The Dealer Group Stack: What 50 Rooftops Need That One Doesn't

A single rooftop runs its intake off a settings page. One manager owns it, one process, one lender panel, one report that reconciles because there is nothing to reconcile against. Fifty rooftops have none of that, and the difference is not volume. Group 1 Automotive reported 253 dealerships holding 313 franchises across 36 brands in first-quarter 2026 results published 30 April 2026.

Dealer group software requirements differ from single-store requirements in kind rather than degree. A group needs configuration that inherits from group and brand level while allowing store overrides, access that follows people across sites, and reporting that survives stores being bought, sold and moved between regions. That is an architecture, not a larger licence.

How much of North American auto retail is owned by groups now?

Group ownership is the growth story in North American auto retail. NADA counted 16,990 franchised light-vehicle dealers in the United States for 2025. Automotive News reported the top ten groups taking 11.1% of US new-vehicle sales, and Kerrigan Advisors recorded 478 dealership transactions in the twelve months to March 2026.

Kerrigan Advisors, in the Blue Sky Report released 15 June 2026, put those 478 trailing-twelve-month transactions at 21% above the prior year and 114% above the pre-pandemic five-year average, with 108 of them multi-dealership deals. Haig Partners, in its Q1 2026 report released 27 May 2026, counted 139 rooftops changing hands in the quarter alone, a 39% rise. Both firms are buy-sell advisors publishing their own research, so read the direction rather than the decimal.

The same report puts the accumulation at roughly 850 dealerships added by the top 150 groups over the past decade, a 25% increase. Kerrigan Advisors' 2026 OEM Survey, published 14 July 2026 from roughly 155 manufacturer executives, found 45% expecting fewer dealers in their networks within five years, up from 33% a year earlier.

Growth has also stopped depending on acquisition. Automotive News, in its 2026 Top 150 Dealership Groups analysis published April 2026, reported that 101 of the 150 groups grew sales and 62 of those grew without adding a store. A group that grows by running its existing rooftops better is a group whose operating systems are load-bearing.

What actually changes between one rooftop and fifty?

The change is in what has to be modelled. One store needs a user list, a process and a report. Fifty stores need a hierarchy, because almost nothing at a group is uniform: people work at more than one site, brand programs differ, lender panels overlap without matching, and the store set changes during the reporting period.

Count the objects rather than the stores. Asbury Automotive Group reported 158 dealerships holding 202 franchises across 34 brands on 28 July 2026; AutoNation reported 323 franchises across 245 stores and 30 brands in its 2025 Form 10-K filed 12 February 2026; Lithia Motors reported 455 stores and 54 brands across three countries in its own 2025 Form 10-K, filed 25 February 2026. The franchise count exceeds the store count every time, so "per store" and "per franchise" answer different questions, and a system that knows only one will quietly answer the wrong one.

Table 1: ten dimensions of an intake and verification stack, as a single rooftop meets them and as a fifty-rooftop group meets them. The right column is what the left column becomes once the same requirement has to hold across stores acquired at different times, selling different brands and moving between regions.

Dimension	One rooftop	Fifty rooftops
Configuration	A settings page, one owner	Group defaults, brand overlays and store overrides, with precedence rules
Identity and access	A staff list	People assigned to several sites, roles spanning regions, access that changes when a person moves
Brand programs	One or two franchises	Dozens of franchise programs with different requirements per store
Lender panel	The local panel	Group agreements, regional panels and store exceptions that overlap imperfectly
Reporting unit	The store	Store, franchise, brand, region and group, each with its own denominator
Comparability	Month over month	Same-store cohorts, because the store set changes during the period
Change rollout	A conversation with a manager	A staged program with a pilot, a schedule and a rollback path
Failure blast radius	One store	Every store inheriting the same default
Adding a site	Initial setup	An inheritance decision plus migration of an acquired store's history
What is being bought	A licence	An architecture, and the operating discipline to run it

Why is the configuration hierarchy an organisational problem?

The configuration hierarchy problem is an organisational problem wearing a technical costume. The lattice of group defaults, brand overlays and store overrides is simple arithmetic. What is hard is that every override has an owner, a history and usually a person who negotiated it, and the software makes those ownership questions explicit for the first time.

