Correlation between sales and website traffic

[Daniel J. Mondello]

How does the math play out Joe I'm not remembering my Statistics class in college at all (probably because I failed it 3 times before passing with a C- how's that for ego humbling Joe).

113,250 (users) / 1700 (sold vehicles) = 1.5 (2014 total car sales for one dealership YTD)

So is it # of sales (a) / number of Visitors (b) * 1000 or =(A/B)*1000 equals number of sales out for every 1000 unique visitors? or 15.01

76,000 (users) / 1520 (sold vehicles) = 2 (2013 total car sales for same dealership YTD)

# of sales (a) / number of Visitors (b) * 1000 or =(A/B)*1000 equals number of sales out for every 1000 unique visitors? or 20

Am I running this correctly?

 

[Daniel J. Mondello]

Dan,

Wow! Sometimes your ego just blasts out of your body and takes over your keyboard! You're writing in circles, I'm not sure what you said. I think you're crossing up definitions and you're surely not reading the thread.

#1).


#2).

Let's start here:

Correlation | Define Correlation at Dictionary.com
"...Statistics. the degree to which two or more measurements show a tendency to vary together"


re: #1.
The group has found so far: THE UNIQUE VISITOR TO SALES RATIO IS .79% to 2% So, if you had read, you'd find that in most cases the correlation between website traffic and sales is on either side of 100:1.


re #2.
We're all stating there is a correlation.


I've added that logic leads me to believe that the correlation is "loose" or elastic. Bill's awesome table illustrate that theory.

...so I reread the thread... You're right, my bad.

 

[Rick Buffkin]

 

[rlamb]

But as a dealer whose advertising is heavily invested in Digital I can tell you months where my group has over 1 million page views are typically much more successful sales months than the ones where the traffic was incrementally smaller.

Keep telling yourselves there's no correlation between Dealer Website traffic and sales and I'll keep laughing while I hand you your ass every month without fail.</story>

What I'm getting at is how do we know that your "1 million page views" influenced more sales rather than the market being strong at those times and therefore more people were searching for cars as a whole and you sold more cars.

The reason I am asking is my website traffic, page views, VDP views, email leads, etc. are all up significantly YOY and continuing to trend up. That said, our sales, while they are up YOY, they are not up anywhere near as much as our website traffic. So...I'm getting a TON more traffic to my site but only a slight increase in sales. Which has me wondering, what is the value of traffic to the site?

I get more web traffic in June then I do in January and I sell more cars in June than January, but that doesn't mean that web traffic causes sales obviously (again, I think lot traffic would probably be a better metric to use instead of sales). So, again, other than bragging about how much site traffic you can bring to a site, how can you demonstrate the value of this traffic?

 

[Daniel J. Mondello]

When you say Traffic are we talking sessions, users, or pageviews or all of the above. Sounds like you may have reached critical mass meaning you are running at your peaks in terms of internet closing, website lead provisioning, etc. What is your SEO and SEM like? Are you optimized to for more eyeballs or more conversions. Are you advertising way beyond a reasonable distance to drive and do business? These are some questions I would be asking in your situation.

 

[joe.pistell]

Dan,

Let me say the almost no one in my memory "get's it" like you do. I am proud to be considered a mentor. Your passion for "the kill" has no equal (& you can't teach that shit

That being said, your strength is your weakness. As you mature, your awesome desire to win will be improved as you become wiser.

Getting to your thoughts... Yes, your calcs are right. IN 2013, that store SOLD 1.5% of all unique visits. In 2014, that store sold 2% of all visits. 2% is a KILLER NUMBER!!!

 

[Stefan]

Toyota



Lexus-36 per 1K
Mazda KIA-28 per 1K

I think this just validates what many have said: so many variables here. I wish I could have a true count for the lot traffic (we all know how difficult is to log the ups).

 

[joe.pistell]

I wish I could have a true count for the lot traffic (we all know how difficult is to log the ups).

Efforting on that Stefan


Consider if data scientists can deploy an algo over 1 million images and break out "cats wearing hats", an automated solution to count ups can't be too far out.

More as it comes...

 

[joe.pistell]

Toyota

View attachment 2155

Wow, BIG numbers Stefan! Toyota, 2014 YTD comes in at 3.3% Thats one sale for every 33 visitors.


Back of the napkin math, if PPC = $4 p/visit and PPC costs $4, all you need is 33 visits for a sale. $4 * 33 = $132 cost per sale!




THIS IS A HIGH LEVEL LOOK ONLY. It's intended to be a "thought starter" for deeper analysis. For example, if your store is big in service, you'll have a lot of visits that never look at cars (reducing your ratio). A sharper analysis would breakout out "visitors exhibiting shopping behavior" vs sales.


Gotta run to another meeting!!

 

[joe.pistell]

Also, Look at that AMAZING jump in web traffic from 2013 to 14 (+57%), leads +15%, Lot Ups +10%, and sales increased roughly +8%.

Killer study Stefan. Very interesting indeed!