At any multi-rooftop dealer group operating above five locations, the CFO cannot answer a simple question: which dollar of advertising spend, across which channel, across which store, produced the most margin-positive sale last month?
Not because the question is hard. Because the infrastructure to answer it doesn't exist.
Each store's campaign data lives in a separate agency account. Attribution windows differ by platform and by vendor. CRM exports don't match ad platform reports. Reporting formats vary by store, by quarter, and by who the agency hired last year. The CFO gets a deck with impressions and click-through rates and a line at the bottom that says "above-benchmark performance" — and none of it connects to a sold unit, a gross margin figure, or a cost per vehicle retailed.
Marketing Is the Only Line Item That Doesn't Have to Prove Itself
Walk through what a dealer group CFO actually controls. F&I has a menu system, a compliance layer, and per-deal profitability tracked to the penny. Parts has inventory turns, obsolescence thresholds, and carrying cost calculations. Fixed ops has technician efficiency ratios and RO-level margin reporting. Service lanes have CP-to-warranty ratios that get reviewed every month.
Then there's marketing. A $3-6 million annual line item — more in large groups — that arrives as a PDF from an agency, summarized in metrics the agency chose, benchmarked against averages the agency selected, with no connection to the data that records what actually sold.
The CFO who would never accept "trust us, it's working" from the parts director accepts it from the marketing vendor every month. That's not a marketing problem. That's a financial controls failure.
The parallel to programmatic advertising's opacity problem is exact: in both cases, the dealer is paying for a black box and being handed a summary that was designed to justify the invoice, not audit it.
How the Silo Architecture Got Built — and Who Benefits From It Staying That Way
The attribution gap didn't happen by accident. It's the natural output of a vendor model that was never designed around dealer accountability.
When a dealer group acquires its fifth, eighth, twelfth store, it doesn't audit the acquired store's marketing infrastructure — it inherits it. That means inheriting whichever agency the previous owner used, whichever platforms that agency prefers, whichever CRM integration (or lack thereof) was already in place. The group ends up with three agencies, four attribution models, two different Google Ads account structures, and a Meta Business Manager that nobody has admin access to anymore.
The agencies, to be clear, benefit from this fragmentation. Siloed accounts mean siloed performance narratives. When Store A is having a bad month, the agency points to market conditions. When Store B is up, the agency takes credit. There's no cross-store baseline to benchmark against, no group-level dashboard that would reveal which account is actually driving results and which is coasting on organic demand. The opacity is structural — and as we've covered before, the agency report is designed to be read by someone without access to the raw data it summarizes.
The groups most hurt by this aren't the ones that grew slowly and organically. They're the ones that acquired fastest — which is precisely the acquisition pace the market is rewarding right now.
The Consolidation Wave Is Accelerating the Problem
NADA's dealer count data tells a straightforward story: roughly 22,000 rooftops in 2000, around 18,000 today. The number of rooftops has fallen while the size of surviving groups has grown. The median group isn't a husband-and-wife operation with two stores anymore — it's a platform entity that has absorbed four to eight points in the last five years and is actively looking for more.
Haig Partners' blue sky reporting documents acquisition activity running at record pace in 2025-2026. Private equity-backed consolidators — groups that came into automotive retail from outside the industry — are among the most active acquirers. These are operators who ran financial controls in other retail verticals and who recognize immediately that marketing is the one department where the normal rules don't apply.
They're not wrong to notice. And they're not going to accept it indefinitely.
The consolidation wave is not just a real estate story about who owns which franchise. It's a technology story about whose infrastructure wins at scale. The groups that own their data stack will outperform the ones that rent it — and the attribution gap is exactly the kind of rented-infrastructure problem that shows up in due diligence and suppresses multiples.
What "Unauditable" Actually Costs
The standard CFO argument for accepting marketing opacity is that it's always been this way, the spend is producing leads, and the cost of rebuilding the infrastructure is higher than the cost of the inefficiency. This argument has three problems.
First, the efficiency assumption is unverifiable. If you can't measure which channels produce margin-positive sales, you can't know whether your current allocation is optimal, wasteful, or actively misallocated. Groups routinely discover, when they first unify their attribution, that two or three channels they deprioritized were dramatically outperforming the flagship agency relationship that consumed the largest budget share.
Second, the cost of the infrastructure gap compounds. Every quarter of siloed reporting is another quarter of optimization decisions made on bad data. Budgets get renewed based on agency-reported CPL, not actual cost-per-vehicle-retailed. Channels get cut because they look expensive in the agency report — not because they produced fewer sales.
