Map the automotive customer journey and you almost always get the same slide: awareness, research, c…
The Automotive Customer Journey Breaks at the System Boundary, Not at the Design
Map the automotive customer journey and you almost always get the same slide: awareness, research, configuration, test drive, purchase, service, repurchase. Clean, plausible, and rarely the actual problem. In automotive retail, the journey doesn't fail because a phase is missing or because the interface is badly designed. It fails at the handovers between organizations and systems. The brand site belongs to the manufacturer, the configurator hangs off the manufacturer's product logic, availability and the actual sale sit with the dealer network, and the parts business runs in a third system with its own storefront. Every one of those handovers is a point where the customer starts over: new navigation, new search, new cart, often a visibly different brand.
This post maps where those breaks actually sit, why they're under extra pressure in 2026, and what role the frontend layer plays. It's not a sales strategy, it's the architecture question underneath one.
The Sales Model Itself Is in Motion
The reason this is urgent right now: the industry hasn't settled who actually sells the car.
In December 2025, Volkswagen announced it is ending the agency model for electric vehicles across Europe and returning to a franchised dealer network with dealers setting their own prices; the agency model stays in place for fleet business (electrive, ecomento). BMW is moving in the opposite direction, shifting its core brand's European sales onto the agency model with Mini as the blueprint, converting passenger car sales in 2027 (auto motor und sport).
Two major European manufacturers, two opposite calls, within months of each other. That matters for the journey because the sales model determines who sets the price, who signs the contract, and who owns the customer relationship. Build that logic into a storefront and you'll be pulling it back out at the next model change. The practical consequence is unglamorous but expensive to ignore: the sales model belongs in the configuration layer, not in frontend code.
Four Breaks That Are Specific to This Industry
1. Brand site to configurator. In many setups, the configurator is its own product with its own rendering logic, its own data source, and its own look. Moving from the emotional brand world into a technically dominated selection tool is the first visible break, and the most expensive one, because concrete purchase intent forms right there.
2. Configurator to dealer network. A finished configuration is worth nothing until it meets availability, a location, and a real offer. That's exactly where the manufacturer context ends in many journeys and a dealer site begins, with a different frontend, a different data state, and different scheduling logic.
3. New vehicles to aftersales. Service, accessories, and parts are the recurring business, but they usually run in a separate storefront, often on a different backend. From the customer's side it's the same brand. From the system's side it's two worlds.
4. Brands and markets. Automotive groups are multi-brand, multi-market organizations. Maintain a separate frontend per brand and per country and every fix, every accessibility requirement, and every campaign change multiplies.
These four have one thing in common: they aren't design mistakes. They're the visible output of a system landscape carved along lines of responsibility rather than along the customer's path.
What the Market Data Suggests
Two data points help with prioritization.
First, willingness to buy online. According to the Automobilbarometer 2026 from Consors Finanz, 60 percent of German respondents can imagine buying a car entirely online, from research through to delivery. At the same time, willingness to buy a new car purely digitally has slipped over time, from 65 percent in 2022 to 63 percent in a recent survey (kfz-betrieb). Openness is high, but it isn't growing on its own. It depends on how well the digital path actually works.
Second, the parts business. Online aftermarket parts trade is growing considerably faster than new vehicle sales. Analyst estimates diverge, landing somewhere between roughly 93 and 130 billion US dollars globally for 2026; The Business Research Company puts the move at 111.97 billion US dollars in 2025 to 130.24 billion US dollars in 2026 (The Business Research Company via GII, Mordor Intelligence). The spread says a lot about how differently these markets get scoped; the direction is consistent across all of them.
The sober reading: vehicle purchase stays hybrid for the foreseeable future, initiated online and frequently closed offline. The predictable, recurring online volume sits in aftersales and parts. If you're allocating digital budget, serve both, but don't treat a fully online new car checkout as the only goal worth funding.
The Frontend Layer as the Part That Stays Stable
When the sales model, the backend landscape, and the dealer structure all move at different speeds, you need a layer that absorbs those differences instead of being rebuilt every time one of them shifts. That's the job of a decoupled frontend layer.
