Platform comparison · lab receipts

Captured 2026-07-22 · PSI mobile · best of 2

+25 to +55 PSI points ahead.
1.4-4x faster largest paint.

Four dealer-website platforms. Five real production dealer sites each, verified by the platform's own fingerprint in the served HTML, measured with Google PageSpeed Insights against our two production templates - same test, same day, every site named below. Medians, not cherry-picks.

The ledger

Median PageSpeed, platform by platform

Google's own test, run on real production sites. Competitor rows are the median of five sites; our rows are the two template sites themselves, exactly as measured.

Site setScore (PSI mobile)LCP (lab)SitesOur lead
Spif.io template-1.spif-staging.com 86/100 3.53stemplatebaseline
Spif.io template-2.spif-staging.com 85/100 3.63stemplatebaseline
DealerCenter60/100 5.01s5+25.5 pts · 1.4x LCP
OverFuel51/100 9.98s5+34.5 pts · 2.8x LCP
Cars For Sale49/100 9.88s5+36.5 pts · 2.8x LCP
Motive30/100 14.37s5+55.5 pts · 4x LCP

competitor score and LCP = median of 5 production sites per platform · best of 2 PSI mobile runs per site · captured 2026-07-22 · bar length = that row's score, 0-100 · lab-vs-lab - how we tested

Your platform on this list? Request a measurement - script off vs on
Full transparency

Every site, named

No anonymous aggregates. Each platform's five measured sites, both raw runs per site, and the real-user field verdict where Google has one. Expand any platform.

Spif.io templates median 85.5/100 · 3.58s LCP · n=2 + expand- collapse

spif = production template sites (template-1/template-2.spif-staging.com), not client sites. Staging domains carry no real-user traffic, so Google has no field sample for them, which is why this page compares lab to lab.

Both templates carry demo inventory - 12 vehicles each, confirmed live on template-1 and template-2 - not a full production catalog. Production dealer sites run larger real inventories.

DealerCenter median 60/100 · 5.01s LCP · n=5 + expand- collapse
  • zoomut.com 60/100 4.41s LCP runs 31 · 60 no field data
  • olmakscars.com 72/100 3.81s LCP runs 72 · 69 no field data
  • lowmilesmotors.com 62/100 15.71s LCP runs 62 · 46 no field data
  • autoshowcasesa.com 47/100 7.04s LCP runs 47 · 47 no field data
  • hrautomotors.com 28/100 5.01s LCP runs 28 · 28 field fail

our lead vs this platform: +25.5 points median score · 1.4x median LCP.

platform: DealerCenter (Nowcom) hosted dealer-website product. Attribution fingerprint: WordPress themes on cdn.dealercenterwsstatic.net/wp-content/themes/dws-v3*/ + JS namespace 'DealerCenter||(DealerCenter={}))' + chat-cf.dealercenter.net / imagescf.dealercenter.net service hosts

OverFuel median 51/100 · 9.98s LCP · n=5 + expand- collapse

our lead vs this platform: +34.5 points median score · 2.8x median LCP.

platform: OverFuel dealer-website platform (overfuel.com). Attribution fingerprint: <meta name="generator" content="Overfuel"> + assets on static.overfuel.com/dealers/<slug>/

Cars For Sale median 49/100 · 9.88s LCP · n=5 + expand- collapse

our lead vs this platform: +36.5 points median score · 2.8x median LCP.

platform: Carsforsale.com dealer-website product (custom-domain sites). Attribution fingerprint: Footer credit 'Powered by Carsforsale.com' (1999-2026 (R) or (C) 2026 TM variants) + vehicle imagery on cdn05.carsforsale.com + signin.carsforsale.com dealer login

Motive median 30/100 · 14.37s LCP · n=5 + expand- collapse

our lead vs this platform: +55.5 points median score · 4x median LCP.

platform: Motive dealer-website platform (motivehq.com, formerly Wackk; OEM-certified). No footer credit exists; the platform signs itself via *.app.ridemotive.com asset/API/tenant hosts (ridemotive.com 308-redirects to motivehq.com).. Attribution fingerprint: assets.app.ridemotive.com / api.app.ridemotive.com / images.app.ridemotive.com + per-dealer tenant subdomains <slug>.app.ridemotive.com

How every number was produced - method and caveats
Reproducible by anyone

How we tested

A comparison is only worth what its method survives. Here is exactly how every number above was produced, including the parts that don't flatter us.

