2,383 SEO plugins, a third of them silent: the SEO category in the WordPress directory

Look for a WordPress SEO plugin and you will find no shortage of comparisons, most of them built around feature lists. We asked a different question: what does the directory itself say? The numbers come from the Plugin Directory Census 01/2026, our full count of all 70,909 listed plugins on 28 August 2026. How we counted is in the census workshop report. This piece looks at one category.

What these numbers can do is limited, but you can check every one of them. They show when a plugin last shipped a release, how many installs report to the update check, which WordPress version the author says they tested against, and how many support questions got resolved. Which plugin is better, they cannot tell you.

How silent is the SEO category?

808 of the 2,383 plugins tagged seo had gone more than 365 days without a release on the reference date. That is 33.9 percent. Across the whole directory it was 41,395 of 70,909, or 58.4 percent. The gap holds at longer thresholds too:

That makes seo one of the liveliest large categories. It is the third most common tag in the directory, after woocommerce and widget. Among the 40 most common tags (table 4 in the census section “How silent is your category?”), only seven have a lower silent share: AI at 15.2 percent, then automation, chatbot, checkout, performance, woocommerce and elementor. security comes in just behind at 35.8 percent, analytics at 41.0. At the far end sit widget at 83.7 and twitter at 94.3 percent.

The gap sits with the small plugins

The obvious guess: SEO plugins are bigger, and big plugins go silent less often. It does not hold up. Of the 2,383 SEO plugins, 60 have at least 10,000 installs, which is 2.5 percent. Across the directory, 3.1 percent clear that mark. And among the bigger ones, the lead is gone:

From 10,000 installs up, SEO plugins go silent about as often as everything else, if anything slightly more often. With 60 plugins, each one moves the share by roughly 1.7 percentage points, so “about the same” is as far as the data goes. The low overall figure comes from below 10,000 installs. Why so many small SEO plugins have recent releases is not something the fields tell us, and we are not going to guess.

What counts as an SEO plugin here

A tag is self-reported. Each author picks up to five, nobody checks them, and the API merges nothing: seo and search-engine-optimization are two separate tags. That cuts both ways.

• The tag catches more than you would expect. Three of the 17 plugins with 100,000+ installs that carry seo are, by name and remaining tags, cache or CDN plugins: LiteSpeed Cache, SpeedyCache and Cloudflare. Add a table of contents, related posts and a Yoast SEO add-on for Advanced Custom Fields.
• And it misses the obvious. Redirection with at least 2 million installs, Breadcrumb NavXT and Broken Link Checker do not carry it, although many people would file their jobs under SEO.

That is why the census map uses a wider group of 23 tags for its SEO colour, from schema and xml-sitemap to redirect and breadcrumbs. At 10,000 installs and up, 109 plugins carry at least one of them, 49 without seo. For choosing a plugin that net is too wide: Redirection for Contact Form 7 and Easy HTTPS Redirection end up in it as well. So we stick to two plain rules:

• Headline figure: the tag seo, spelled exactly like that. It is the only one for which the census has values across all 70,909 listed plugins.
• Full SEO suite: a plugin tagged seo plus at least one sitemap tag (xml-sitemap or sitemap) and at least one schema tag (schema or structured-data). Exactly ten plugins with 10,000+ installs meet that rule.

The second rule is ours, but anyone can rerun it. The file census-plugins-01-2026.json holds all 2,199 plugins with 10,000+ installs, with names and values, under CC BY 4.0:

jq -r '.plugins[] | select((.tags | index("seo")) and (.tags | index("xml-sitemap") or index("sitemap")) and (.tags | index("schema") or index("structured-data"))) | .slug' census-plugins-01-2026.json

A suite that spends its tags differently slips through the rule. If it has 100,000 installs, it turns up in the second table below.

The ten full SEO suites on the reference date

Sorted by install band, and within a band by number of ratings. That is the order the census map uses, not a verdict. For the seven suites with 100,000+ installs, the name links to the census profile with every value and a comparison with its size group.

Active installs: lower bound of the band; 4M+ means 4,000,000 to 4,999,999, and the 10M+ bucket is open-ended. Days: since the last release, counted to the reference date. Tested up to and PHP minimum: self-reported by the authors. Rating: stars, number of ratings in brackets. Support: threads marked resolved in the wordpress.org forum in the two months before the reference date. Names without marketing taglines.

None of the suites was silent. Nine of the ten had shipped their last release no more than 30 days before the reference date, the tenth 261 days before. For comparison, the census size-group medians: 8 days at a million installs and up, 12 days at 100,000 to 999,999, 31 days at 10,000 to 99,999.

