PSG Vendor Mobile Performance Benchmark 2026

When you’re shortlisting PSG vendors, you’ll typically compare pricing, features, industry fit, and support quality. Mobile website speed rarely makes the list — but it’s one of the few signals you can verify independently, without relying on a vendor’s own claims. We measured all 278 PSG-approved vendors to give you that benchmark.

Why We Measured This

PSG buyers typically shortlist five to ten vendors before making a decision. Most of the comparison work relies on vendor-supplied material: quotes, demos, case studies. These are valuable, but they’re not independent.

A vendor’s website speed is different. It’s public data, measured by Google’s own tools, and it reflects a genuine investment (or lack of investment) in user experience. A vendor who scores well on mobile has optimised their public-facing presence for real visitors on real devices. That’s a meaningful signal about how they approach quality.

We ran Google PageSpeed Insights across all 278 PSG-approved vendors using the mobile strategy in April 2026. The results are available to any buyer who wants to check them.

What the Data Shows

More Than Half of PSG Vendors Score Below 50 out of 100

The median Lighthouse performance score across all 278 vendors is 47 out of 100. That means more than half of all PSG-approved vendors score below the pass-fail line on mobile.

Lighthouse scores run from 0 to 100. Scores in the 90s are considered fast. Scores in the 50s and 60s are moderate. Scores below 50 indicate meaningful room for improvement.

Put simply: if a vendor scores 65 or above, they are already performing better than most of their PSG-approved peers.

Most Vendors Are Measured by Simulation Only

Google’s tools offer two types of data. The first is lab data — a controlled simulation run on a standard device under standard conditions. The second is field data, drawn from real visitors using Chrome, known as the Chrome UX Report (CrUX).

Field data is more meaningful because it reflects actual user experience. But it only exists for sites with enough real visitor traffic.

Of the 278 vendors we measured:

• 33% have field data — enough real visitors for Google to measure their actual performance
• 67% are lab-only — their scores come from a simulation, not real-user observation

This means for most vendors, you’re seeing a simulated estimate. That’s still useful for comparison, but it’s worth knowing the limitation.

Fewer Than One in Four Measured Vendors Pass All Three Core Web Vitals

Among the 91 vendors with enough real-user traffic for at least some field measurement, the results are mixed — and not every vendor is measured on every metric:

Passing all three together is harder. Of the 47 vendors measured on all three metrics, only 11 (23%) pass every one on mobile.

What This Means for Buyers

You don’t need to be technical to use this data. A few practical implications:

A score of 65 or above already puts a vendor ahead of most. The bar across PSG-approved vendors is genuinely low. You don’t need to find a perfect scorer — a vendor in the 65–80 range is performing well relative to the field.

Check the score before your first meeting. Visit pagespeed.web.dev, enter the vendor’s website URL, select the Mobile tab, and read the score. It takes about two minutes. Use it as a conversation starter: “We noticed your site scores X on mobile — how do you approach digital experience for your clients?”

Don’t use it as the only signal. Mobile speed is one factor. Category fit, pricing, support quality, and references matter far more. Speed is a useful tiebreaker, not a replacement for full due diligence.

Grantla shows this data inline. Every PSG solution page on Grantla now includes the vendor’s mobile performance verdict, so you can see it while you’re already comparing solutions — no separate lookup needed.

For a wider Singapore B2B view that sits alongside this PSG-specific snapshot, our sister site’s Most Singapore B2B Websites Fail Google’s Speed Test measures the same patterns across a broader B2B sample.

What Core Web Vitals Actually Measure

Core Web Vitals are three specific signals Google uses to measure real-world page experience on mobile:

LCP — Largest Contentful Paint How long it takes for the main content on the page to appear. The Good threshold is 2.5 seconds or faster. A slow LCP means visitors are staring at a mostly blank screen for longer than they should.

INP — Interaction to Next Paint How quickly the page responds when someone taps a button, opens a menu, or fills in a form. The Good threshold is 200 milliseconds or faster. A poor INP makes a site feel laggy and unresponsive on mobile.

CLS — Cumulative Layout Shift Whether page elements move unexpectedly as the page loads — buttons jumping, text shifting, images pushing content down. The Good threshold is a score of 0.1 or lower. High CLS makes a page feel unstable and can cause accidental taps.

All three together give a rounded picture of whether a site is genuinely pleasant to use on a mobile phone.

Methodology

What we measured: All 278 PSG-approved vendors active in April 2026.

Tool: Google PageSpeed Insights API, which combines Lighthouse (lab simulation) and the Chrome UX Report (real-user field data).

Device strategy: Mobile — this reflects how most users browse, and it applies the stricter mobile scoring thresholds.

Level: Origin-level scores (the vendor’s main domain), not individual pages.

Snapshot date: April 2026.

Data types:

• Lab data (Lighthouse): A controlled simulation run by Google’s tooling. Consistent and comparable across all vendors, but not drawn from real visitors.
• Field data (CrUX): Drawn from real Chrome users visiting the site. More meaningful but only available for vendors with sufficient public traffic.

Both data types are public. Neither is provided by the vendors themselves.

Limitations to Keep in Mind

Marketing site speed is not product speed. We measured vendor websites, not the software products they deliver. A vendor with a slow website might deliver an excellent SaaS platform. These are separate things.

Lab scores vary slightly between runs. Lighthouse simulations are consistent but not perfectly stable — a score can shift by a few points between snapshots. Treat scores as ranges, not precise figures.

No CrUX data means lab-only. Vendors with little public web traffic don’t generate enough Chrome user data for field measurement. Their scores are entirely simulated. This is a limitation of the methodology, not a judgment on the vendor.

Scores change. Vendors update their websites. A vendor who scores 45 today might score 75 in three months after a rebuild. This is a snapshot, not a permanent ranking.

Use this data as a conversation starter and a baseline comparison — not as a definitive verdict on any vendor’s capabilities.

Related Guides

• PSG Vendor Desktop Performance Benchmark 2026 — The desktop story: median 61/100, plus SEO, accessibility, and best-practices signals
• Fastest PSG Vendor Websites 2026 — The top performers by Lighthouse score, named and ranked
• How to Check a PSG Vendor’s Website Speed — The 2-minute buyer’s checklist using PageSpeed Insights
• Core Web Vitals Explained for Singapore SMEs — What LCP, INP, and CLS mean in plain English
• Is PSG Worth It? — Honest ROI assessment for Singapore SMEs
• PSG Grant Explained — Full eligibility, process, and claims guide
• PSG Pitfalls That Get Applications Rejected — Mistakes to avoid before you apply
• Common Grant Pitfalls — mistakes to avoid when applying for Singapore grants
• Singapore Grants Glossary — key grant terms and definitions
• 本指南中文版 — 2026年PSG供应商手机端速度基准报告（中文）