Cars.co.za was on page 1 of Google for 10 of my 30 car searches on both days I checked, and AutoTrader for 9. No other marketplace I track held a spot on both days. Count a search if the site was there on either day, and the order flips: AutoTrader 23, Cars.co.za 19.
I chose the searches as plausible customer searches, like “used cars for sale pretoria”, “sell my car for cash” and “car finance south africa”. They’re a list I picked, not a sample of all car searches. I sampled each one on two days, 6 and 7 November, and counted a site only if it was on page 1 both times.
That rule mattered more than the race. Page 1 moved a lot between the two days, the two leaders split the field, and for 14 searches neither leader was on page 1 on both days.
How much does page 1 move between two days?
More than I expected. I compared the two days search by search, and none of the 30 had the same page 1. Across every site in the panel, 56% of the website-and-search pairs on page 1 on 6 November weren’t there for the same search on 7 November.
The leaders’ daily totals look steady. Cars.co.za was on page 1 for 15 searches on the first day and 14 on the second. AutoTrader went from 14 to 18. But the searches changed underneath them. Only 10 of Cars.co.za’s held on both days, and 9 of AutoTrader’s.
Further down, the effect is starker. CarFind is the extreme case: 2 searches on day one, 5 on day two, and not one in common. ChangeCars had 2 and then 1, with no overlap. Both end at 0 on the two-day count.
So one look at Google is a snapshot, not a position. If you track your own ranking, check it on two days before you act on it.
One look at Google is a snapshot, not a position.
How close is the race between Cars.co.za and AutoTrader?
Your guess first. Same 30 searches for every site. Which marketplace held a page-1 spot on both days most often?
Made your pick? Show the result
One search close. But they didn’t hold the same searches.
| On page 1 on both days | Searches |
|---|---|
| Both leaders (3) | used cars for sale bloemfontein; used suv for sale; used vw polo for sale |
| Cars.co.za alone (7) | sell my car for cash; used cars for sale durban; used cars for sale port elizabeth; buy a car online south africa; used bakkies for sale; cheap cars for sale; used toyota hilux for sale |
| AutoTrader alone (6) | trade in my car; used cars for sale pretoria; second hand cars for sale; used cars under r100000; car dealerships johannesburg; toyota dealer pretoria |
| Neither (14) | listed in the next section |
Cars.co.za’s own seven lean to regional and by-model used-car searches, plus selling for cash. AutoTrader’s own six lean to trade-in, budget and dealer searches. I can see which searches each site holds. I can’t see why, and I won’t guess.
Now the audience numbers. AutoTrader’s estimated search audience is about 2.5 million visits a month, and Cars.co.za’s about 986,000. The two biggest audiences of the 19 brands with an estimate are the two page-1 leaders.
CarFind is the exception. Its estimated audience, about 158,000 visits a month, is the third-largest of the 19 with an estimate, yet it held nothing on both days. These visit numbers are DataForSEO’s modelled estimates, built from the searches a site already ranks for, not the companies’ own figures.
Which searches did neither leader hold?
Fourteen of the 30, almost half the panel:
| Neither leader on page 1 on both days (14) |
|---|
| sell my car online; how much is my car worth; car valuation south africa; sell my car johannesburg; used cars for sale johannesburg; used cars for sale cape town; used cars under r200000; used ford ranger for sale; car dealers cape town; new cars for sale south africa; certified pre-owned cars south africa; car finance south africa; bank repossessed cars for sale; car for sale by owner |
On those 14, no outside site held more than 2 searches on both days, so they’re searches to look into, not proof of an opening. The Meerkat Index hub, "Who owns page 1"
I didn’t see Auto Mart, CarFind, ChangeCars, Surf4Cars or Truck & Trailer on page 1 on both days for any of the 30. None of my 30 searches mentions trucks or trailers, so Truck & Trailer’s 0 says little about its own market.
What does a buyer see first on a phone?
I photographed five of the seven marketplace home pages at phone size on 7 November. They’re five of the seven, not a sample of every marketplace. Two showed a notice on the first screen. None covered most of it.
Then the automated phone-size checks, the same three steps on every site: home page, one offer page, one contact action. Four of the seven marketplaces cleared all three checks: Auto Mart, AutoTrader, CarFind and Truck & Trailer. I didn’t submit an enquiry on any site.
The other three are “partly checked”, which isn’t the same as failing. Cars.co.za’s offer page refused my automated check (error 403) in both passes on 7 November, so neither the offer page nor the contact step behind it was checked. On ChangeCars, the contact step couldn’t be checked in either pass. Surf4Cars showed my check a bot check or an empty page, both times. A bot check stops a script. It tells you nothing about what a person sees.
Which marketplace scores best on Lighthouse’s mobile test?
CarFind scored 61, the highest marketplace score. The six I could test ran from 28 to 61. Surf4Cars couldn’t be tested. None had a median of 90 or more, the mark Lighthouse counts as good. Each number is the median of 5 retained runs, a lab score under my test conditions, not a measure of real visitors.
CarFind also shows why the two boards stay separate. The highest Lighthouse score among the marketplaces belongs to a site that held no page-1 spot on both days. The two most visible marketplaces both scored in the 30s. Blend those into one number and you’d hide both facts.
What can you take from this if you don’t sell cars?
What I observed, and what’s worth testing on your own business. It takes a browser and two afternoons.
- Write down 20 to 30 searches your customers type. Leave out your own company name and your competitors’ names.
- Search each one on two different days. Count your site only if it’s on page 1 both times. Don’t count Facebook or Reddit posts.
- Note who else is there, search by search. Where nobody holds a search twice, look into the competing results and what the searcher wants before you read it as a gap.
- Open your home page on your own phone. Can you reach one offer page and one way to contact you? Does anything cover the first screen?
- Run a Lighthouse mobile test five times and keep the middle score, because single runs move around too much to trust.
Your numbers won’t be comparable to the Index. They don’t need to be. They’re about your customers, and that’s the list that tells you whether a new buyer can find you.
They’re about your customers, and that’s the list that tells you whether a new buyer can find you.
How this was measured
I chose the 30 searches before collecting anything, and ran them on Google (South Africa, desktop, English) on 6 and 7 November 2026. A site counts only if it was on page 1 on both days. Automated phone-size checks ran on 6 and 7 November, with the same three steps on every site and two passes each. Speed is Lighthouse’s mobile score for the home page, the median of 5 retained runs under my test conditions. Estimated visits are DataForSEO’s modelled figures, shown as context only.
I don’t measure visitors, leads, sales or revenue. No company paid to be included. The full method, all 30 searches and every change after review are here: How the Meerkat Index measures, and how to correct us. Something wrong? Email [email protected] and I’ll re-check within a week.
For a quick read on your own business, take the free Business Momentum Score. It takes about 4 minutes, names your weakest area from being found to winning enquiries, and gives you three next steps. It’s separate from the Index. https://score.rhinoberry.com/