Vespoke / decision infrastructure for car retail

The dealers who build their own software win the next ten years.

Every system a dealership already pays for, feeding one warehouse the group owns.

DMSdeals · service · GL CRMleads · notes · history SERVICE CRMappointments · check-ins WEBSITEclicks · forms · first-party MANUFACTUREROEM portals · allocation MARKET + APIScomps · recalls · book values /00 THE WAREHOUSE one customer. one record. owned by the group. TEXTS + EMAILoutreach, receipted DASHBOARDSone live picture ASKplain-English answers AUTOMATIONSwork that runs itself AI AGENTSthe data, safely AI PLATFORMfed real context

Six sources already paid for. One warehouse the group owns. Every output in its own definitions.

Built & running in 16 rooftops Live in production - not a slide deck Operator-built - by a working Director of Operations Month-to-month - the group keeps everything

Why now

You've seen this movie before.

Classifieds took nine years to lose 70% of a $19.6 billion channel. This shift is running the same curve in a fifth of the time.

% OF STARTING VALUE, BY YEARS SINCE THE SHIFT BEGAN

Newspaper classified revenue (2000 → 2018) AI compute cost, GPT-3.5-class (2022 → 2024)
0% 50% 100% 0 yrs 5 10 15 years since the shift began −69% in 9 years −89% in 18 −99.6% in 2 years

Last time the collapsing line was a revenue channel. This time it is the price of building software. NAA via Poynter · Stanford AI Index 2025.

The problem

The DMS, the CRM and the AI platform earn their keep. The other ten tools are the problem.

Behind them sits a long tail of point solutions - each built for one job, on its own contract, with its own copy of the data.

$29Kaverage monthly software spend per rooftop, including the DMSCar Dealership Guy
< 1 in 3dealers satisfied with the customer data they receiveCox Automotive
$42Ka year in integration fees to move that data between vendorsDealertrack

Why build

More software will not solve this. Building will.

The tools are not the expensive part any more. Coordinating them is - and those connections multiply far faster than the tools do.

CONNECTIONS THAT HAVE TO AGREE, AS TOOLS ARE ADDED

3 TOOLS 3 connections 6 TOOLS 15 12 TOOLS 66 Every connection is a reconciliation someone owns.

Double the tools and the reconciliation work more than quadruples. A warehouse removes it - the numbers settle once, in one place.

280×cheaper to run the intelligence under custom software2022 → 2024
$575K→$1Mrevenue per employee after Klarna rebuilt its stackin one year
35%of enterprises have already replaced a SaaS tool with software they build78% will build more

WHAT A SINGLE STORE SPENDS, PER YEAR

RENT — 8–12 tools + integration fees $78K–138K / yr BUILD — the layer underneath, once one-time, then yours

Rental recurs and rises. A build is capital the group keeps.

This isn't a strange idea

The best dealers already build their own.

Among the largest groups, the software cited as an advantage is the software they built.

OperatorSizeWhat they built
Asbury~150 storesClicklane - their own digital retailing tool, instead of renting one
Go Auto70+ rooftopsTheir own CRM, from scratch, fifteen years ago
Crain Automotive16 stores"Crain Vision" - DMS, payroll, expenses, open ROs and daily sales on one screen
SutherlinMulti-storeA service-lane appraisal engine, because no vendor would build it
One independentSingle store, $125MTheir own recon software - for less than a year of the subscriptions it replaced

Five years ago this was a billion-dollar-group move. All examples public, from the Car Dealership Guy podcast.

What we build

One warehouse. Fifteen systems. All in production.

Everything below is live in a 16-rooftop group today. It stacks in order - the warehouse first, then the layers it powers. Click anything to see the problem it kills, what it replaced, and the numbers.

FOUNDATION/00Warehouse

Latest impact // Barnes Crossing Auto Group // 16 rooftops

The group that stopped renting its own reality.

One owned warehouse. Fifteen production systems. 26,135 vehicles a year flowing through a data layer Barnes Crossing controls - built from the inside by the operator who runs it.

Engagement model

How the work is structured.

1 storePilot first, proven on the group's own numbers before wider rollout.
90 daysTo something live and in use - not a six-month discovery phase.
Month‑to‑monthNo multi-year term, no auto-renew. If it stops earning its keep, it ends.
RetainedOn exit the group keeps the code, the warehouse and the data.

The enemy is unaccountable spend, not vendors.

Start with the data.

A 45-minute working session. No deck. And the terms are part of the point: pilot one store first, month-to-month, and the group keeps everything we build.

Book a working session