Re-targeting and Cars.com version of Re-targeting

[Jason Petkov]

Dealers are going to double the price they are spending on Cars.com for banner ads? Banner ads have a CTR of somewhere between .07 down to .03 of a percent.

The banners are not on cars.com. They are across an ad network displayed to users that have visited cars.com and searched for a vehicle. I would look into adjusting creative if you're only getting a CTR of .07 to .03 seems extremely low.

 

[ddavis]

The banners are not on cars.com. They are across an ad network displayed to users that have visited cars.com and searched for a vehicle. I would look into adjusting creative if you're only getting a CTR of .07 to .03 seems extremely low.

Jason, I understand where the ads are. That is .03 to .07% which is fine if you are Coke or Taco Bell but if you are a single point dealership, that is a problem. Change the creative? When was the last time you clicked on a banner ad?

I went back to the Dataium study commissioned by Cars.com. I remember it mentioned banner ads.

"Other digital advertising such as social media and local market banner advertising: 8.5% of total budgets; 20% of digital"

It only discussed the amount dealers spend on this but nothing about effectiveness. I'm shocked. They were not shy about slamming SEM.

Reading that study, if you do the opposite of what the "average" dealer is doing, you might be on the right track.

 

[csabatka1]

I have run several retargeting ads on both Yahoo! and Google networks in the automotive and other industries and have seen similar lackluster results in all.

A recent Google Network retargeted display ad for automotive.
May 2013
33,563 Impressions
93 Click-Throughs .277%
1 - Conversion

A point that you have to remember is Google will run your display ad on free Android Apps downloaded from Play Store.

The highest click-through rate I've seen from retargeting campaign was a 0.7%.



For $2000-$3500/month you're better off spending your money elsewhere.

 

[Jason Petkov]

I have run several retargeting ads on both Yahoo! and Google networks in the automotive and other industries and have seen similar lackluster results in all.

A recent Google Network retargeted display ad for automotive.
May 2013
33,563 Impressions
93 Click-Throughs .002%
1 - Conversion

A point that you have to remember is Google will run your display ad on free Android Apps downloaded from Play Store.

The highest click-through rate I've seen from retargeting campaign was a .07.

I am not sure if you added an extra number on the clicks or left one off the impressions? 33,563/93 = 0.277% CTR not .002%

Here is my approach:

1. I only use 300x250 and 160x600 and disable visibility mobile devices. (My past numbers have shown those are the best performing sizes so I stick with them. Also keeps you off of Play)
2. I usually start with at least 3 versions of the same banners with minor color changes and eliminate over 1-2 weeks based on CTR for that banner. Usually within 4-5 days you will see which banners need to go
3. Exclude poor performing placement sites weekly. <- This is key, you can weed out very quickly high impression / low click sites from displaying your ads which hold back your entire campaigns performance.
4. Add / remove targeted Zip codes as needed based on performance. DMA is a great easy way to target people, but it doesn't really show you how different zones within your DMA are performing and doesn't give you the ability to remove
5. Set frequency capping (This will help keep the unnecessary impressions down, and not annoy your past visitors.)

It takes me a little extra time, but the results have been worth it for me. I average around .4% - .8% CTR with average CPM between $8-11.

 

[isales]

Well if you take the medium of those averages to estimate how many vistors you will get off 50k impressions then take your average conversion ratio on your website and so on and so on....

those numbers don't seem to bad to me.

 

[ewalraven]

I'm not a huge fan of this stuff (remarketing) outside of doing it in-house with google, but it's hard to debate the value of doing it on other vendor networks without some solid numbers.

With that said, here's our numbers from last month (brag?):
Impressions: 19.5k
Clicks: 25
Conversions: 1

Anybody that has some success with it. I'd be interested in knowing.

 

[Jason Petkov]

Also, to clarify about ad networks, Cars360 serves our ads via an ad exchange that gives us access to Google, AOL, and Yahoo networks.

