"It is not the strongest or the most intelligent that will survive but those that can best manage change."

—Charles Darwin, Naturalist

Dealers have spent years competing for clicks. Now they are competing for the recommendation. Adobe's newest AI traffic data shows shoppers are arriving from AI sources with more focus, more intent, and more value, which puts dealership websites in a very different kind of spotlight.

Traffic referred by AI sources to U.S. retail sites rose 393% YOY in Q1 2026. They also converted 42% better, stayed 48% longer, viewed 13% more pages per visit, and drove 37% higher revenue than non-AI traffic.

Those numbers should grab any retailer’s attention.

If AI is becoming a discovery layer for shopping, then dealer websites need to be legible to both people and machines. Inventory pages need clean trim data, transparent pricing, payment context, and accurate merchandising.

Service pages need plain-language explanations, real menu pricing, and scheduling paths that do not bury the lede. Fixed ops content, F&I explainers, trade-in pages, and FAQ sections all become part of how AI tools understand and recommend a store.

1️⃣ Optimize for AI-driven discovery.
AI assistants are becoming a new front door for shoppers. Dealers with clear inventory data, trustworthy pricing, strong page structure, and useful reputation signals give those tools more to work with and create stronger visibility in recommendation-driven shopping.

2️⃣ Turn your website into a decision engine.
Every major page carries more weight when shoppers arrive with higher intent. SRPs, VDPs, service pages, trade tools, and finance content perform best when they deliver clear details, helpful context, and obvious next steps that move shoppers closer to action.

3️⃣ Connect the full customer journey.
AI-assisted shopping flows across research, trade, finance, and service in one continuous path. Dealers that align those touchpoints create a more complete digital experience, stronger signals for AI systems, and more opportunities to capture high-intent traffic across the store.

Anthropic just shipped a powerful new pairing: Claude Opus 4.7 and Claude Design.

Opus 4.7 is their new general-availability model, and Anthropic is positioning it as a stronger engine for long-running, detail-heavy work, especially coding, agent workflows, finance analysis, and high-resolution visual tasks.

The more interesting release for many teams is Claude Design.

It sits on top of Opus 4.7 and turns Claude into a visual collaborator for prototypes, slides, one-pagers, mockups, and marketing assets. You can start from a prompt, upload docs, refine with inline comments and sliders, then export to Canva, PDF, PPTX, or standalone HTML.

Anthropic also says it can read a team’s code-base and design files to build a reusable design system, then pass finished work straight into Claude Code for handoff.

Put simply, Claude is getting better at turning rough thinking into presentable work. For dealership teams juggling events, promos, decks, and mockups, that kind of momentum goes a long way.

Have you ever wondered why your chatbot tends to lose the plot after a while? One minute it is tracking perfectly, and the next it is forgetting key details, missing nuance, or giving answers that are confidently incorrect.

That usually comes down to context limits. AI works within a fixed amount of conversational memory, often called a “context window.” Once you approach that limit, the AI starts losing track of what was written at the top. And, it can't scroll back up.

The tricky part is knowing when you've hit the limit. Most platforms won't tell you, they’ll just automatically consolidate everything, which is why it's worth getting into the habit of regularly asking the chatbot to summarize everything important that's been discussed.

Try this prompt:

Then do not just blindly trust the summary. Review it. The model is good at spotting patterns, but you are still better at knowing which details are make-or-break. If something is missing, ask, "What did you leave out and why?" That question is sneaky useful because it forces the model to show you how it decided what mattered.

Later, when the chat starts getting crowded, you can drop that file back in or start a fresh conversation with it. Same context, cleaner workspace, better odds the bot stays on track.

AI is already embedded in most dealership stacks, but very little of it is being governed with the same rigor as the rest of the business.

Mackenzie Wiltrout gets into what actually separates signal from noise, where the effectiveness of AI comes down to how well it’s anchored to trailing data, how clearly success is defined, and how tightly outputs map to decisions that impact revenue.

She digs into how purpose-built models outperform generic tools when they’re trained on dealership-specific inputs, and how teams that treat AI as a diagnostic layer, not just a production layer, start to uncover patterns that shift strategy across marketing and operations.

Watch the full interview and see how top operators are tightening the loop.

Thanks for reading, Friend! Enjoy the wins, but remember to stay humble.