Automotive Dealerships Have the Data, They Just Can't Access It
An AI audit for automotive dealerships reveals how DMS data hostage-taking by CDK and Reynolds blocks AI and where modern systems like Tekion open the door.
A dealer principal I spoke with last year had a phrase for his dealership management system: “the world’s most expensive jail.” Every vehicle in his lot, every deal his finance team had ever closed, every service event on every customer’s car - all of it was in CDK. None of it was accessible without paying CDK or one of their approved integration partners thousands of dollars a month for the privilege.
He was not exaggerating. The automotive retail industry has one of the richest operational datasets of any industry - vehicle inventory, transaction history, service records, customer lifetime value, F&I product attachment rates - and that data has historically been held hostage by DMS vendors who monetized access to it as aggressively as they monetized the software itself.
That is the starting point for every automotive dealership AI audit. The data exists. The question is what it costs to access it, and whether the dealer is on a platform where that answer is “reasonable” or “prohibitive.”
The DMS Landscape: A Spectrum from Open to Locked
The choice of DMS is probably the single most consequential factor in a dealership’s AI readiness. The platforms vary dramatically in how they handle data access.
CDK Global: Data Services at a Price
CDK is the market leader, running a large percentage of mid-to-large franchise dealerships in the US. CDK has a data access program called CDK Data Services that provides API access to DMS data - dealer information, inventory, repair orders, customer records, and financial summaries.
The practical reality of CDK’s data access program is twofold. First, it is expensive. Third-party integrators pay significant monthly fees for API access, and those costs get passed to dealers who want custom integrations built on top. Second, CDK’s relationship with third-party integrators has been contentious. The 2023-2024 antitrust litigation involving CDK’s data access restrictions and the settlement that followed brought industry attention to practices that dealers had been frustrated about for years. Access has improved somewhat post-settlement, but CDK’s data services model is still a commercial relationship first, an open ecosystem second.
For a dealership on CDK, the audit findings typically show: rich data exists, API access is technically available, but integration costs need to be factored into the AI ROI calculation from the start. A data extraction layer for a CDK dealership can add $2,000-$4,000/month in platform fees before a single line of AI code is written.
Reynolds and Reynolds: The Approved-Vendor World
Reynolds ERA is the other major DMS in franchise automotive retail, and its data access model is even more restrictive. Reynolds runs what it calls an “approved vendor” program - third-party developers who want to integrate with ERA data must apply to be Reynolds-approved, sign commercial agreements, and pay ongoing fees.
The practical result is that building on Reynolds data is slower and more expensive than building on almost any other major business software platform. Reynolds has faced legal challenges related to these practices, but the approved-vendor model remains the primary access pathway. A dealership on Reynolds that wants to build AI on top of their operational data is looking at a longer lead time and higher vendor coordination cost than a dealership on any other platform.
This is not a reason to not pursue AI - but it is a constraint that needs to surface in an honest systems assessment. We flag it clearly and factor it into the priority matrix.
Tekion: The Modern Open Ecosystem
Tekion is the new entrant that has changed what “modern DMS” looks like. Built cloud-native with a REST API from the start, Tekion exposes inventory, deal data, service records, customer data, and financial information through a well-documented API with reasonable access terms. There is no approved-vendor program and no five-figure monthly integration fee for a motivated developer to start pulling data.
For a dealership that has migrated to Tekion - or is considering it - the AI readiness picture is completely different. The data access question is answered before the audit even starts. The conversation moves immediately to what to build rather than whether building is feasible. Tekion is growing rapidly in the market specifically because franchise dealers are aware of this difference.
VinSolutions and DealerSocket: The CRM Layer
A significant number of dealerships run a separate CRM on top of their DMS, and the two most common are VinSolutions (part of Cox Automotive) and DealerSocket. Both have solid REST APIs with good documentation.
VinSolutions’ API gives access to leads, customers, vehicles of interest, and communication history. For AI use cases centered on the sales and customer communication side - lead response automation, prospect follow-up, lost lead re-engagement - VinSolutions is a workable foundation even when the underlying DMS is CDK or Reynolds. The tradeoff is that CRM data is a subset of total dealership data. Service history, F&I performance, and deal structure still live in the DMS.
Dealertrack, another Cox Automotive product used primarily for F&I and compliance, has its own API and handles credit applications, financing, and deal submission. It is wireable but narrow in scope.
What the Data Hostage Story Means for AI
The core thesis of the automotive dealership AI audit is this: a dealership probably has more customer and operational data than almost any other service business of comparable size, and yet that data is often harder to access for AI purposes than a smaller business’s data sitting in a modern CRM.
A 30-person service company using HubSpot can have a voice agent reading and writing customer records in a week. A $30M automotive dealership on Reynolds might spend three months on vendor negotiations before a single integration is live.
This is why we assess DMS access explicitly in the first week of any automotive audit. The findings shape every subsequent decision.
