Most dealers review vendors when something goes wrong. A bad month, a blown budget, a rep who stops returning calls. By then you have already paid for the problem. A better approach is a fixed cadence. Every 90 days, you pull the same numbers, ask the same questions, and make the same kinds of decisions. It turns vendor management from a fire drill into a routine, and it puts you back in charge of the scoreboard.
Why 90 Days
A month is too short to judge a marketing channel fairly. Cars take time to move, and one slow week can swing a small sample. A year is too long to let waste run. Ninety days is enough to see a real trend in cost per car sold, and short enough that a bad vendor cannot coast for long. It also lines up with how most dealers already think about the quarter.
There is a practical reason too. A shopper who first found you in January may not sign until March. Judge that source in January and it looks dead. Judge it across the full quarter and the picture is fair. Ninety days is long enough to let those slow-cooking deals close and land on the source that actually started them.
What to Pull
Gather the same data every quarter so you are comparing like to like. At a minimum:
- Total spend by vendor and channel. Gross dollars, before any co-op offset.
- Sold cars attributed to each channel. From your DMS, on one consistent definition, not from vendor dashboards.
- Cost per car sold by vendor. The headline number for the whole review.
- Close rate and gross per unit by source. So a cheap channel selling junk does not out-rank a pricier one moving real metal.
- Quarter-over-quarter trend. Direction matters as much as the current number.
The close rate line is the one dealers skip and regret. Say a lead site delivers three hundred leads at a low cost per lead, but your team closes them at four percent, while your own website leads close at twelve. The cheap source is not cheap once you count the desks, phones, and follow-up it burns to sell one car. Pulling close rate and gross alongside spend is what stops a low sticker price from hiding a high real cost.
What to Measure Against
Rank every vendor on cost per car sold and return by dollar. Put them in the same table, on the same math, and the picture gets clear fast. Some vendors will carry their weight. Some will be expensive but defensible because of the gross they bring. And some will have no honest answer for what your money bought. The goal is not one perfect metric. It is a fair, side-by-side comparison you control.
Picture two vendors side by side. One runs a cost per car sold of four hundred dollars but the units come in at eighteen hundred front gross. The other is three hundred a car but the gross averages nine hundred. The pricier vendor is the better buy, and only a table that puts spend, units, and gross in the same row will show you that. The number in isolation lies. The comparison tells the truth.
What to Decide
A review that ends without decisions is just a meeting. Each quarter, every vendor lands in one of four buckets:
- Scale. Strong cost per car sold and room to grow. Move more budget here.
- Hold. Working fine, no reason to change. Leave it alone and watch the trend.
- Fix. Underperforming but worth a conversation. Set a specific target and a deadline, in writing.
- Cut. No honest case for the spend. Give notice and move the money.
Tell each vendor which bucket they are in and why. The good ones will respect it. The rest will tell you a lot by how they react. A vendor you put in the Fix bucket with a clear target either comes back next quarter with better numbers or comes back with excuses. Both answers are useful, and both are ones you engineered instead of waited for.
How to Make It Stick
The review only works if it happens on schedule whether or not anything is on fire. Put four dates on the calendar for the year. Assign one person to own the data pull. Keep the format identical every time so the comparison is clean. And build the numbers on data you own, from your DMS and your own spend records, so no vendor can move the goalposts between reviews.
Keep the old reviews in one folder. A year in, you can lay four quarters side by side and see which vendors trend up, which drift down, and which have quietly doubled their cost per car while the rep kept smiling. That folder is worth more at renewal time than any pitch a vendor will ever hand you.
Run it four times a year and vendor accountability stops being a confrontation and becomes a habit. The vendors earning their keep get rewarded, the ones coasting get caught, and your budget goes to the channels actually selling cars.
Frequently asked questions
How long does a quarterly review actually take?
Once the format is set, the data pull is an hour or two and the decisions are another hour. The first one takes longer because you are building the table from scratch. After that you are updating numbers you already know how to find, which is why keeping the format identical every quarter matters so much.
What if a vendor’s numbers look great one quarter and terrible the next?
That is exactly what the trend line is for. One strong quarter can be a seasonal push or a lucky batch of leads. One weak quarter can be a slow month for the whole store. Do not scale or cut on a single data point. Watch two or three quarters and act on the direction, not the spike.
Should I share the cost per car sold number with the vendor?
Yes, when you want them to fix something. A vendor cannot hit a target it never saw. Showing your number turns a vague complaint into a specific ask, and it tells the vendor your store runs on data. The ones worth keeping will use it to improve. The ones who argue with your own DMS are telling you which bucket they belong in.
These are the numbers you should own, not rent from a vendor. See how VCTRS helps dealers hold every vendor accountable on our vendor accountability page.