• Stop being a LURKER - join our dealer community and get involved. Sign up and start a conversation.

Dan Sayer

Boss
Staff member
Dec 4, 2009
476
518
Awards
7
First Name
Dan
I'm back in retail after being on the vendor side (which I think everyone in retail should do at least once and vice versa) for nearly two years. Prior, I spent 14 years on the retail side mainly in Internet/BDC/eCommerce/Dig Ad. Around 2010, I was struggling how to measure ROI of classified vendors [insert joke here]. Leads didn't make sense so started focusing on their cost per VDP. Back then I was holding AutoTrader to under a dollar and Cars.com to $.85 and really used that through 2016. During that time, if they wanted a package up-sell I needed to see that cost stay the same or lower for it to make sense and ultimately keep the product.

Fast forward to 2018 and it looks like my stores (same group I was at prior) are running just below $.70 per VDP view with both AutoTrader and Cars.com (CarGurus is quite a bit lower per). My question to you is do you use Cost Per VDP as a KPI and if not, what do you use in 2018? Analytics/Referrals, Leads, Actions on site, etc?
 
I'm back in retail after being on the vendor side (which I think everyone in retail should do at least once and vice versa) for nearly two years. Prior, I spent 14 years on the retail side mainly in Internet/BDC/eCommerce/Dig Ad. Around 2010, I was struggling how to measure ROI of classified vendors [insert joke here]. Leads didn't make sense so started focusing on their cost per VDP. Back then I was holding AutoTrader to under a dollar and Cars.com to $.85 and really used that through 2016. During that time, if they wanted a package up-sell I needed to see that cost stay the same or lower for it to make sense and ultimately keep the product.

Fast forward to 2018 and it looks like my stores (same group I was at prior) are running just below $.70 per VDP view with both AutoTrader and Cars.com (CarGurus is quite a bit lower per). My question to you is do you use Cost Per VDP as a KPI and if not, what do you use in 2018? Analytics/Referrals, Leads, Actions on site, etc?
Hey Dan, yes it's still a KPI that we look at and hold our third party listing sites accountable to.

Below are 2 example slides with fake sample data showing the KPIs we look at for our third party listing sites.

My digital marketing team has a dedicated performance manager that travels around to each of our stores and presents on all of our different vendors in an unbiased way and makes recommendations to help improve our ROI / increase sales.

PS: We're using Google Sheets & Data Studio to create this centralized reporting system.

third party slides 1.JPG

autotrader slide 2.JPG

slide 3.JPG
 
Last edited:
Interesting, @reverson. Nice reports. To clarify, are the "VDP Views" numbers in your reporting reflective of referral traffic VDP views on your sites or are you pulling in the VDP Views counts from the 3rd party sites? I was holding them accountable to the View counts on the 3rd party site.
 
Interesting, @reverson. Nice reports. To clarify, are the "VDP Views" numbers in your reporting reflective of referral traffic VDP views on your sites or are you pulling in the VDP Views counts from the 3rd party sites? I was holding them accountable to the View counts on the 3rd party site.
Correct, we are only counting the VDP views on the third party listing sites.

We take the 3rd party sites’ reports and manually input the data into Google Sheets which Data Studio then pulls from. It only takes us a couple hours each month to do for all of our stores so we haven’t really felt the need to look into creating a custom API data connection.

Website referral clicks are included in the metric called engagement which also includes saving a vehicle, printing the vdp, etc.
 
  • Like
Reactions: Dan Sayer
Great stuff, thanks @Dan Sayer for asking the question, and I always sit up in my chair when @reverson shares his analytical approach. I don't expect all to agree with my approach, but for the 3rd party classifieds (TPCs), I measure:

- Cost per VDP
- Cost per lead (lead/chat/text/call)
- VDP/SRP ratio (how often you "win the click")
- VDP/lead ratio (how often the platform generates hand-raisers from VDP)
- Cost per sale (only as good as the CRM and in-store processes)

Now when I'm consulting with dealers, and representatives of the TPCs are in the meeting, they are quick to point out, "George, the lead counts you're tracking are not the ONLY leads/showroom traffic we're providing the dealers." To which I say, you are 100% correct. However, that effect is felt evenly across all TPCs, and by measuring these metrics and comparing them across all TPCs, I am able to compare apples to apples. It is then up to the dealer to decide, are they overpaying one provider vs. another? Or do they feel the exposure is worth it, and they are willing to hold their nose?

