Client: Chevrolet / General Motors Partner: Autobytel Initiative: Digital commerce proof of concept Market:Washington, D.C.
“We sold more cars in the first day with the Chevy Showroom than we sold in our entire first month with GM BuyPower!” — Vice President of U.S. Sales and Service, General Motors
The signal GM couldn’t ignore
In 1999, General Motors launched GM BuyPower to transform how people bought vehicles online. It had serious investment, national promotion, and thousands of participating dealerships. What it didn’t have was a connection to how consumers actually wanted to shop.
BuyPower digitized the traditional automotive brochure. Customers could research vehicles, sure — and then they ran straight into the same old wall: unclear pricing, uncertain inventory, dealer negotiations, and no clean path to purchase. The technology worked as designed. The model didn’t.
By 2001, GM was considering a new digital marketplace built around actual inventory and real pricing. Before writing another large check, the company wanted evidence that a different approach could work. Autobytel was brought in to help develop and test that alternative.
Reframing the problem
The original question was “how do we sell vehicles online?” We reframed it: what information and control do customers need before they’re willing to take the next step?
That distinction changed the whole project. Instead of treating the website as another promotional channel, we treated it as a system for reducing uncertainty. Customers didn’t need more vehicle advertising. They needed to know which vehicles were actually available, what each one included, what the dealer had added, what the invoice and sticker prices were, what they’d really be expected to pay, and how to claim the one they wanted.
Our hypothesis was simple: transparency and real inventory would produce more buying activity than a polished but static online catalog.
Building the Chevy Showroom
Working closely with the GM team, I led development of a regional proof of concept called the Chevy Showroom. My team owned the product architecture, technology, user experience, interface design, and the inventory and pricing data systems underneath it all.
Every night, the platform imported detailed vehicle data from GM dealerships: configurations, options, packages, dealer-installed additions, invoice pricing, sticker pricing, and actual discounted pricing. For the first time, shoppers could search for a vehicle by the precise features they wanted instead of browsing generic model descriptions. They could find a real car on a real lot, review its exact equipment and pricing, print a virtual window sticker, see the discounted price, and reserve it at a participating dealership.
The Showroom didn’t put the existing dealership experience online. It removed the parts of that experience people dreaded.
The hardest part wasn’t the code
The biggest obstacle was organizational. GM came in with deeply held assumptions about how consumers would shop online and, especially, how much pricing information they should be allowed to see.
Rather than accepting those assumptions as requirements, we tested them against actual digital shopping behavior. We helped the GM team separate practices that protected the existing process from features that created value for the customer. That meant challenging conventional thinking, educating stakeholders, and dragging every decision back to the customer’s experience — sometimes more than once.
The goal was never to win the internal debate. It was to design a test that would let the market settle it.
A controlled test before a national bet
Instead of launching another national initiative, GM tested the Chevy Showroom in a single market: Washington, D.C. The pilot involved a handful of Chevrolet dealers and almost no paid promotion. Nearly all of its traffic came from a single link on Autobytel inviting visitors to try the new experience.
That set up a clean question. Could a transparent, inventory-driven experience in one region outperform a nationally promoted platform with vastly more brand reach, dealers, and media support?
It took about a day to find out.
The results
Within its first 24 hours, the Chevy Showroom facilitated more vehicle sales than GM BuyPower had generated in its entire first month.
Sit with that comparison for a second. BuyPower operated nationally, included thousands of dealerships across multiple GM brands, and had substantial media backing. The Showroom had one regional market, a few Chevy dealers, minimal promotion, and one traffic source.
GM extended the 90-day pilot to six months. Over that extended run, the Showroom sold more vehicles than BuyPower had sold since its launch.
GM ultimately chose not to expand the model, worried about competition and conflict within its dealer network. But by then the customers had already voted, and the core strategy was validated.
What the experiment proved
BuyPower’s problem was never a lack of investment, technology, distribution, or promotion. It was a mismatch between the experience GM had designed and the experience customers needed.
The Showroom won because it replaced internal assumptions with observable buyer behavior. It gave customers specificity instead of generic information, transparency instead of negotiation anxiety, actual inventory instead of hypothetical configurations, and a direct next step instead of a research dead end.
Most importantly, it gave GM evidence.
The adaptive marketing lessons
Years before I called it adaptive marketing, this project demonstrated principles that would later become central to The Adaptive CMO.
Follow the signal, not the playbook. BuyPower’s disappointing results weren’t just a performance problem. They were the market telling GM the underlying model was wrong.
Treat strategy as a hypothesis. We didn’t assume transparency and inventory access would work. We built a controlled test and let customers validate or reject the idea.
Build systems around real behavior. The platform connected live inventory, detailed vehicle data, pricing, search behavior, and dealership action — organized around the customer’s decision process, not GM’s org chart.
Use technology to accelerate learning. The tech mattered because it made a new experience testable. It didn’t just automate an existing process; it let GM discover what customers actually valued.
Measure outcomes that change decisions. Page views and traffic wouldn’t have proven anything. Vehicle reservations and sales did.
The lasting lesson
Large organizations tend to assume transformation requires a large launch. The Chevy Showroom demonstrated the opposite. One market, a handful of dealers, and almost no advertising taught GM more about digital automotive commerce than its national platform ever had.
That’s the advantage of an adaptive organization. It doesn’t need to predict the future perfectly. It builds systems that reveal what to do next.
Build Marketing That Gets Smarter as It Moves
The Chevy Showroom is an early example of the operating principles explored in my book, The Adaptive CMO: The Secret to Revenue, Relevance, and Results.
The book provides a practical framework for replacing static playbooks with marketing systems that recognize patterns, test assumptions, respond to buyer behavior, and improve continuously.