Three questions decide the design and none is technical. Who can override a group default, and who approves it. When group policy and a store override disagree, which wins. When a new store joins, does it inherit the group's configuration on day one or start from its own setup and converge later.

The second is where groups stall. A group that cannot tell its best-performing general manager that the group standard applies to their store does not have a configuration problem. It has a governance problem, and no vendor selection resolves it. Software with overrides at every level faithfully encodes whatever settlement the group actually operates, including the settlement where nobody is in charge.

Stores arrive carrying history. A group of fifty was assembled over years, and each acquisition brought its own systems, contracts, process and people, most of whom stayed. The F&I manager who knows which lender takes which profile on a Friday afternoon keeps that in their head and in a spreadsheet, and an integration plan that says "we will standardise" is proposing to delete the thing that made the store worth buying. Sometimes that is right. It is never free, and it is never a settings change.

Brand requirements are the constraint a group cannot standardise away. A store selling a luxury import and a store selling a domestic truck brand answer to different manufacturer programs, facility expectations and retail requirements. The Kerrigan Advisors 2026 OEM Survey found 43% of manufacturer executives expecting to require new image facilities within five years, a brand-level requirement landing store by store. Group uniformity has a ceiling set outside the group.

What a full standardisation costs is on public record. Asbury Automotive Group disclosed on 28 July 2026 that roughly 70% of its stores had converted to the Tekion dealer management system, up from about 25% at the end of 2025, with completion expected in the autumn. The same release excluded $6.0 million of implementation expense from adjusted results and described near-term operational effects during the conversion. A single-platform rollout across 158 stores is about a year of work with a cost line and a disclosed drag while it runs, and any group evaluating a platform is buying that project whether or not the demo mentions it.

Why are group reporting roll-ups harder than they look?

Roll-ups are hard because the store set is not constant. A group's reported numbers are not the sum of its stores' numbers when stores were bought and sold during the period, which is why every public dealer group reports same-store results alongside totals. The reporting primitive is a cohort definition, and cohort definitions are policy rather than configuration.

The filings show it directly. Lithia Motors disclosed acquiring 17 stores and divesting 12 during 2025 in its Form 10-K filed 25 February 2026, with net investment of $751.0 million expected to add nearly $2.4 billion in annualised revenue. Asbury Automotive Group reported second-quarter 2025 total revenue up 3% against same-store revenue up 5%, in results released 29 July 2025, and Penske Automotive Group reported F&I gross profit per unit of $1,812 on a same-store basis in full-year 2025 results released 11 February 2026.

Two mechanical problems sit under that: the denominator has to hold across brands and franchises, and partial periods need a consistent cohort rule. The third is history, and nobody specifies it until it breaks. When store 34 moves from one region to another in April, does last quarter's regional total rewrite itself. Both answers are defensible. An operational trend line wants the current structure applied backwards so the comparison is like for like; a review of what a regional manager was accountable for wants the structure as it stood. A system that does only one forces the group into whichever answer the vendor happened to build, discovered the first time a regional number moves without anyone touching a deal.

Effective-dated organisational structure is the least glamorous item on a group requirements list and the one most often missing.

What belongs on a dealer group software requirements list during procurement?

A group procurement list for intake and verification works better as operational and commercial questions than as a feature matrix. The useful ones test the hierarchy, the rollout, the reporting behaviour and the exit. These eight separate vendors quickly, and asking them of incumbents too gets better answers from everyone.

1. Show a configuration change made at group level and then overridden at one store, in the product. Who makes each change, who approves it, and where does the record of who changed what and when live.
2. When a new store joins, what does it inherit on day one, and how long until it transacts. Name the last group our size that did it and how long that took.
3. When a store moves between regions, what happens to last quarter's regional numbers, and can the system produce both answers.
4. Which of our brand programs have been configured on this platform before, by name, and which would be new work.
5. What does year three cost for fifty stores, including implementation, per-store fees, integration work and the internal headcount to run it.
6. If we stop paying, what comes back, in what format, how long it takes and what it costs, answered separately for configuration, audit records and applicant history.
7. Name three groups our size live on the platform today, and one that left. What happened.
8. When something is built for one group in your book, do the rest of us wait, and who decides that queue.

