-
This thread is just the tip of the iceberg.The people ahead of the curve aren't Googling for answers — they're already in here, having the conversations you haven't found yet. DealerRefresh is free.Get the full picture →
Is PPC a window to all visitor behavior?
- By LarryBruce
Ok my first thought is that this post is more about the types of conversions than why the visitor converted or how to get more of them to convert. According to a study by TMP Directional and ComScore 46% of all lead conversions happen on the phone (http://www.dmconfidential.com/blogs/column/Search_Engines/2446/), and a recent MIT study in 2009 showed that a call is 22 times more likely to get a sale than any other type of conversion (http://www.leadresponsemanagement.org/lrm_study). So it shouldn’t be by accident that calls are respectively 3 to 1 and 20 to 1 that should be by design. As a matter of fact when we build out landing experiences for PPC campaigns we intentionally push a call over a form submission for all the reasons above and in this post.
Additionally, I'm not sure I would have used new, used and service as my segmentation buckets for this test. It shouldn’t be shocking that a new and used visitors both behave similarly on your site, the funnels for finding a new or used car and converting for each are exactly the same so in the end the visitor has no choice but to behave the same way, that’s how you designed it.
This brings in my biggest contention with website providers today, ALL OF THEM! There is very little a dealership can do to differentiate themselves in any website provider’s platform, particularly in the most important areas, the listings page and the details page because these are shared across every dealer in the platform, as Alex S would put it “Equal Incompetence”.
“The page view difference in this study is a result of less used cars of the same type or model to choose from. that’s not behavior that’s logistics.” Very, very few visitors are coming to a dealerships site to build a car. By the time they get to the dealers site they have already decided what they want to buy and it is now a matter of do you have it, how much is it and how do I get it.
You did however hit on the top 2 things that a visitor is looking to do on a dealerships main website:
1. Find a car
2. Schedule a service appointment
“DO” is the operative word here. The problems you should be solving on a dealerships website are ones that help visitors do the things they want to do, faster.
As for conversion, well I think I would want to see this broken out by branded vs. non branded terms (dealer name vs. high value non-name related terms), sales and service. A blended 12% rate isn’t good in my opinion, we want to see branded terms on a website convert at 30% plus. NOW, let me caveat that with we don't have a dealership doing that yet, my company is new and the oldest client I have right now is 60 days old. 30% plus is our goal and this is why I believe it is possible. The user typed in your name into a search engine their clear intent was to find your dealerships website. The only reason they don't convert is because you don't give them what they need to do so.
We believe we can achieve this through a series of off main site pages or microsites and our behavioral targeting technology to intercede at certain pages in the main website and/or redirect to offsite pages that focus on what we can determine as user intent. By doing this we can also split test the offsite pages and offers in real time, something not possible with any website company I know of today. I say this because I do not believe anyone can get to a 30% plus conversion rate without a clear test driven strategy, if you're just guessing you’ll never get it.
“18% for branding (name related key terms), I wouldn’t call that a pretty big number. I would call that a little less than average number for the same reasons I stated above.”
As for buying your name, first thing I would say is “keywords don't convert landing do”. With that said I would ask to see that data. Are the search terms the same? Is the landing the same? Were both the organic listing and the paid listing on page one and was the organic listing in the same position respectively to the paid listing (ex: paid listing in position 2 and organic listing in position 2)? If not that would skew your results of the test dramatically. I would wager there are some differences that would cause this to happen that have nothing to do with the ad being a sponsored link. Again you have to go back to user intent which is clearly to find that store and by human nature they are going to complete their intent unless there are some dramatic circumstances that stop them from doing so.
Last but certainly not least “Deep Linking vs Off page landing” this is actually the softball in this whole post. YOU NEVER, EVER WANT TO SEND PAID TRAFFIC BACK YOU YOUR RANDOM ACCESS MAIN WEBSITE…EVER!
1. With deep linking you still have all of the site navigation that will distract your customer out of the conversion funnel lowering you conversions.
2. When you deep link you are assuming that one page is going to do it we have very very very few landing experiences that are one page most are 3 some as many as 10 pages. With a deeplink or a single page you are assuming a one size fits all mentality and this post started out dispelling that notion immediately. Let me add to this before someone brings it up, a dynamic landing page where you are changing the picture or a header isn’t a custom landing that is closely connected to your promise. That’s just perfuming the pig; it doesn’t do a damn thing for conversion.
3. There is absolutely no way to split test a deeplink page with in any providers website right now for a lot of the reasons I stated above in this comment and if you can spit test you can’t get high conversions or continuous improvement.
Those are the top three there are many more but t this comment is already too long.
Well there you have it, be careful what you ask for you just might get it!
Hope that helps, of course I am sure that there are questions or clarifications on what I have said here everyone knows I am certainly willing to answer them and help where I can.
