6.3 seconds to 1.8. As much as 35% more conversions.6.3 seconds to 1.8. As much as 35% more conversions.
Every page under two seconds on low-4G, guaranteed. Mtrix measures the speed
your shoppers actually feel on their own phones, then shows it next to the
revenue it earns — across the stores we run, that drop drove as much as 35%
more conversions on the same traffic.
Performance/Speed Insights
lumière.co
Production
DesktopMobile
Last 7 days
S
Real Experience Score98GreatP75 · mobile
First Contentful Paint0.9s
Largest Contentful Paint1.4s
Interaction to Next Paint64ms
Cumulative Layout Shift0.02
Time to First Byte0.28s
Mobile · real visits · last 7 daysReal Experience Score
Great — above 90
More than 75% of shopper visits had a great experience — measured on
their phones, not in a lab.
Weighted from FCP · LCP · INP · CLS · TTFB on real visits.
How the score is weighted →
P75P90P95P99
PagesTemplates61,204 real visits
Poor<50
No poor pages
Needs improvement50–90
Nothing to improve
Great>90/23K99/products/aero-knit-tee12K98/collections/new-in8.4K97/checkout6.1K96/cart5.2K98/search3.9K95
View all
The speed your shoppers lived — not a lab score.
The load times your customers' own phones recorded on your live store. That's the
only speed that converts, so that's the only speed we measure.
Before6.3s
→
After1.8s
Conversionup to +35%
Across the stores we run · same traffic, same period
A lab scores a URL. Mtrix scores your checkout.
Your Shopify speed score and Google PageSpeed are single lab runs on one borrowed
phone — useful for debugging, blind to your live traffic, and weeks stale. Mtrix is
your real shoppers, today, on the pages that move money — your advertorial, your
checkout, your upsell — with the dollars each slow one is bleeding.
Lab run
Your Shopify / PageSpeed score
89
https://yourstore.com/pageone device · one run · one URL · ~28-day-old field data
One Lighthouse run on a borrowed phone — great for debugging, blind to your live traffic.
It folds in Google’s field data — but that field data is about 28 days stale.
One blended number per URL — no slicing by campaign, page type, or country.
No idea what a slow checkout costs you.
Mtrix
Your shoppers, your funnel, your dollars
By funnel stage · Mobile · last 24h
Advertorial82Needs work–$1.9k/mo
Checkout71Needs work–$3.1k/mo
Upsell64Needs work–$2.2k/mo
your shoppers · today · by funnel stage · in dollars
Every real device, network, and city your shoppers actually use.
Fresh today — a regression you ship Tuesday shows up Tuesday, not next month.
Sliced by page type, campaign, country, and more.
Each slow page priced in the conversions it’s costing you.
Lab tools earn their place pre-launch. Even Google notes its lab result may not capture
real-world conditions — real-user data is the only way to know for sure.
See which page — and which device — is losing the sale.
Every page gets one score and a plain verdict — Great, Needs work, or Poor — with the
four metrics behind it, so it’s never a black box. Toggle Desktop and Mobile to catch
the gap your phone shoppers feel, walk the funnel stage by stage, and slice by
campaign, country, or device — 27 dimensions in all. You fix the page that’s actually
costing you the sale, not a blended average.
Checkout · Mobile
71Score
Needs work
28-day trend
FCP1.6s
LCP2.9s
CLS0.08
TBT240ms
The four metrics behind the score — never a black box.
Funnel stage
Checkout · Desktop 96 (Great) · Mobile 71 (Needs work) — your
highest-intent page is the slow one on mobile.
Under 2 seconds on low-4G. Every page. Guaranteed.
Not “up to” fast. Not “fast on Wi-Fi.” Every page Mtrix ships loads in under two
seconds on a mobile, low-4G connection — the honest worst case your shoppers hit on
the train or a crowded signal. And it’s verified by your real shoppers, not a lab
score — a promise the simulator-based guarantees can’t make.
SLA
Verified by real shoppers
Every page we ship. Under 2.0s. Mobile low-4G.
Verified by real shoppers.
Why low-4G
A weak mobile signal — on the train, in a parking lot, on a crowded network — is
where shoppers feel slow the most and bail. We guarantee the hardest case, not
the easy one.
If a page we ship misses under 2s on low-4G,
we fix it free — before it counts against you.
Promises a number on a simulator. A perfect lab score can still crawl for real shoppers.
vs
Mtrix guarantee
Measured on your shoppers’ real devices and networks. We guarantee what your customers actually feel.
Checkout · Mobile · live
Speed score
71Needs work
LCP 2.9s · TBT 240ms
Conversion rate
2.1%+0.6 pts after fix
revenue / visitor $1.84
Watch the session
A/B test running · faster checkout
Load before they leave. Convert before they go.
The speed your shoppers feel, priced in the revenue it earns — and a promise no
lab-score tool can make: under two seconds on low-4G, every page, guaranteed. Find
the page that’s costing you, fix it, and watch the number move.