- Nov 4, 2026
- 4
- 0
- Awards
- 1
- First Name
- Ibrahim
Closing % by source is the right lens, but a lot of buyers never submit a form at all, so every source looks worse than it is.

1. Manual attribution - the rep is giving credit to a source that may not be the actual source.
2. We're counting solds in the month they sold and leads in the month they were submitted. That can skew a close %. A q3 average view should be long enough to compensate for that though.
| Platform | Q3 Close Rate | Cost per Sale |
|---|---|---|
| Autotrader | 9.4% | $532 |
| CarGurus | 6.8% | $735 |
| Cars.com | 6.6% | $758 |
| Carfax | 4.7% | $1,064 |
| Edmunds | 4.0% | $1,250 |
| TrueCar | 3.8% | $1,316 |
| Overall | 6.4% | $781 |
Dealers discuss which third-party lead sources perform best, with Q3 close rate data showing Autotrader leading at 9.4% and TrueCar trailing at 3.8%. A key insight is that raw close rates are misleading without factoring in cost per lead — at $50 per lead, the average effective cost per vehicle sold across platforms runs around $781 — and attribution data is further distorted by single-source CRM systems, alphabetical drop-down bias, and customers naming big-brand sites they recognize rather than the actual ad they clicked. Practical advice includes using an LLM to analyze a full year of lead data before cutting any provider, and noting that high-funnel car shoppers browse third-party and OEM sites rather than dealership websites.