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

Website Conversion Counting

Jason Petkov

Green Pea
Jun 6, 2013
22
3
First Name
Jason
How is everyone tracking website conversions? I have been on the fence about using a few different techniques. Here are a few things that I am considering using to factor some more realistic results.

1.) Excluding all unique visitors that have NOT been been on SRP or VDP.
2.) Track the time on site and page views of users that HAVE submitted a lead and count the visitors that match or above the average but have not submitted a lead as a lead or half lead.
3.) Possibly count visitors that have been to the site more than once as 1/2 a lead.

I would use the above 3 methods on a quarterly sample and not a monthly.

I know these numbers would show improvement over current tracking, and that's not what I am looking to accomplish. I want something that's going to provide better ratios for A/B testing and more accurate results, proving that it's not all about the "Click". I don't want the reports to turn into Cobalt's convoluted soft leads high funnel, low funnel marketing CYA reports.

Cobalt tracks people that visit your hours and directions pages, and lumps in your service/parts leads with sales leads. They are extremely graceful at manipulating the results of their websites performance especially when you compare them to the end of the month CRM sales results. And not to forget how every month they are "excited" they are to share the results with you!

So what are your methods? What do you count? And most importantly how are you using your info to grow and improve?
 
  • 1Like
Reactions: Brad Burlingham

Jason Petkov

Green Pea
Jun 6, 2013
22
3
First Name
Jason
I did see some of the other ones, but nothing was recent. I've always found it bad forum etiquette to resurrect old material. If I missed a current one my apologies, feel free to close the thread.
 

csabatka1

Rust & Dust
Jan 7, 2013
149
82
First Name
Chad
What I found is everyone has their own measurement and vendors especially skew that measurement to their benefit. You're best developing your own standard measurement and don't deviate from that measurement.

For example here's what I use for our Lead Conversion Rate on our Website(s) and mobile site - I seperate mobile from desktop numbers:
Shoppers - Unique Visitor who visits at least one inventory page
Lead - Fresh (careful including these), Phone (unique tracking on sites), or Form Submission
Form Lead - Exactly that, form submission from VLP or VDP
Phone Lead - Phone call from tracking number
Solds - Website Lead who purchased

To Calculate:
Lead Conversion Rate = Leads/Shoppers
Lead To Sold Rate = Solds/Leads
Shopper To Sold Rate = Solds/Shoppers

When Comparing numbers, I only compare like quarters, like months to previous year, and previous month, maybe two months back. One thing I've notice is your numbers will fluctuate throughout the year and then trend like other years. Comparing like months will help eliminate yearly trends. You will have outlier months - like January of 2012 we had 70+ degree temps vs January 2013 we had 1'+ snow and cold temps. Make note of those in your analysis.
 

joe.pistell

Uncle Joe
Apr 7, 2009
4,076
1,585
First Name
Joe
1.) Excluding all unique visitors that have NOT been been on SRP or VDP.
2.) Track the time on site and page views of users that HAVE submitted a lead and count the visitors that match or above the average but have not submitted a lead as a lead or half lead.
3.) Possibly count visitors that have been to the site more than once as 1/2 a lead.

Jason,

Here's an outline of my analytics model that I ran at my last store. I wanted the "Scoring system" to give weight to the our customers behavior. HTH.

What % of buyers sent in a lead prior to purchase? (pretty standard stuff here)

Next, because shoppers are over'whelmingly stealth, I wanted to know:

What % of shoppers were at your site prior to purchase?
is it 20% is it 80%? (It will likely be >70%).

This is an important number that few decision makers can visualize. The idea is to let all players over you to know the importance of your efforts and how your efforts can add more sales without shoppers sending in a lead.

Additionally, you'd like to ask your buyers, "how many times were you at our website prior to purchase?"

Now, you have a numerical "value" of the lead and a value to the damn invisibile web shopper that drives us all mad.

