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

How much is a slow website costing me? $$$

Mar 21, 2012
1,041
1,360
Awards
10
First Name
Ryan
New year, new website?

Website speed is critically important, but how can you quantify the impact?

Here's how:

First, calculate how much ad spend you're flushing down the toilet:
1) Gather the click data from your Facebook and Google Ads.
2) Compare this to the session data from GA4 for the same campaigns.
3) Work out the discrepancy between ad clicks and actual website visits.
4) Calculate this as a percentage of your total ad clicks.
5) This percentage represents potential customers lost due to slow loading times.
6) Multiply this by your total ad spend to see the financial impact of wasted ad spend that never even makes it to your website.

Next, let's quantify the lost opportunities in terms of leads, sales, and profit:
1) Multiply the discrepancy in ad clicks and website visits by 2% to estimate lost leads.
2) Assume 10% of these leads could have become sales.
3) Multiply the lost sales by an average profit of $4,000 each.
4) Now, you have an estimate of the total gross profit lost.

If you want to really make yourself sick - annualize these figures by multiplying by 12. This will show you the potential yearly impact of a sluggish website. To play devil's advocate and be optimistic - divide these numbers in half and in half again, you likely may still become queasy.

Now don't get me wrong, it's natural to have some variance between ad clicks and site visits due to factors like accidental clicks or private browsing blocking UTMs. And it's important to keep in mind that ad clicks from people scrolling on Facebook are more likely to be less patient than people actively searching on Google.

However, you may begin to notice some trends, especially if you have multiple website providers. Some website providers are significantly faster than others. This exercise may even trigger you to consider changing website platforms.

So where should you focus your attention first? Your vehicle details page.

With a large percentage of ad units like VLA's and AIA's driving traffic direct to VDP, the VDP is by far the single most important page of your entire website. You want your VDP's to load fast, be optimized for conversions, and offer alternative recommended vehicles to maximize every ad click.

Send this over to your ad agency today and ask them to run these numbers for you. I'd recommend breaking it out by ad platform (Google/Facebook), destination type (VDP/SRP/Home), and website vendor. The results may surprise you.

Dealer margins are heading downwards for 2024. Dealers need to get ahead of the curve and begin to aggressively focus on cost reduction. This is one easy way to identify wasted ad spend and set yourself up for success in 2024.

Here's to a great 2024 everyone!
 
I know you mentioned this but I think the accidental clicks are a pretty decent number, say 40-50% of the clicks that don't register as sessions. The user closes the browser before anything renders. I'm not sure there's a way to separate this from the total. Although I completely agree with your premise part of this is unavoidable. Using your portal site as an example Google tag manager was the 3rd thing to render so the slowness you are referring could be from this not rendering right away on the site or not asynchronously. It important for every dealership to ensure their GTM renders right away, like Garber's, so you can get an accurate measurement.

Screenshot 2024-01-02 150432.png

Thinking about what causes website slowness I think it important to talk what causes this. Generally 4 categories:
1. The host 2. Base website 3. Third party code 4. User

As a dealership you can really only influence 3 of these and you may not be able to get your provider to change the host or their base website code, leaving #3 as your primary focus. This is also typically the main issue as the add-ons have implementation issues.
 
  • Like
Reactions: Ryan Everson
I know you mentioned this but I think the accidental clicks are a pretty decent number, say 40-50% of the clicks that don't register as sessions. The user closes the browser before anything renders. I'm not sure there's a way to separate this from the total. Although I completely agree with your premise part of this is unavoidable. Using your portal site as an example Google tag manager was the 3rd thing to render so the slowness you are referring could be from this not rendering right away on the site or not asynchronously. It important for every dealership to ensure their GTM renders right away, like Garber's, so you can get an accurate measurement.

View attachment 8361

Thinking about what causes website slowness I think it important to talk what causes this. Generally 4 categories:
1. The host 2. Base website 3. Third party code 4. User

As a dealership you can really only influence 3 of these and you may not be able to get your provider to change the host or their base website code, leaving #3 as your primary focus. This is also typically the main issue as the add-ons have implementation issues.
Yeah one thing that a dealer can't control (outside of switching website providers) is server response time. This takes place before even scripts, images, etc even begin to load.

1704227001176.png
So I'd encourage everyone to not just run a speed test on your homepage, but also on your SRP and VDP. The results can vary quite a bit.

But the speed test I trust most is the "Mississippi Rule." If I can count to 3 or 4 Mississippi's while waiting for a VDP to even begin loading on my gigabit internet, that's way too slow and without a doubt costing me business.
 
  • Like
Reactions: DrewAment
Thinking about what causes website slowness I think it important to talk what causes this. Generally 4 categories:
1. The host 2. Base website 3. Third party code 4. User

As a website vendor, we ended up building a tool that lets us toggle off all the 3rd party integrations so we can demonstrate to the dealer that, although these tools are often great, they are the reason your OEM is complaining about your page speed and they do have an impact on website experience.

If you can, implement vendor plugins after the page load and without shifting the contents of the page around.
 
  • Like
Reactions: Ryan Everson
Might be worth mentioning the bounce rate. This can often be a bit misleading. It wouldn't be unusual to get a higher bounce rate in the AM, and towards 5pm. A number of people visit your site simply to get the service phone number or address which are typically on the homepage. So much so, we've often disabled ads during peak service hours since people will google the store... click on our ad.. just to get the number and 'bounce'.
 
Might be worth mentioning the bounce rate. This can often be a bit misleading. It wouldn't be unusual to get a higher bounce rate in the AM, and towards 5pm. A number of people visit your site simply to get the service phone number or address which are typically on the homepage. So much so, we've often disabled ads during peak service hours since people will google the store... click on our ad.. just to get the number and 'bounce'.
Regarding bounce rate, I completely agree. Not really a worthwhile metric, outside of tracking your own trends per traffic source.

I wouldn’t consider the discrepancy in ad clicks and GA4 traffic a traditional bounce though. This means the user didn’t even do a 1-page “bounce”, they clicked the back button before the page even began to load.

Some website providers have a solid 3-4 second server response time, which means the GA4 script doesn’t have a chance to load before the user gets frustrated by the slowness and bails. Let’s refer to it as the bail rate %!
 
  • Like
Reactions: Todd Thompson
Yikes. I got back into Search Console for the first time since seeing this post and either a lot has changed in here or our sites have completely fallen apart. I've always been told that dealer sites notoriously rank poorly in a lot of the metrics that things like Search Console track. I'd be interested to see a dealer site with "good" Search Console data for reference. For instance, CWV says I have 85 poor URLs, 14 that need improvement, and 0 good ones. Should I be freaking out?