Numbers 6 and 7 are the ones the vendor publishing this article answers with the most hedging, which is why they are on the list. Exit terms are where every vendor's incentives run against the customer's, and one that will not price and schedule a full extraction before signature is asking for trust it has not earned. Number 7 is worse, because the true answer usually involves something the vendor got wrong. Number 8 sits beside them: a vendor running white-labelled deployments has a finite build queue, and a group that never asks where it sits in it finds out later, at a worse moment.

What does the group picture look like in Canada?

Canada publishes almost nothing about group structure. CADA's 2024 Data Report profiles 3,783 franchised new-car dealerships and carries the only public Canadian distribution of ownership by group size: 91.65% of owners running one to five stores in 2024, 5.18% running six to ten, 2.33% eleven to twenty-five, 0.60% twenty-six to fifty, and 0.22% more than fifty.

That table counts owners, not stores. The one-to-five share rose from 88.54% in 2023 to 91.65% in 2024, which reflects the reporting base rather than Canadian auto retail deconsolidating.

AutoCanada gives the one listed data point. It reported 64 franchised dealerships across eight provinces representing 23 brands, plus 12 in Illinois, in fourth-quarter 2025 results released 18 March 2026, alongside roughly $115 million of annualised run-rate cost savings from its transformation program and disclosed operational disruption during implementation. Canada has no equivalent of the Automotive News Top 150 and no published buy-sell transaction count.

What we couldn't verify

Nobody publishes how many dealer groups run a single intake or verification configuration across all their stores, in either country. We looked at NADA, CADA, Automotive News, Kerrigan Advisors, Haig Partners, DesRosiers and the filings of six public dealer groups. The number that would settle this article's central question is not public anywhere.

Technology counts per store are measured indirectly. DealerSignals, an independent auto-retail data publisher, reported on 16 June 2026 from a scan of 1,760 dealerships that franchise dealers run an average of 5.2 of 15 core retail-technology categories against 2.8 for independents. That method detects presence on a website rather than depth of use, which the report states.

Rollout timelines and costs are disclosed in useful detail by one filer. Asbury's conversion percentages and $6.0 million implementation expense are the clearest public figures we found, and one company's project is not a benchmark. Nobody publishes average implementation duration, per-store configuration effort, or how often a rollout is paused. Kerrigan Advisors and Haig Partners use different transaction definitions and periods, so their counts are not comparable, and the Automotive News Top 150 figures come from a ranking built on reported group data rather than audited filings.

Common questions

What are dealer group software requirements that a single store does not have?
Configuration inheritance across group, brand and store levels with defined precedence, access that follows people working at several sites, effective-dated organisational structure so reporting survives stores moving between regions, same-store cohort logic, and a staged rollout with a rollback path. A single rooftop needs none of these.

Why can't a dealer group just buy more licences of a single-store system?
The requirements are relational rather than additive. Fifty stores generate store-by-brand-by-lender-by-role-by-region combinations that a single-store design has nowhere to hold, and group reporting needs a history of the organisational structure itself, which a system built for one rooftop has no reason to keep.

Why are group reporting roll-ups difficult?
The store set changes during the reporting period. Lithia Motors acquired 17 stores and divested 12 during 2025, per its Form 10-K filed 25 February 2026, which is why public groups report same-store results alongside totals. Roll-ups also need one consistent denominator across stores, franchises and brands.

How consolidated is North American auto retail?
NADA counted 16,990 franchised light-vehicle dealers in the United States for 2025. Automotive News reported the top ten groups taking 11.1% of US new-vehicle sales in its 2026 Top 150 analysis, and Kerrigan Advisors recorded 478 dealership transactions in the twelve months to March 2026, released 15 June 2026.

What should a dealer group ask a vendor about exiting the platform?
What comes back, in what format, how long it takes and what it costs, answered separately for configuration, audit records and applicant history. A vendor unwilling to price and schedule a full extraction before signature has answered the question, and this applies to an incumbent as much as to a new vendor.

---

Related reading: the F&I office as the bottleneck, mapping the dealer-lender handoff, what saved floor time is worth.

Carousel builds intake and verification infrastructure, including white-labelled deployments for dealer groups. See the auto intake suite