Every few weeks, a vendor shows up at a dealership with a screenshot of Google PageSpeed Insights. Big red “43.” Lots of frowny-face icons. Then the pitch: “Your site is slow. You’re losing shoppers. Let us build you a new one.”

And then the magical screenshot. The new site. A beautiful, glowing “99.”

Here’s the question nobody asks in that meeting:

Does the shopper actually notice?

Let’s do the math

A PageSpeed score of 99 maps to a page load of about one to two seconds. A score of 43 maps to about four to six seconds. The gap is roughly three to four seconds.

Three to four seconds. About ten blinks of an eye.

In a lab, that gap is enormous. In the hands of a vendor’s slide deck, that gap is “catastrophic.” On a mobile phone in your shopper’s living room, that gap is: “Huh, my internet’s being weird tonight.”

What a shopper actually thinks when a page is slow

Nobody shopping a $40,000 vehicle on their phone sits there with a stopwatch going “This dealership has failed to properly optimize their Largest Contentful Paint.”

What they actually think, in order:

1. “One bar? Really?”
2. “Verizon is garbage in this neighborhood.”
3. “Did I disconnect from Wi-Fi?”
4. “My phone needs an update.”
5. “Everyone must be streaming right now.”

Nowhere on that list is “this dealer clearly doesn’t invest in a modern tech stack.”

They don’t have that thought. They can’t have that thought. They don’t know what a modern tech stack is, and they don’t care.

The research everyone cites is for buying a $15 phone case

Here’s the dirty secret of the page-speed pitch. The famous “53% of mobile users abandon a page that takes more than three seconds” statistic? That was built on commodity ecommerce behavior. Low-intent sessions. Impulse purchases. Someone deciding whether to buy a travel pillow at 11pm.

A shopper researching a $40,000 vehicle is not that person.

That shopper is on page seven of a mental journey. They’ve been looking at trims. They’ve been comparing payments. They have a spouse in the room asking about cargo space. They have investment, consideration, and a credit score they are mentally pre-stressing about. They are not going to abandon that search because the Jeep Grand Cherokee page took four seconds instead of one.

They will blink. They will glance at their signal. They will keep going.

What they do abandon for

High-intent shoppers do leave sites. For real reasons. Specifically:

Prices that don’t match the ad. They came for a $27,988 Civic. They see $31,500 with “dealer-added accessories.” They leave.

No inventory on the lot. Three VDPs deep and every unit is “in transit.” They leave.

Payment calculators that return “call for pricing.” They leave, and they resent the dealership for wasting their time.

Lead forms that demand 14 fields. They leave halfway through.

A phone number that goes to a voicemail at 7pm on a Tuesday. They call the next dealership on the list.

Notice what’s not on that list. Page speed. Core Web Vitals. Largest Contentful Paint.

So why does the vendor pitch work?

Because dealers don’t know what’s actually happening on their site. A vendor with a graphics-rich dashboard can make almost anything look urgent. “Your site is slow” is easier to sell than “your inventory feed is misaligned with your paid campaigns and your BDC isn’t picking up the phone after 6pm.”

Page speed is the mechanic who tells you your air filter needs changing when you came in for an oil change. Technically true. Rarely the thing that was going to strand you on the highway.

The honest version of the page-speed conversation

Page speed matters at the extremes. If your site takes twelve seconds to load, you have a real problem. If your mobile images are 8 megabytes each, you have a real problem. If your scripts are blocking the first paint for a full ten seconds, you have a real problem.

Those problems are also obvious. You don’t need a vendor to tell you a site is broken. You open it on your phone, you see it hang, you know.

The gap between a 43 and a 99? That’s not a shopper problem. That’s a Google-ranking problem. It affects your SEO. It affects how aggressively Google serves your pages in paid auctions. Both of those are real. Neither of them is “shoppers are abandoning your site.”

Those are different problems. They deserve different conversations. And they certainly deserve different budgets.

So, what about SEO? What about the AI crawlers?

Fair question. Both are real. Let’s handle them.

Google uses Core Web Vitals as a ranking factor. Not a huge one, but a real one. If you and your closest competing dealer are tied on everything else that matters, like content, backlinks, local relevance, and the fundamentals of a well-structured site, Google uses page speed as a tiebreaker. That’s documented. That’s true.