01

Lab vs lab - framed honestly

same test, both sides, same day

Our templates run on staging domains with no real-user traffic, so Google holds no field (CrUX) sample for them yet. That makes this a lab-vs-lab comparison: the identical PageSpeed Insights mobile test, run by Google's servers on both sides on 2026-07-22. These are not our local runs, and we are not comparing our lab against their field. One number we won't hide: under PSI's mobile throttling our templates' LCP measures ~3.58s, above Google's 2.5s “good” bar. The story of this page is the relative gap, measured identically: +25 to +55 points, 1.4-4x on largest paint.

One more caveat, stated plainly: Both templates carry demo inventory - 12 vehicles each, confirmed live on template-1 and template-2 - not a full production catalog. Production dealer sites run larger real inventories.

02

How sites were selected

fingerprint-verified, not assumed

Five real production dealer sites per platform, selected from larger hand-verified pools for dealer-type fit and geographic spread. Every candidate's homepage HTML was fetched live and the platform's own fingerprint - footer credit, asset CDN, or generator tag, quoted per platform in the site lists above - confirmed present in the served markup before the site entered the sample. Sites that could not be verified (bot-challenged, unfetchable) were excluded rather than padded in.

03

The test protocol

best-of-2 favors the competitor

PageSpeed Insights API v5, Lighthouse mobile performance, 2 runs per URL - keep the best. A kept best can only flatter the measured site: lab noise inflates it, never deflates it, so every competitor number above is that site's better face. All 22 URLs - ours included - were measured the same way on 2026-07-22.

GET https://www.googleapis.com/pagespeedonline/v5/runPagespeed
  ?url=https://<any-site-above>/&strategy=mobile&category=performance&key=<key>
# 2 runs per URL, keep the best; compare medians per platform
04

Why medians, not bests

single runs swing

Lab runs are volatile: the widest pair in this dataset measured 39 and 68 on back-to-back runs of the same URL (choosecardeal.com). That is why every platform figure quoted on this page is a median of five sites - never a single-site best - and why both raw runs for every site are published in the lists above.

05

The field supplement

where real-user data exists

Lab is the only fair both-sides comparison here, but where Google has real-user field data on the competitor sites we publish the verdict per site above. The strongest signal in this dataset: 5 of 5 measured Motive sites fail Core Web Vitals in real-world field data (28-day p75 of actual Chrome users). Many small independents carry too little traffic for a field sample at all - those are marked “no field data,” not counted against anyone.

Provenance

lab + field data captured 2026-07-22 via the PageSpeed Insights API v5 (Lighthouse mobile / CrUX).

spif = production template sites (template-1/template-2.spif-staging.com), not client sites.

Both templates carry demo inventory - 12 vehicles each, confirmed live on template-1 and template-2 - not a full production catalog. Production dealer sites run larger real inventories.

best-of-2 runs per URL - the kept best favors the measured site.

platform attribution hand-verified in served HTML; fingerprints quoted per platform above.

medians per platform; per-site raw runs published in full. No site was dropped for scoring well.

Evidence build · Spif.io · dataset captured 2026-07-22. Every figure on this page is computed from the published dataset at build time - the page cannot say something the data doesn't.

Already on another platform? Start where you are.

The Spiffy Speed Script works on your current site - one script tag in the head, measured honestly, wins and losses published. And when you want the numbers at the top of this page, that's the platform.