I also like measuring, and discussing with the dealership their SRP to VDP ratio, as I feel it is meaningful. Maybe so many years at Dealer Specialties and presenting to dealers on the importance of merchandising, and "winning the click", which I still think is critically important. This number tends to range from 0.8% to 2.8% (big swings). CarGurus tends to have a lower VDP/SRP ratio vs the other big-2. I attribute that to the fact that CG's SRPs display more vehicles per page than the others. Carfax.com has the highest VDP/SRP ratio by a long shot for most dealers, I don't have a definitive opinion on why, perhaps their shoppers are further down the funnel?

Finally, the VDP lead ratio is interesting, how often the VDP is compelling enough to generate lead. This varies by provider, but in general Cars.com is higher than ATC or Gurus, and by far the #1 is Edmunds with a VDP to lead ratio that is 3x-5x larger than any of the TPCs. @Ian Isch could share his opinion, but I would guess it is the way they entice the shopper to submit lead information in order to unlock prices.

My sample set is small, I work with under 20 dealers today, but I can tell you of all the monthly reporting and analytics I provide the dealers, they love the TPC data the most.
 
What, exactly, is the purpose of your third party spend? Is it supposed to build brand awareness of your dealership? Is it to drive traffic to your website? Is it to generate leads in the form of Calls, Texts, Forms? Is it to SELL CARS?

Or is it to shut up Old Bob, your archaic used car manager who insists it works without any data?

One of my biggest gripes with Third Party Inventory sites is the way they block dealer access to data in an attempt to prevent dealers from seeing their true performance. If they were serious about transparency, then they would allow dealers the ability to download historical performance data in a csv format. Not partial data. Not for the last month. Not in a formatted Excel or PDF file. But all historical performance data.

None of them do this because if they did, dealers would drop them like a bad habit or demand price reductions.

Another of my pet peeves is the way they use other lead gen methods and advertisers to generate leads/actions that they roll up into their reports as "engagements." The perception is that a car shopper goes to one of their sites, searches for a car, sees a SERP, clicks on a dealer's car that generates a VDP, and then converts into some kind of lead. But, I've seen way too many cases where these third party sites use other parties, like Craig's List, to generate phone calls or leads. Or they will use retargeting ads to generate leads. In my opinion, this is deceptive. In my experience the quality of these "leads" sucks.

But to your point, is Cost Per VDP still relevant?

Yes and no. Its a measure, and sometimes you have to use what's at your disposal. So from that standpoint, it's relatively useful to compare performance between Third Party Vendors, and as a benchmark for performance so when your rep comes in and tries to raise your rates you have some sort of rational basis for making a decision. But beyond that, it's questionable.

The truth is it's way too easy to manipulate top level KPI's. Increasing the number of VDPs and lowering the Cost Per VDP means nothing if it doesn't translate to your bottom line as sales.

Unfortunately way too many GMs still use the mindset that, "if it sells me one or two more cars a month, it's worth it."

And given the nature of OEM Co-op programs and the political pressure marketing managers are under, often times it's just not worth the hassle of cutting this dead weight from your budget and reallocating the spend to more productive uses. Especially if the only real alternatives you have are OEM approved Co-op SEM vendors.
 
  • Useful
Reactions: Alexander Lau
Great stuff, thanks @Dan Sayer for asking the question, and I always sit up in my chair when @reverson shares his analytical approach. I don't expect all to agree with my approach, but for the 3rd party classifieds (TPCs), I measure:

- Cost per VDP
- Cost per lead (lead/chat/text/call)
- VDP/SRP ratio (how often you "win the click")
- VDP/lead ratio (how often the platform generates hand-raisers from VDP)
- Cost per sale (only as good as the CRM and in-store processes)

Now when I'm consulting with dealers, and representatives of the TPCs are in the meeting, they are quick to point out, "George, the lead counts you're tracking are not the ONLY leads/showroom traffic we're providing the dealers." To which I say, you are 100% correct. However, that effect is felt evenly across all TPCs, and by measuring these metrics and comparing them across all TPCs, I am able to compare apples to apples. It is then up to the dealer to decide, are they overpaying one provider vs. another? Or do they feel the exposure is worth it, and they are willing to hold their nose?

