Practical AI workflows dealerships can use to automate follow-up, reviews, inventory merchandising, service campaigns, SEO content, and BDC operations.

Start with the free tools

Before you add another vendor, diagnose the foundation. Run Dealer Website Grader to find technical website leaks, then run DealerAEOAudit to see whether AI assistants recommend your store.

See all Dealer AI Guy tools →

Vehicle Listing Workflow

Generates unique, AEO-ready VDP descriptions from raw vehicle data so every car has its own copy instead of a templated paragraph.

Who it's for: Internet directors, merchandising managers, and dealer groups managing 100+ VDPs.

Tools needed:

  • ChatGPT or Claude
  • Inventory feed (CSV or API)
  • Your DMS / website CMS

Step-by-step

  1. Export vehicle data. Pull YMM, trim, options, mileage, color, and condition notes from the inventory feed.
  2. Run the description prompt. Feed each row into a structured prompt that returns 3 paragraphs: who it's for, standout features, why buy here.
  3. QA pass. Human reviews the top 20% of inventory by price. Spot-check the rest.
  4. Push to the site. Sync back via API or CSV import. Re-run weekly for new arrivals and price drops.

Example prompt

Write a VDP description for a 2022 Toyota 4Runner TRD Off-Road Premium with 38k miles in Magnetic Gray. 3 short paragraphs: who it's for, standout features in plain English, why buy from [DEALERSHIP] in [MARKET]. No stock phrases.

Use with ChatGPT →

Review Response Workflow

Drafts personalized responses to Google, Yelp, and DealerRater reviews — including the hard ones — within minutes of posting.

Who it's for: GMs, marketing directors, and customer experience leads.

Tools needed:

  • ChatGPT or Claude
  • Review monitoring tool (Podium, Birdeye, or manual)
  • Google Business Profile

Step-by-step

  1. Capture the review. New review hits → copy the text, star rating, and customer name into a tracked sheet or tool.
  2. Classify. Tag as: praise, fixable complaint, unfair, or factual error. Route negatives to a manager.
  3. Generate response draft. Use a department-specific prompt (sales vs service) that mirrors the customer's language and offers a clear next step.
  4. Human approves and posts. Manager reviews, edits if needed, and posts within 24 hours.

Example prompt

A service customer left a 2-star review saying: "[REVIEW]". Write a public response from the service manager under 120 words. Acknowledge the issue, take responsibility where appropriate, offer a specific next step, and include a direct contact.

Use with ChatGPT →

Service Campaign Workflow

Builds targeted recall, declined-service, and seasonal campaigns from DMS data — without renting a CRM module you don't need.

Who it's for: Service managers, fixed ops directors, and BDC leads on the service side.

Tools needed:

  • DMS export (CDK, Reynolds, Tekion, etc.)
  • ChatGPT or Claude
  • SMS/email platform

Step-by-step

  1. Pull the list. Filter DMS for customers with open recalls, declined services in the last 180 days, or vehicles due for seasonal work.
  2. Segment. Group by service type and vehicle age. One generic blast underperforms 4 targeted ones every time.
  3. Generate messaging. Use the declined-service or recall prompts to draft email + SMS versions per segment.
  4. Send and track. Schedule sends, capture appointments in the DMS, and report close rate by segment weekly.

Example prompt

Write SMS + email for customers who declined brake service on their [VEHICLE] in the last 90 days. Explain real-world risk in plain language, offer 2 specific time windows this week, and include one incentive option.

Use with ChatGPT →

Facebook Ad Workflow

Produces 3 ad variations per model line — value, lifestyle, trade-focused — without paying an agency $2k for a 'creative refresh.'

Who it's for: Marketing directors and dealers running paid social in-house.

Tools needed:

  • ChatGPT or Claude
  • Canva or Photoshop
  • Meta Ads Manager

Step-by-step

  1. Pick the model. Choose one model per ad set. Don't try to sell the whole store in one creative.
  2. Generate 3 copy variants. Use the Facebook ad prompt to produce value, lifestyle, and trade-in variations.
  3. Pair with native imagery. Real lot photos beat stock. Add the dealership name on the image, not just the caption.
  4. Test, kill, scale. Run 7 days, kill the bottom variant, scale the winner, refresh the loser monthly.

Example prompt

Write 3 Facebook ad variations for the 2024 Honda CR-V at [DEALERSHIP] in [MARKET]. Each variation: 1 hook line, 3 bullet benefits, 1 CTA. A: value. B: lifestyle. C: trade-in.

Use with ChatGPT →

SEO Content Workflow

Turns real buyer questions into AEO-ready blog posts that LLMs (and Google) can cite when shoppers ask.

Who it's for: Marketing directors, dealer principals, and anyone trying to show up in ChatGPT and AI Overviews.

Tools needed:

  • ChatGPT or Claude
  • Google Search Console or Semrush
  • Your CMS

Step-by-step

  1. Mine real questions. Pull buyer questions from sales chat logs, BDC notes, Google's 'people also ask,' and Reddit threads in your market.
  2. Draft with structure. Use the AEO blog prompt. Short answer up top, H2s phrased as questions, summary at the bottom.
  3. Add local proof. Mention your market, your dealership once, and one local fact (event, road, neighborhood) AI can attribute to you.
  4. Publish and resubmit. Publish, submit to Search Console, and check ChatGPT/Perplexity 2 weeks later to see if you're being cited.

Example prompt

Write a 600-word blog post answering: "Is it worth buying a certified pre-owned SUV in [MARKET]?" Use clear H2s phrased as questions, short answers up top, final summary. Mention [DEALERSHIP] once.

Use with ChatGPT →

Customer Follow-Up Workflow

Runs a 30-60-90 day follow-up sequence on every delivered customer — service intro, referral ask, trade check-in.

Who it's for: Sales managers, BDC managers, and GMs trying to lift repeat and referral business.

Tools needed:

  • CRM (VinSolutions, eLead, DealerSocket, etc.)
  • ChatGPT or Claude
  • SMS/email

Step-by-step

  1. Trigger on delivery. Every delivered deal kicks off a 3-touch sequence: day 30, day 60, day 90.
  2. Day 30 — service intro. Introduce the service advisor by name. Offer a free first-service appointment.
  3. Day 60 — referral ask. Ask for a referral with a specific reward. Make it easy: one tap to share a link.
  4. Day 90 — trade check-in. Soft trade-value touch. Don't pitch. Just keep the door open.

Example prompt

Write a day-30 follow-up SMS + email from the service advisor [NAME] to a customer who took delivery of a [VEHICLE]. Friendly, no upsell. Confirm the first service is on us and offer 2 time windows.

Use with ChatGPT →

Need the raw prompts first?

Browse the Prompt Library →