AI search is similar. Google’s AI Overviews, ChatGPT, Perplexity, and Claude all have crawler-latency budgets. A site that renders cleanly and fast is more likely to show up in an AI-generated answer than a site that doesn’t. That’s also real.

So yes. Page speed affects SEO. Page speed affects AI visibility. The 43-to-99 jump matters for both of those things.

Here’s the honest version of what that means.

Page speed is a ranking input. It is not a shopper experience input. Those are two different conversations, with two different decision frameworks, and two very different price tags attached.

A dealer who wants to improve organic rankings and AI visibility should absolutely care about page speed. That’s a technical SEO conversation, which usually involves a few hundred to a few thousand dollars of optimization work. Compress the images. Clean up the scripts. Enable caching. Fix the fonts. Most of that can be done without rebuilding the website at all.

A dealer who is told their site is slow and therefore shoppers are abandoning them, and that the solution is a $2,000 to $10,000 full site rebuild, is being sold a ranking fix dressed up as a conversion fix. Those are not the same product. Those are not the same price. And one of them is based on a shopper behavior that doesn’t actually happen.

Speed matters. For Google. For AI crawlers. For paid-ad Quality Scores.

It does not matter the way the vendor says it does.

But doesn’t a slow website cost us leads?

Here’s the one every vendor leans on hardest. And here’s where the pitch really falls apart.

Studies get cited. Akamai says 100 milliseconds costs 7% of conversions. Google says 53% of mobile users abandon pages over three seconds. Deloitte says a tenth of a second lifts conversions 8%. Those numbers sound terrifying when they’re on the vendor’s slide.

Here’s what nobody mentions. Every one of those studies was run on commodity ecommerce. Shoppers at checkout, buying a $40 phone case, a sweatshirt, a travel pillow. People with no emotional investment in the purchase. People whose entire relationship with the brand lasts 90 seconds.

That is not your car shopper.

Your shopper has been researching for weeks. They have a spouse in the room. They have a trade-in parked in the driveway they’re mentally pricing. They have a credit score they’re pre-worrying about. They have a payment they’re trying to hit. They are not going to abandon that search because the inventory page took four seconds instead of one.

Then there’s the vendor case study. “This dealer went from a 40 to a 95 and leads went up 23%.” Sounds like proof. Except that dealer also got a new website. With better forms. Better photos. A cleaner VDP. Clearer calls to action. A BDC workflow refresh. Twenty things changed at once, and the vendor is attributing the lift to the one they sold.

Here is what actually costs dealers leads. In order:

• Lead forms that ask for ten fields when three would do.
• “Get Best Price” CTA’s that ask for customer info instead of just showing the price.
• “Call for price” CTA’s where a payment estimator should be.
• VDPs with three photos instead of thirty.
• Mobile layouts where the submit button is behind a chat widget.
• A BDC that answers leads in four hours instead of four minutes.
• Phone numbers that go to voicemail at 6:30pm on a Tuesday.
• Prices in the ad that don’t match prices on the site.

Those are the real lead killers. Those are the things that actually cost you deals.

Page speed in the four-to-six-second range? That is not even in the top twenty.

Fix the form. Fix the price transparency. Fix the BDC response time. If after all of that you still want to spend money on a faster site, by all means do it. But do it because you want better Google rankings. Not because you were told your shoppers are leaving over something they do not notice and do not blame you for.

The question to ask the next vendor who shows you a PageSpeed slide

Just one. Say it with a straight face.

“Can you show me one dealer who actually lost a sale a shopper described as a page-speed issue?”

They won’t be able to. Because that’s not how high-intent shoppers think, that’s not how they describe their experience, and that’s not what they blame when they don’t buy. They blame their signal. They blame the dealer’s finance team. They blame the fact that the car wasn’t actually on the lot. They blame their spouse.

They have never, in the history of car buying, left a dealership’s website and told their partner, “The Largest Contentful Paint was unacceptable.”

The point

Dealer sites can and should be built well. Speed matters. Code quality matters. User experience matters.

But the vendor who sells you a $2,000 rebuild on the strength of a “43 to 99” PageSpeed jump is selling you a Google-ranking improvement dressed up as a shopper-behavior fix. Those are not the same product.

The shopper is still blaming Verizon.