The Inventory Problem
Vehicle inventory is one area where data is often more accessible than it appears. Most dealerships export inventory to third-party listing platforms (Cars.com, AutoTrader, CarGurus), and those platforms maintain their own inventory feeds. Some dealerships also use inventory management tools like vAuto (also Cox Automotive) which has its own API.
If the DMS data access situation is difficult, inventory data can sometimes be sourced from these downstream feeds rather than the DMS directly. It is a workaround, not a solution - the feeds are typically delayed by hours rather than real-time, and they don’t include deal history or service records. But for AI use cases that only need inventory data (website chat assistants, online appraisal tools), it avoids the DMS access problem entirely.
Service Lane: The Underused Opportunity
Fixed operations - service and parts - is consistently the highest-margin department in a dealership and the one where AI has the clearest near-term ROI. Service appointment scheduling, declined services follow-up, recall outreach, and maintenance reminder campaigns are all AI-tractable workflows.
The data needed for these use cases is primarily customer-vehicle history and open recall data. Customer-vehicle history lives in the DMS service records. Open recall data is publicly available through NHTSA’s API (free, no commercial agreements required). This combination - DMS service data plus NHTSA recall data - enables several high-value AI workflows even for dealerships with challenging DMS access situations.
For clients where we have built automotive voice agents and dashboards, we learned that service lane automation often delivers faster ROI than sales-side AI because the appointment booking workflow is more standardized and the data needed is narrower. That pattern holds in our audit findings too: service lane is usually the first priority in the AI blueprint for automotive clients.
The AI Blueprint for a Franchise Dealership
After a full automotive dealership audit - covering the DMS access situation, CRM data quality, inventory management setup, and service operation data - the blueprint we deliver usually phases out something like this:
Phase 1 targets the wireable-now opportunities: service appointment automation using the DMS or a scheduling layer built on top of it, recall and maintenance outreach using NHTSA data plus customer records, and lead response automation through the CRM API (VinSolutions or DealerSocket). These can be live in 6-10 weeks regardless of which DMS the dealer is on.
Phase 2 addresses the DMS access situation directly - either building the integration through CDK Data Services or Reynolds’ approved-vendor program, or evaluating whether a Tekion migration makes sense given the long-term AI roadmap. This is where the most significant investment decision sits.
Phase 3 is the intelligence layer: predictive service marketing (which customers are due for maintenance based on mileage patterns and service history), lead scoring based on engagement and vehicle search behavior, and F&I product recommendations based on deal structure patterns. These require clean, connected data across DMS, CRM, and service systems - which is why they come last.
The AI readiness audit is where dealers discover which phase they can start in and what the cost of reaching each subsequent phase actually looks like.
Frequently Asked Questions
We’re on CDK. What does AI integration realistically cost for us?
The honest answer is that CDK data access fees add meaningful overhead to any AI project. Third-party integration access through CDK Data Services runs $1,500-$4,000/month depending on the data objects you need access to and which integration partners are involved. This is a recurring cost, not a one-time fee. The audit will calculate this as part of the total cost of ownership and compare it against the projected ROI of the specific use cases you want to pursue, so you can decide whether the math works.
We’re considering switching to Tekion. Should we do the AI audit before or after the migration?
Before, ideally. A migration to Tekion changes the AI readiness picture significantly - the audit will show you what becomes possible post-migration that isn’t possible today, which makes the migration decision easier to justify. The audit’s systems assessment will also flag any data migration considerations (service history, deal records, customer data) that need to be part of the Tekion implementation plan.
Our CRM is VinSolutions. Can we build AI on that without touching the DMS?
For sales-side use cases, yes. Lead management, follow-up automation, BDC AI support, and customer communication workflows can all be built on the VinSolutions API without requiring direct DMS access. The limitation is that you’re working with CRM data only - you won’t have visibility into service history, F&I performance, or deal structure from the DMS side. That’s fine for Phase 1 sales AI. For the full intelligence picture, you eventually need the DMS connection.
What about the BDC (Business Development Center)? Is that a good AI starting point?
BDC is one of the highest-value AI entry points in automotive retail, precisely because it is the most repetitive, scripted work in the dealership. Inbound lead response, appointment setting, missed call follow-up, and service scheduling calls all follow predictable conversation patterns that AI handles well. We have built voice agents for automotive clients specifically for inbound BDC workflows and the ROI case is strong - a voice agent handling after-hours calls and overflow during peak periods costs a fraction of an additional BDC rep and never misses a call.
What does an automotive dealership AI audit typically cost and how long does it take?
The audit is a structured 2-3 week engagement. The deliverables include a full process map of your sales, F&I, service, and BDC workflows; a data and systems assessment covering your specific DMS, CRM, and ancillary platforms; a quantified waste report; a risk map; and a phased AI blueprint with a priority matrix. The audit price is separate from any implementation cost - it gives you an independent view of what to build before you commit to building anything.
Ready to Get Started?
Tell us what you're working on. We'll review every submission and respond within 24 hours.
Founder of Bosar Agency. Building AI systems that actually work for service businesses — voice agents, chatbots, and automations that capture leads 24/7.