How to Respond to Cars.com Leads in Under 60 Seconds

This is the playbook we use at the dealerships running our AI Closer. It’s broken into three phases: measure, automate, optimize. Each phase has concrete steps you can do this week.

Phase 1: measure where you actually are (1 hour)

You can’t improve a number you haven’t measured. Most dealerships dramatically misestimate their Cars.com response time — usually in the optimistic direction.

Step 1.1: pull the last 90 days of leads from Cars.com Dealer Portal

Log into your Cars.com Dealer Admin → Leads → Export. Pull a CSV with: lead arrival timestamp, dealership response timestamp, lead status. If your portal doesn’t expose response timestamp, pull it from your CRM.

Step 1.2: calculate response time per lead

For each lead, response time = (first outbound message timestamp) − (lead arrival timestamp).

If the first outbound is a generic auto-reply (“Thanks, we’ll be in touch!”), exclude it. Real response time = first personalized reply.

Step 1.3: bucket the results

Build a histogram with these buckets: under 1 min, 1-5 min, 5-30 min, 30 min - 1 hr, 1-4 hrs, 4-12 hrs, 12-24 hrs, over 24 hrs.

You’re looking for two things:

• What % of leads got a response in under 5 min? Industry top quartile is ~9%. Below that is the norm.
• What’s the median response time? Industry median is 17 hours. If you’re under 1 hour you’re already top quartile.

Segment by day of week and hour of day. The gap is almost always at nights + weekends.

Step 1.4: calculate the dollar value of the gap

Pull your 90-day close rate on leads that got a response within 5 minutes vs >1 hour. The gap is usually 2-3×. Multiply (5-minute close rate − >1-hour close rate) × (number of leads currently >1 hour) × (your average front-end gross). That’s the number on the table.

Example. 200 leads currently over 1-hour response, 9% gap in close rate, $4,500 gross = $81K of gross profit you’re leaving annually per rooftop. Most dealerships’ number is higher.

Phase 2: automate the response (1 day)

This is the actual work. Three options ranked by impact.

Step 2.1: pick the response channel

SMS is the highest-leverage starting point because:

• Cars.com leads include a phone number 95%+ of the time.
• SMS open rates are 98% within 3 minutes; email is 20% within an hour.
• Buyers reply to SMS at 6-9× the rate of email.

Email and webchat come second. Voice is its own project (different post).

Step 2.2: pick the AI platform

Your three real options:

1. Purpose-built dealership AI (Synthevo, etc.) — Cars.com ADF integration, DMS query for live inventory, dealership-trained tone. Setup 30-60 min. ROI typically positive month 1.
2. Generic AI chatbot (Drift, Intercom, ManyChat) — broader features, weaker dealership integration. Setup 1-2 days. ROI marginal because tone and inventory accuracy are weak.
3. Build your own with OpenAI/Claude API + your dev team — most flexible, slowest to ship. 6-12 weeks before you have something production-grade.

For most dealerships, option 1 is the right call. The build-vs-buy math doesn’t favor build at this price point.

Step 2.3: configure the response template

The AI replies need to do four things, in order:

1. Acknowledge the specific vehicle — “Yes, the 2022 Civic SE you asked about is still here, $19,495.”
2. Answer the buyer’s actual question if there is one — “Mileage is 28K, one previous owner.”
3. Suggest a concrete next step — “Want to come see it at 11am, 2pm, or 6pm today?”
4. Sound like a person — contractions, short sentences, no “Thank you for your interest!”

Most platforms ship with reasonable defaults. Tune them by reviewing the first 50 conversations the AI runs and editing the system prompt where it goes off-tone.

Step 2.4: set up escalation rules

The AI shouldn’t try to:

• Negotiate price
• Hold a vehicle without a deposit
• Commit to specific financing terms
• Discuss trade-in valuations beyond a soft ballpark

When the conversation crosses any of these triggers, hand off to a BDC rep with full context. Most platforms have these rules pre-configured; verify they match your dealership’s policies.

Step 2.5: deploy in shadow mode for 7 days

Don’t flip the AI live to customers immediately. Run it in shadow mode (it generates replies, but the BDC reviews and sends manually) for the first 5-7 days. You’ll catch tone issues, factual errors (the AI quoting the wrong price), and policy edge cases before they reach customers.

After 7 days of shadow mode, audit the AI’s replies. If 95%+ are sendable as-is, flip live.

Phase 3: optimize the loop (ongoing)

Once the AI is live, three weekly habits:

Step 3.1: review the worst conversations weekly

Pick the 10 conversations with the lowest engagement (buyer didn’t reply, or replied negatively). Read them in full. The AI’s first reply is usually where things went wrong. Update the system prompt or templates accordingly.

Step 3.2: A/B test the opening line monthly

The first reply has the highest leverage on conversion. Try variants — “Hey [name]!” vs “Hi [name],” vs no greeting at all. Acknowledging the vehicle first vs greeting first. Suggesting a specific time vs offering options. Run each variant for 200 leads and compare reply rate.

Step 3.3: feed the platform new policies as they change

When your dealership runs a new promotion, changes business hours for a holiday, or updates trade-in policy, the AI needs to know. Most platforms have a “knowledge base” or “policy” surface for this. Update it the same day the policy changes.

What good looks like, after 90 days

A dealership running this playbook end-to-end at 90 days typically sees:

• Median response time: 22 seconds (down from 47 minutes).
• % of leads responded to in under 5 min: 99% (up from 9%).
• Booked appointments per 100 leads: 24 (up from 11).
• Close rate on internet leads: 11% (up from 7%).

Net effect: ~50% more cars sold from the same lead volume.

What to do today

Try the AI yourself before you do anything else in this guide: request access to our live demo. We’ll text you from a real number in ~30 seconds. That’s not a screenshot, not a slide deck — that’s the actual AI we run for 50+ dealerships, replying to a stranger’s text in real time.