Hi Rock, thanks for chiming in and clarifying the networks that you will be serving the ads on. Do you by chance have an example results report that dealers will receive? And what power would dealers have over their campaigns? Could we for example eliminate sites which our ads show on, eliminate IP addresses from seeing them etc.?

 

[csabatka1]

I am not sure if you added an extra number on the clicks or left one off the impressions? 33,563/93 = 0.277% CTR not .002%

Here is my approach:

1. I only use 300x250 and 160x600 and disable visibility mobile devices. (My past numbers have shown those are the best performing sizes so I stick with them. Also keeps you off of Play)
2. I usually start with at least 3 versions of the same banners with minor color changes and eliminate over 1-2 weeks based on CTR for that banner. Usually within 4-5 days you will see which banners need to go
3. Exclude poor performing placement sites weekly. <- This is key, you can weed out very quickly high impression / low click sites from displaying your ads which hold back your entire campaigns performance.
4. Add / remove targeted Zip codes as needed based on performance. DMA is a great easy way to target people, but it doesn't really show you how different zones within your DMA are performing and doesn't give you the ability to remove
5. Set frequency capping (This will help keep the unnecessary impressions down, and not annoy your past visitors.)

It takes me a little extra time, but the results have been worth it for me. I average around .4% - .8% CTR with average CPM between $8-11.

Yes, the 0's got me in that post. .277% was supposed to be the calculation for May.

0.7% was highest I've seen.

 

[Rock Irvin]

Hi Jason. The reporting for Cars360 focuses in on the increasing audience and awareness for the dealership, by using the Cars.com audience.

We give metrics on how many shoppers are in your area (i.e. Buffalo), and how many of those shoppers have shopped the vehicle brand that you have targeted (i.e. Buffalo Ford shoppers). That is your "relevant audience" and we would only serve ads to folks shopping in Buffalo who have shopped Ford. We want to, first off, show you that we are targeting the most appropriate audience for our clients.

Through our awareness metrics we show the number people we served your ad to, your penetration into the "relevant audience", and the top dealerships those people were shopping as well as you, and instead of you. We, of course, show the impressions and impression frequency per person. The main key here is to, not only show that we served impressions, but also give insights on the audience and detail on how we have improved the dealership's reach within the relevant audience.

Lastly, we show insights on the number of people who took actions on Cars.com after seeing the ad.

We want our reports to mirror the value of Cars360, increasing a dealership's brand reach and frequency among the Cars.com audience.

 

[Chris Leslie]

These types of programs have been around for years. The idea of having ads follow you around the web in hopes of increasing CTR was a good idea 10 years ago. I've worked with all types of ad networks and publishers or years and here is what I see..

cars.com is working with the yahoo network which is called the "right media exchange" this ad network similar to googles "double click" are filling unsold ad space for millions of publisher sites.many of these sites probably have nothing to do with cars o anything rated to cars but, because cars is working with yahoos exchange the millions of publisher sites old theoretically be onside red partner sites.

When you work with the exchanges yourself you get options of day parting when your ads get shown, along with time parting, page placement preference and even frequency capping. You lose all that with his program.


Another misconception is that retargetting should yield higher ctr's.the reality is though it will always deliver less simply because he amount of total impressions increase at a rate not equal to clicks. So if you do need to measure this don't use the cost per click method. Rather, use the cost per thousand.

If you're thinking this program will sell more cars I can assure you it won't. Retargetting campaigns when used should direct to simple lanibg page in hopes of gathering a bit more info. Also, if you're running a cobalt site you need to make sure your landng pages are hosted elsewhere due o possible traffic loss from mobile devises via the auto re-direct cobalt has.

With so many options, variables and software out there cars.com shouldn't even be tempting to be any good at this part of digital marketing. I could spend a whole day explaining why you should steer cler of it as it will not produce what you want it to.

In short retargetting program are about impressions and getting as many as you can for as little as you can. Adding any addition KPI'Swill only discourage you. You're better off taking those dollars and using them for educating yourself on how you can build and Do this program in house.