and data claims
For cloud, identify SaaS, dedicated hosting or hosted legacy architecture. Request availability definitions, incident history, RPO, RTO, backup and recovery test evidence, maintenance rules and capacity model. For APIs, request objects, fields, events, write operations, authentication, sandbox, rate limits, overages, versioning, monitoring and data-export rights.
For privacy and security, assess roles, least privilege, MFA, logging, encryption, vulnerability management, subprocessors, transfer mechanism, retention, deletion, incident notification and independent assurance. GDPR requires risk-appropriate controls and privacy by design, but a certification or cloud provider does not make the dealer automatically compliant.[3] ENISA guidance can structure evidence requests, although NIS2 scope must be assessed separately.[4]
6. Apply fair public evidence statuses
| Named product/market | Open/API evidence | Security evidence | Interpretation |
|---|---|---|---|
| Nextlane Platform, Europe | Confirmed: official open-API positioning | Confirmed: AWS transformation and stated EU residency objective | Exact DMS and migration status require proposal validation |
| Pinewood platform, global/Europe | Confirmed: DMS API statement and a named integration | Confirmed: public ISO statements | Scope, reports and commercial API access require validation |
| Tekion ARC, UK | Confirmed: API agreement exists | Confirmed: trust portal lists certifications and encryption | Continental Europe maturity Not assessed |
| Omnetic, European public site | Not publicly confirmed: no reviewed technical catalogue | Not publicly confirmed: no reviewed certification/residency matrix | Request proposal evidence; do not infer absence |
| bee2link OpenFlex, Europe | Not publicly confirmed: no reviewed general catalogue | Not publicly confirmed | Product-specific due diligence required |
Public status is an orientation tool. Procurement evidence may change it. A vendor should be invited to correct factual errors and provide current confidential proof under an appropriate process.
7. Finish with implementation, references and contract
Reference calls should match country, dealer size, OEM complexity and scope. Ask what changed after contract, what required workarounds, what data failed, how long adoption took, how incidents were handled and what the reference would do differently. Do not ask only whether users like the product.
Make acceptance criteria contractual. Cover data completeness and reconciliation, critical workflows, integrations, performance, security, training, cutover and support. Price migration iterations, environments, API usage, messages, storage, report work, travel, indexation and change requests. Define service levels, escalation, exit export, transition support, deletion and retained access to statutory records.
8. Where Omnetic fits
Omnetic should be shortlisted where the RFP values shared customer and vehicle context, sales and aftersales CRM, used-car lifecycle depth, pricing and stock actions, and structured mobile inspection. Its defensible distinction is continuity from insight or evidence to an owned operational action.
A fair Omnetic proposal must still prove every gate for the named countries, OEMs and modules. Public claims such as scale, certifications and native-module count need current definitions. Security, architecture, API, SLA and migration evidence should be evaluated with the same standard applied to every vendor.
Limitations
The weights are illustrative and must not be copied without dealer priorities. The public comparison is selective and does not score implementation quality. Not publicly confirmed never means absent. Legal, security, tax and accounting requirements need specialist validation.
Frequently asked questions
Scope, market matrix, workflows, data, integrations, security, migration, support, pricing, exit and evidence instructions.
Use predeclared criteria, gates and evidence levels against the exact proposed product and market.
Usually no. Distinguish demonstrated, configurable, dependent, roadmap, not publicly confirmed and not assessed.
Use a compact set covering sales, used vehicles, workshop and finance, with exception cases.
Use identical scripts and data, record evidence, set weights first and allow factual correction.