This page is about one tool’s data: the Mobile-Friendly Checker. For how these scores sit alongside the other measures, the distributions in full and the methodology, see SEO Score Benchmarks.
Which mobile check is missing, and how often
The tool takes a URL and runs eight checks: the viewport tag and the two values inside it, whether zoom is blocked, responsive CSS, responsive images, fixed-width layout and the charset declaration. The percentage below is the share of runs in which the check did not pass.
| Check | Did not pass | Weight |
|---|---|---|
| No fixed-width layout | 55% | Partial |
| Responsive images (srcset) | 53% | Partial |
| Responsive CSS (media queries) | 31% | Hard fail |
| initial-scale defined | 19% | Partial |
| width=device-width | 13% | Hard fail |
| Viewport meta tag | 11% | Hard fail |
| Zoom not disabled (a11y) | 8% | Partial |
| Charset declared | 5% | Partial |
The top and bottom halves of the table tell two different stories. The two items at the top — a fixed-width layout in 55% of runs and no responsive images in 53% — are the most common, and both are partial credit lost: the page still works on a phone. The three at the bottom are far rarer but scored as hard fails, because without them the page genuinely does not work on a phone. The average scan passes 6.0 of the eight checks.
The cliff inside the score
What separates this measure from the others is the shape of its distribution. The 25th percentile is 78 and the 10th is 26 — 52 points apart. 62% of scans collect at 80 and above, 11% drop below 40, and only 1% sit in the 40–59 band between them. The grade mix says the same: 586 scans graded A and 163 graded F, with just 29 in the D band in between.
The gap is a property of the scoring. The viewport meta tag and the width=device-width value inside it carry half the available points between them — and both live in the same single line:
<meta name="viewport" content="width=device-width, initial-scale=1">
If a page has that line, three checks pass at once; if it does not, three fail at once. There is no middle state, which is why there is a hole in the distribution. The numbers line up too: the viewport tag is missing in 11% of runs and width=device-width in 13% — against 11% of scans falling below 40. They are the same pages. The practical consequence is that a page scoring badly here does not need a list of improvements, it needs one line. Adding it moves the page straight to the edge of the top band.
The category ranking looks at first like it contradicts that: Responsive Layout is the weakest category in 80% of runs against only 15% for Viewport & Scaling. It is not a contradiction. "Weakest category" is a ranking that only means anything on pages that have the viewport tag. Without it a page is already on the floor; with it, the layout itself is the only weak point left.
What is left on the pages that are fine
The pages on the right of the cliff — the large majority of the sample — keep missing the same two items. A fixed-width layout is still present in 55% of runs: usually an old table, a fixed-size image, or a width="960" left behind by a template. Responsive images are missing in 53%. The second does not break mobile-friendliness, but it does mean a phone downloading desktop-sized images — a cost paid on speed and data rather than on layout.
One item sits at the bottom of the list but deserves separate mention: zoom is blocked for the user in 8% of runs. That is not an SEO problem, it is an accessibility one: user-scalable=no, added to stop a layout breaking, also stops a reader with low vision enlarging the text. If the layout is already responsive, the line is not needed.
Test a page for mobile-friendliness
Frequently asked questions
What is a good mobile-friendly score?
In this dataset the median is 84/100 and 62% of scans reach 80 or above. But on this measure the number matters less than which side of the split you are on: the 25th percentile is 78 while the 10th is 26. There is almost nothing between 78 and 26. If you are under 60 here, this is not a "tighten it up" situation — it most likely means the viewport tag is absent altogether.
Why is the gap between the 10th and 25th percentile so large?
Because half the score lives in a single line of HTML. The viewport meta tag and the width=device-width value inside it carry half the available points between them — and both live in the same tag. If a page has that tag, both pass; if it does not, both fail. There is no middle state. That is why 11% of scans (155 runs) drop below 40 while the rest collect above 60. The score is a threshold, not a slope.
Does a missing srcset make a page not mobile-friendly?
No — it is scored as partial credit lost, not a hard fail. The page still renders correctly on a phone; the problem is that images are downloaded larger than they need to be. It is one of the most common gaps on the list all the same, with no srcset in 53% of runs. The effect shows up in data usage and load time rather than in layout, which is why it is really paid for on the speed side.
Is it harmful to stop users zooming?
Yes, and it still happens in 8% of runs in this sample. Adding user-scalable=no or maximum-scale=1 has been a designer’s shortcut for years to stop a layout breaking under zoom; the cost is that readers with low vision cannot enlarge the text. There is also no longer a reason for it: if the layout is genuinely responsive, disabling zoom buys nothing.
The sample, the methodology, the anonymization and the CC BY 4.0 licence are all set out in the methodology section on the hub.