Additionally, I'm not sure I would have used new, used and service as my segmentation buckets for this test. It shouldn’t be shocking that a new and used visitors both behave similarly on your site, the funnels for finding a new or used car and converting for each are exactly the same so in the end the visitor has no choice but to behave the same way, that’s how you designed it.
This brings in my biggest contention with website providers today, ALL OF THEM! There is very little a dealership can do to differentiate themselves in any website provider’s platform, particularly in the most important areas, the listings page and the details page because these are shared across every dealer in the platform, as Alex S would put it “Equal Incompetence”.
“The page view difference in this study is a result of less used cars of the same type or model to choose from. that’s not behavior that’s logistics.” Very, very few visitors are coming to a dealerships site to build a car. By the time they get to the dealers site they have already decided what they want to buy and it is now a matter of do you have it, how much is it and how do I get it.
You did however hit on the top 2 things that a visitor is looking to do on a dealerships main website:
1. Find a car
2. Schedule a service appointment
“DO” is the operative word here. The problems you should be solving on a dealerships website are ones that help visitors do the things they want to do, faster.
As for conversion, well I think I would want to see this broken out by branded vs. non branded terms (dealer name vs. high value non-name related terms), sales and service. A blended 12% rate isn’t good in my opinion, we want to see branded terms on a website convert at 30% plus. NOW, let me caveat that with we don't have a dealership doing that yet, my company is new and the oldest client I have right now is 60 days old. 30% plus is our goal and this is why I believe it is possible. The user typed in your name into a search engine their clear intent was to find your dealerships website. The only reason they don't convert is because you don't give them what they need to do so.
We believe we can achieve this through a series of off main site pages or microsites and our behavioral targeting technology to intercede at certain pages in the main website and/or redirect to offsite pages that focus on what we can determine as user intent. By doing this we can also split test the offsite pages and offers in real time, something not possible with any website company I know of today. I say this because I do not believe anyone can get to a 30% plus conversion rate without a clear test driven strategy, if you're just guessing you’ll never get it.
“18% for branding (name related key terms), I wouldn’t call that a pretty big number. I would call that a little less than average number for the same reasons I stated above.”
As for buying your name, first thing I would say is “keywords don't convert landing do”. With that said I would ask to see that data. Are the search terms the same? Is the landing the same? Were both the organic listing and the paid listing on page one and was the organic listing in the same position respectively to the paid listing (ex: paid listing in position 2 and organic listing in position 2)? If not that would skew your results of the test dramatically. I would wager there are some differences that would cause this to happen that have nothing to do with the ad being a sponsored link. Again you have to go back to user intent which is clearly to find that store and by human nature they are going to complete their intent unless there are some dramatic circumstances that stop them from doing so.
Last but certainly not least “Deep Linking vs Off page landing” this is actually the softball in this whole post. YOU NEVER, EVER WANT TO SEND PAID TRAFFIC BACK YOU YOUR RANDOM ACCESS MAIN WEBSITE…EVER!
1. With deep linking you still have all of the site navigation that will distract your customer out of the conversion funnel lowering you conversions.
2. When you deep link you are assuming that one page is going to do it we have very very very few landing experiences that are one page most are 3 some as many as 10 pages. With a deeplink or a single page you are assuming a one size fits all mentality and this post started out dispelling that notion immediately. Let me add to this before someone brings it up, a dynamic landing page where you are changing the picture or a header isn’t a custom landing that is closely connected to your promise. That’s just perfuming the pig; it doesn’t do a damn thing for conversion.
3. There is absolutely no way to split test a deeplink page with in any providers website right now for a lot of the reasons I stated above in this comment and if you can spit test you can’t get high conversions or continuous improvement.
Those are the top three there are many more but t this comment is already too long.
Well there you have it, be careful what you ask for you just might get it!
Hope that helps, of course I am sure that there are questions or clarifications on what I have said here everyone knows I am certainly willing to answer them and help where I can.
Is PPC a window to all visitor behavior?
Every customer who arrives at your site has a unique experience. If you don’t take the time to understand the multitude of possible experiences on your site, you might be missing enormous amounts of opportunity. Ask yourself a simple question: “What are the different problems I can solve for my customers?” You’ll find out quickly that the many solutions you offer your clients are not found on your site. This is the basic premise behind search personas. To read more on search personas, check out Marketing in the Age of Google.
Paid search allows us to segment data to better observe these funnels of customers and understand how they behave differently, therefore justifying the need to treat them differently. We separated Used, New and Service customers and dove into the data.
We looked for high-level indicators of behavior: Page views per visit and Calls per form submission. Page views:
Used and New customers behaved very much alike, visiting 2 and 3 pages per visit respectfully and landing solidly at 3 calls per form submission.