For an automotive setup, that means:
- One frontend, multiple backends. The OEM commerce system, dealer data, and the parts shop feed the same presentation layer instead of each bringing its own frontend. Laioutr supports over 50 backends (source: laioutr.com/en/why-laioutr), including the systems common in group environments; for SAP-based landscapes there's a dedicated entry point on headless frontend for SAP Commerce Cloud.
- One component library across brands and markets. Multi-brand and multi-market run off the same library, so a fix is made once and live everywhere. That's the core of the multi-brand and multi-market capabilities.
- Configurator and checkout inside the same architecture. Instead of an embedded third-party system, the selection flow runs in the same component model as the rest of the page. How we frame selection and checkout flows is on the page for configurators and checkout flows.
- Frontend quality as a platform property. Median LCP in live frontends is 1.2 seconds, measured as a Q2 2026 field-data median (source: laioutr.com/en/why-laioutr). For media-heavy configurators that isn't a side note, it's the condition for anyone finishing the flow at all.
Architecturally, this is the frontend slice of a Composable Digital Experience Platform: the systems underneath stay what they are, the customer-facing view is composed above them.
Two Business Models, Two Sets of Requirements
Automotive isn't one use case. The consumer side around vehicles, accessories, and service follows retail logic with a local dimension: availability at the nearby dealer, appointment booking, pickup. The patterns in the multichannel retail growth kit are the right starting point there.
Parts and workshop business is B2B: customer-specific pricing, fast entry by part number, quote requests, recurring orders. That's what the components in the B2B growth kit address, and the direct-to-business angle is covered on the page for B2B wholesale and manufacturer direct sales.
Run both models on the same frontend layer and you configure them differently rather than building them twice.
Drawing the Line: What This Isn't
This isn't a general omnichannel argument. How touchpoints stay consistent across channels is covered in Omnichannel Customer Journey, and what AI-mediated journeys change for commerce teams is in The AI-Mediated Customer Journey. The automotive case is its own thing because the break doesn't run between channels, it runs between legal entities: manufacturer, dealership, parts trade. No personalization feature repairs a journey where the selling organization changes halfway through. That's an architecture and governance question first, a design question second.
Our Take
The automotive industry has discussed the customer journey as a sales topic for years while treating it technically as a set of separate projects: a configurator here, a dealer portal there, a parts shop in a third budget. As long as the frontend layer gets rebuilt per project, every change to the sales model lands as a rework ticket in all three at once. The alternative is undramatic: decouple the presentation layer once, leave the systems underneath where they are, and treat the sales model as configuration rather than as an architecture decision. Given that two of Europe's largest manufacturers are currently turning their models in opposite directions, that's less a bet on the future than an insurance policy.
FAQ
Why isn't modernizing the configurator enough? Because the configurator is rarely where the journey drops off. The drop happens at the handover right after it, when a configuration has to meet availability and a real offer in the dealer network. A better configurator with no connected next step just moves the problem downstream.
Do we have to replace our commerce backend for this? No. The whole point is that the backends stay. The frontend layer sits above them and talks to the existing systems, and the backend remains replaceable later without rebuilding the frontend.
What about dealers running their own websites? That's the norm, and it isn't a contradiction. What matters is whether dealer sites run on the same component base and use the same data sources. Then brand presentation and data state stay consistent without taking independence away from the dealers.
Is the effort worth it if most deals close offline anyway? Especially then. The digital part is the lead-in, and it decides how prepared and how confident someone is when they walk into the dealership. On top of that sits the parts and service business, which recurs online in a predictable way.
Next Steps
If you're working out how to bring the brand site, configurator, dealer network, and parts commerce onto one frontend layer, book a 30-minute demo. We'll walk the four break points against your actual setup.
More From the Laioutr Platform
About the author: Marcel Thiesies is Co-Founder of Laioutr and works on how frontend architecture absorbs organizational boundaries instead of passing them on to the customer.