Third, the data that would answer the question is already being generated. Your DMS records every sold unit, every gross margin figure, every deal structure. Your CRM records every lead source. Your ad platforms record every conversion event. The CRM knows who bought; the ad platform doesn't — and the gap between those two systems is where the CFO's real question goes to die. The problem isn't data scarcity. It's that nobody has built the connective layer that stitches them together at the group level.
The Metric That Replaces CPL
Cost-per-lead is the metric agencies report because it's the metric they can control. Lead volume is a function of budget and bid strategy — with enough money and the right auction posture, you can generate leads at almost any CPL you target. What you cannot control, from inside a campaign dashboard, is whether those leads close — and at what gross margin.
The metric that actually matters to a dealer group CFO is cost-per-vehicle-retailed, broken down by channel, by store, and by vehicle type. That metric requires connecting the ad platform's conversion data to the CRM's lead record to the DMS's closed deal. It requires a consistent attribution model across every store in the group, not the three different models that three different agencies are using simultaneously.
Most groups don't have this metric. The ones that do — that have built or acquired the infrastructure to produce it — make fundamentally different budget decisions than the ones flying on CPL alone. They know which channel produces the buyer who shows up with a trade-in and takes the full F&I menu. They know which store's Google PMax campaigns are producing conquest customers versus re-engaging existing household names. They can rank every dollar of marketing spend by its margin contribution, not just by its click volume.
That's not a marketing capability. That's a CFO capability. And right now, most CFOs at multi-rooftop groups don't have it — because the data that would produce it lives in accounts they don't fully own or control.
How AUTONOMi Solves This
AEGIS — AUTONOMi's AI operating system for automotive advertising — is built to run every paid acquisition channel across every store in a group from a single platform. Google Ads, Meta, Microsoft, TikTok — every channel AEGIS manages flows through dealer-owned accounts, operated under the same campaign architecture, optimized against the same budget logic. The agency model, by contrast, fragments each of those into a separate account, a separate reporting format, and a separate performance narrative.
That structure is what begins to make the CFO question answerable — not fully, and not in a way that traces a click through to a sold unit (that requires CRM and DMS data that exist outside the ad platforms), but in the dimension where the answer has been most opaque: which channel, at which store, is producing the lowest cost-per-conversion against a consistent attribution model. AEGIS tracks campaign performance against ad-platform-reported conversions, normalized to the same measurement baseline across every store in the group. That's not the complete answer, but it's the answer the current agency model actively prevents.
The budget logic isn't passive. AEGIS runs continuous rebalancing across channels and campaigns — reading platform performance data, detecting underperforming accounts, and reallocating spend toward the channels where results are strongest. The AI doesn't wait for a monthly deck. It reads the platform data on a continuous basis, without a human intermediary, and every action it takes is logged in AXIOM's audit trail so the CFO can see exactly what was moved and why.
This is also how AUTONOMi eliminates the silo problem at the point of acquisition. When a group adds a new store, AEGIS doesn't inherit the previous agency's account structure — it builds the new point into the same campaign architecture the group already operates. The attribution model doesn't fragment. The reporting baseline doesn't reset. The CFO can compare the new store's campaign performance against the rest of the group from month one, in the same platform view, under the same cost accounting.
For PE-backed consolidators specifically, this matters at the deal level. The agency model that managed campaigns store-by-store is structurally obsolete in a group context — and AEGIS is built for the operating model that's replacing it.
The CFOs Who Close the Gap First Will Set the New Benchmark
The groups that solve this in the next 18 months will look back at this period the way manufacturing CFOs look back at the adoption of enterprise resource planning in the 1990s: the inflection point where some operators got real-time cost visibility and used it to permanently outmaneuver the ones still running on spreadsheets and quarterly reports.
The dealer groups that don't close it will keep renewing agency contracts based on CPL, keep misallocating budget toward channels that look productive by the wrong metric, and keep watching their cost-per-vehicle-retailed drift upward while the groups with better infrastructure take share. The PE-backed consolidators entering the market right now didn't come from outside the industry because automotive retail is easy — they came because it's an industry where the operating standards are low enough that anyone willing to install modern financial controls can outperform the incumbents.
Marketing attribution is one of those controls. It's not sophisticated. It's not optional in any other retail vertical that operates at this scale. And the technology to do it — across every channel, every store, in a single platform view — exists right now. If you want to see what your group's spend is actually producing, start with a 30-day pilot and answer the question your CFO has been unable to ask.