If this interests you, the one page post sale survey I used is here in the DR forums.. somewhere ;-) I ran nearly 20,000 sales thru it and here are some stats to help you see what I was looking at

http://www.usedcarking.com/

  • 85% visited the site prior to purchase
  • 70% of these we're at the site 3 times or more prior to purchase
  • Result= 60% of all sales were at the dealer's website 3x or more prior to purchase.
  • Sales from internet leads (not phone ups) = 22% of all sales

I did not have time to deep dive leads behavior (break out leads by # of visits). If anyone has this, I'd love to give that a look :)

So... The delivery survey can give you extra visibility into that damnable invisible shopper and their impact to your org's bottom line. The survey showed me that I needed to give different weights to different points of measurement. For me, it was quite clear that cultivating a killer multiple visitor experience was where the ROI was.

Leads are cool too! ;-)
Consider looking into building a lead profile, where you'll know the favorite paths a shopper takes to each lead source. Let's entertain a hypothetical example: The "sweetspot" for a Tradein form could be: >2 visits, seen >6 Used VDPs or 4 New VDPs, has NOT filled out any other forms and 40% come in after store hours.


HTH
Uncle Joe
 
Mar 23, 2013
28
2
First Name
Chris
This is something I am also interested in getting a set formula to analyze. I agree with you Jason. I caught on to Cobalt's tactics long ago for over inflating their website results. I feel they are getting over on less educated people analyzing the data. Anyways, sorry for posting a little off topic. Now that I know I'm not alone maybe I will spend more time coming up with something based off of others suggestions. My initial thought is I think that one could use Cobalt's tool to lay a foundation for your formula and then make your own modifications. Like you mentioned, h & D page views shouldn't be considered in the conversion ratio. At least not in my opinion. However, I do think it does have some significance I think. Kind of like ad prints on cars.com or autotrader. I would like to see a study on h&d visits and ad prints and the relevance of these actions to a dealers bottom line. Is the relevance so small that it's not significant? Are the vendors using these actions to only inflate their conversions? I would like to think that you can attribute these actions to sales every month. The question is...what is the average number of customers that actually bought after performing one of these actions on your site. If we knew this number then maybe we could incorporate these actions into our conversion ratio.
 

ideveroux

Boss
Feb 13, 2010
9
4
Awards
1
First Name
Ian
How is everyone tracking website conversions? I have been on the fence about using a few different techniques. Here are a few things that I am considering using to factor some more realistic results.

1.) Excluding all unique visitors that have NOT been been on SRP or VDP.
2.) Track the time on site and page views of users that HAVE submitted a lead and count the visitors that match or above the average but have not submitted a lead as a lead or half lead.
3.) Possibly count visitors that have been to the site more than once as 1/2 a lead.

I would use the above 3 methods on a quarterly sample and not a monthly.

I know these numbers would show improvement over current tracking, and that's not what I am looking to accomplish. I want something that's going to provide better ratios for A/B testing and more accurate results, proving that it's not all about the "Click". I don't want the reports to turn into Cobalt's convoluted soft leads high funnel, low funnel marketing CYA reports.

Cobalt tracks people that visit your hours and directions pages, and lumps in your service/parts leads with sales leads. They are extremely graceful at manipulating the results of their websites performance especially when you compare them to the end of the month CRM sales results. And not to forget how every month they are "excited" they are to share the results with you!

So what are your methods? What do you count? And most importantly how are you using your info to grow and improve?
My Golden Rule is "there is no such thing as bad data". COUNT EVERYTHING.
1. Don't exclude customers that didn't hit your SRP or VDP. Use that data to find out why. Why didn't they goto my SRP or VDP? What did they find, or not find, that kept them from these pages?
2. There's no such thing as a half lead. If their time on site was too short, dig into the data and find out why (excluding bounces of course). What brought them to the site (source) and what were they looking for that they left so quickly. Was my page load time too long when clicking between pages, etc...
3. Unique users data should filter out those that have visited the site more than once.

Lastly, don't rely on your website vendor to provide the best data. You gotta roll up your sleaves and dig for it. Google Analytics and your CRM are great tools to get what you need.