To be fair, the difference between new and used shoppers seems pretty small. But we’re talking an average of one less page view over a full month’s worth of clicks to all of the dealers who use our paid search. That’s pretty significant. I think the explanation for the difference is pretty simple. There are a lot of extra tools and resources for the new car customer. Build and research tools and reviews are definitely contributors here.
Service customers came in at 2 page views per visit and a whopping 20:1 call to form ratio.
The service customer showed their preference for the phone loud and clear. This is a confirmation of our gut instincts, but it’s certainly good to see it on paper. We are seeing that service customers only view 2 pages per visit.
Now the number that everyone talks about is conversion. We looked at conversion across all of the keywords in the new, used and service departments:
12% is a good number, I was pleased to see that. There is more to this story though.
There is a set of keywords that cannot be attributed to a specific profit center. Things like the dealership name fit into this category and these keywords tend to do a lot of heavy lifting… they drive a lot of traffic. Because searchers who use dealership name cannot be deep linked to a profit center, the page views per visit goes up to 5. Calls per form rises to 5:1 as this is a grouping of sales and service customers.
The interesting story unfolds with a look at conversion on dealership terms however:
18%. This is a pretty big number in the conversion world. Given what we know about used and new sales performance though, I think there is a story here. Customers do their research, move up and down the purchase funnel considering different makes/models etc. During this process they are taking in advertisements from all of the local dealers. When it’s time to make the leap to leaving a lead what do they do? They Google the Dealership name, head to the site and contact the dealership. Proof.
There are certainly other stories to be told here and I would love to hear other perspectives on what folks take from this data.
What does this mean to the dealer: Advertise on your name, regardless of the medium. Too many people ask “Why should I bid on my name in paid search, we’re #1 in organic?”
Answer: Yes, the number of people who click on the dealership name in PPC is much less compared to the number of people who click on the organic listing. However, the conversion rate on the people who clicked the PPC ad is 4 times higher. To take this a step further, PPC is a quick and easy way to capture all those customers who may misspell your dealership's name. Google does a good job of trying to correct searchers' mistakes, but they're not perfect.
I can only say all of this because PPC gives us the most accurate view of our visitors' behavior. Is it a clearer window to your visitors? I would say yes.
Paid search allows us to segment data to better observe these funnels of customers and understand how they behave differently, therefore justifying the need to treat them differently. We separated Used, New and Service customers and dove into the data.
We looked for high-level indicators of behavior: Page views per visit and Calls per form submission. Page views:
Used and New customers behaved very much alike, visiting 2 and 3 pages per visit respectfully and landing solidly at 3 calls per form submission.
To be fair, the difference between new and used shoppers seems pretty small. But we’re talking an average of one less page view over a full month’s worth of clicks to all of the dealers who use our paid search. That’s pretty significant. I think the explanation for the difference is pretty simple. There are a lot of extra tools and resources for the new car customer. Build and research tools and reviews are definitely contributors here.
Service customers came in at 2 page views per visit and a whopping 20:1 call to form ratio.
The service customer showed their preference for the phone loud and clear. This is a confirmation of our gut instincts, but it’s certainly good to see it on paper. We are seeing that service customers only view 2 pages per visit.
Now the number that everyone talks about is conversion. We looked at conversion across all of the keywords in the new, used and service departments:12% is a good number, I was pleased to see that. There is more to this story though.
There is a set of keywords that cannot be attributed to a specific profit center. Things like the dealership name fit into this category and these keywords tend to do a lot of heavy lifting… they drive a lot of traffic. Because searchers who use dealership name cannot be deep linked to a profit center, the page views per visit goes up to 5. Calls per form rises to 5:1 as this is a grouping of sales and service customers.
The interesting story unfolds with a look at conversion on dealership terms however:
18%. This is a pretty big number in the conversion world. Given what we know about used and new sales performance though, I think there is a story here. Customers do their research, move up and down the purchase funnel considering different makes/models etc. During this process they are taking in advertisements from all of the local dealers. When it’s time to make the leap to leaving a lead what do they do? They Google the Dealership name, head to the site and contact the dealership. Proof.There are certainly other stories to be told here and I would love to hear other perspectives on what folks take from this data.
What does this mean to the dealer: Advertise on your name, regardless of the medium. Too many people ask “Why should I bid on my name in paid search, we’re #1 in organic?”
Answer: Yes, the number of people who click on the dealership name in PPC is much less compared to the number of people who click on the organic listing. However, the conversion rate on the people who clicked the PPC ad is 4 times higher. To take this a step further, PPC is a quick and easy way to capture all those customers who may misspell your dealership's name. Google does a good job of trying to correct searchers' mistakes, but they're not perfect.
I can only say all of this because PPC gives us the most accurate view of our visitors' behavior. Is it a clearer window to your visitors? I would say yes.