I also like measuring, and discussing with the dealership their SRP to VDP ratio, as I feel it is meaningful. Maybe so many years at Dealer Specialties and presenting to dealers on the importance of merchandising, and "winning the click", which I still think is critically important. This number tends to range from 0.8% to 2.8% (big swings). CarGurus tends to have a lower VDP/SRP ratio vs the other big-2. I attribute that to the fact that CG's SRPs display more vehicles per page than the others. Carfax.com has the highest VDP/SRP ratio by a long shot for most dealers, I don't have a definitive opinion on why, perhaps their shoppers are further down the funnel?

Finally, the VDP lead ratio is interesting, how often the VDP is compelling enough to generate lead. This varies by provider, but in general Cars.com is higher than ATC or Gurus, and by far the #1 is Edmunds with a VDP to lead ratio that is 3x-5x larger than any of the TPCs. @Ian Isch could share his opinion, but I would guess it is the way they entice the shopper to submit lead information in order to unlock prices.

My sample set is small, I work with under 20 dealers today, but I can tell you of all the monthly reporting and analytics I provide the dealers, they love the TPC data the most.


Thanks, @georgenenni for the tag. This is a very interesting topic as in my previous life a long time ago I work at Cars.com (through Gannett) and the Cost Per VDP was a KPI we pushed and was a great strategy to move away from a "lead" conversation. At Edmunds, you will see less VDPs but our VDPs are currently designed to achieve a lead form with the "Reveal Price" option. Our goal is to prove our value with hard conversions and then the Cost Per VDP and Engagement are all additional value. We are always testing, and trying to find the right balance.

Truly impressed with @reverson and his team with how detailed they are looking at digital marketing. I think these metrics are important to gauge the value which always depends of course on how much your spend is with any vendor.
 
I love the executive summary that @reverson is preparing for his stores!

@georgenenni -- If I remember correctly, my old boss, Dale Pollak used to say that SRP/VDP ratio is useful when tracking performance on a specific third-party over time, but isn't nearly as useful when comparing multiple third parties. Let's say you have three different listing sites, one has 20 vehicles per page, another has 25, and yet another has 50. A customer clicks into one VDP on the first results page of each site; you then get a 5% click rate, a 4% click rate, and 2.5% click rate, respectively. Dale always referred to the VDP as the 'money metric' while click rate was something to monitor, just not as important.

Not all VDPs are necessarily created equal. Another metric to consider watching might be Engagements per 100 VDPs (using Ryan's definition of engagement). More engagement per VDP would signal more engaged shoppers -- much like more time on site and more pages viewed on your own site might signal a more engaged shopper. One site might yield a cheaper VDP, but with far less engagement. Just a thought.
 
I love the executive summary that @reverson is preparing for his stores!

@georgenenni -- If I remember correctly, my old boss, Dale Pollak used to say that SRP/VDP ratio is useful when tracking performance on a specific third-party over time, but isn't nearly as useful when comparing multiple third parties. Let's say you have three different listing sites, one has 20 vehicles per page, another has 25, and yet another has 50. A customer clicks into one VDP on the first results page of each site; you then get a 5% click rate, a 4% click rate, and 2.5% click rate, respectively. Dale always referred to the VDP as the 'money metric' while click rate was something to monitor, just not as important.

Not all VDPs are necessarily created equal. Another metric to consider watching might be Engagements per 100 VDPs (using Ryan's definition of engagement). More engagement per VDP would signal more engaged shoppers -- much like more time on site and more pages viewed on your own site might signal a more engaged shopper. One site might yield a cheaper VDP, but with far less engagement. Just a thought.

Good points @ed.brooks , though one point on your "...5% click rate, a 4% click rate, and 2.5% click rate...", that would be assuming every vehicle on the SRP belong to that dealer. Since SRPs are made up of varying dealers, I'm not sure the number of vehicle per page matters as much. I agree that VDP is one of the money metrics, but you gotta get that shopper to choose you first.