All were tested against 6.9 or later, seven against 7.1, the WordPress core version on the reference date. As their PHP minimum they state 7.0, 7.2 or 7.4. None states PHP 5.

Ratings need their count next to them. The stars run from 4.0 to 4.9, but the 4.0 rests on 9 ratings and a 4.8 on as many as 27,819. Of 140 support threads in the last two months, 127 were marked resolved (90.7%); pooled across all plugins with 10,000+ installs, it was 66.5%. Then again, 116 of the 140 threads belong to the three largest suites. Where one or two threads come in over two months, the share tells you little.

What else carries the seo tag

The other ten plugins with 100,000+ installs tagged seo, sorted the same way. The column with the remaining tags shows what the author considers the plugin to be.

Two of these ten had gone more than a year without a release, and five state PHP 5.x as their minimum. One is a date, the other an author's statement; neither is a finding about the code. A small tool that has done the same job for years does not need a release every month. What a PHP 5 floor still tells you is covered in our piece on WordPress plugins that declare PHP 5.

How to check an SEO plugin yourself

Whether a suite has the features you need is not in any of these fields. The description on wordpress.org and a test on a copy of your site answer that. The fields help with a different question: is the plugin your titles, sitemaps and structured data depend on still getting releases, and is anyone answering in the forum? Six things you can read off any directory page, no census needed:

• Last updated. In the sidebar. Benchmarks from the reference date: a median of 8 days at a million installs and up, 12 at 100,000 to 999,999, 31 at 10,000 to 99,999. Over 365 days counts as silent in the census, not as abandoned. What to do about a silent plugin is in our guide to updating WordPress plugins.
• Tested up to. The WordPress version the author says they tested with. If it is three or more major versions behind, the directory page shows a warning. On the reference date, 13 of the 60 SEO plugins with 10,000+ installs were below 6.9 (21.7%), against 18.1% for all plugins that size.
• PHP version. Compare it with the PHP version on your server. A low minimum does not mean the code fails on PHP 8. It is a statement, not a test.
• Issues resolved in last two months. This counts the wordpress.org forum only. Whatever a vendor answers elsewhere is not in it. With a handful of threads the share is chance; with dozens it is a signal.
• Number of ratings. 4.9 stars from 378 ratings is a different finding from 4.8 stars from 27,819. Read a few of the latest one-star reviews; they show where users get stuck.
• Active installations. The lower bound of a band, counting installs, not websites. For choosing a plugin the band is enough; the number gets no more precise.

Whether a suite slows your site down is not in any field either. Measure it yourself, before and after installing; the last step of our guide to speeding up a slow WordPress site shows how.

Already running one of the plugins in the tables, or something else entirely? In the plugin check on the census page you enter your slugs, one per line, and see the install band, the last release on the reference date and how each plugin compares with others of similar size. The slug is in the address of the directory page; for Yoast SEO it is wordpress-seo. The lookup runs in your browser.

Limits of this report

• Reference date 28 August 2026. Releases have shipped since, so the days in the tables are different today. Current values are on each plugin's directory page.
• Tags are self-reported. A plugin without seo is missing from the headline figure. A plugin with it counts, even if it is a cache plugin.
• Installs are not websites. They count installations that report to the update check, as the lower bound of a band with one significant digit.
• Tested up to and the PHP minimum come from the authors. Nobody verifies them.
• Support counts the wordpress.org forum only, and only two months of it.
• Silent is not abandoned, and no number here is a security statement. Whether a plugin has vulnerabilities does not follow from any measured field. Which layers a security plugin covers is in our comparison of WordPress security plugins.
• Listed plugins only. Anything not listed on the reference date is in none of the numbers, and not listed does not mean closed.
• Issue 01 is a baseline. It supports no statement about trends, for SEO or anything else.

Conflict of interest

We sell WordPress plugins ourselves, and seven plugins of ours are in the directory. None of them carries the tag seo (checked against the plugin API on 9 October 2026). Two fall into the wider group of the census map: Hafen Core with the tag schema and Linkjet with redirect. Both have fewer than 10,000 installs and appear in no table in this report. Three of the profiles linked here (Rank Math SEO, All in One SEO, LuckyWP Table of Contents) show a box marked as our own advertising for Linkjet. The values are unaffected and collected the same way for every plugin.

The census report itself does not analyse any category we sell in separately. This report takes one category on its own, and with Linkjet we sell in the wider SEO group. That is why the conflict is spelled out here. If you think a value is wrong, write to moin@hafenstudios.com; every correction to the census is listed with its date in the census version history.

Frequently asked questions