{"id":324,"date":"2026-08-17T11:52:06","date_gmt":"2026-08-17T06:22:06","guid":{"rendered":"https:\/\/jnkplugins.com\/blog\/wordpress-plugin-directory-census-2026\/"},"modified":"2026-08-17T11:52:06","modified_gmt":"2026-08-17T06:22:06","slug":"wordpress-plugin-directory-census-2026","status":"publish","type":"post","link":"https:\/\/jnkplugins.com\/blog\/wordpress-plugin-directory-census-2026\/","title":{"rendered":"15,637 WordPress Plugins Were Published and Never Touched Again"},"content":{"rendered":"

On 17 August 2026 we pulled every single plugin listing in the WordPress.org directory \u2014 all 69,838 of them \u2014 and looked at one unglamorous field: the gap between the day a plugin was added and the day it was last updated. For 15,637 plugins, those two dates are the same day<\/strong>. They were published, and then nobody ever touched them again.<\/p>\n

That raw number needs a haircut before it means anything, and we’ll do that in a moment. But the shape of it survives the haircut: 8,851 plugins were published more than two years ago and have never received a single update<\/strong>. That’s 12.7% of the directory. Not “hasn’t been updated lately” \u2014 never, not once, not a version bump, not a readme tweak.<\/p>\n

Below is what we found, how we counted it, and the raw data so you can check our arithmetic. We have no product to sell you in this post and there are no affiliate links in it.<\/p>\n

The honest version of the number<\/h2>\n

Of those 15,637 never-updated plugins, 5,576 were published within the last year. A plugin released two months ago that hasn’t shipped an update yet is not abandoned. It’s new. Quoting the 22.4% figure without saying that would be the kind of statistic that spreads fast and falls apart the moment somebody checks it.<\/p>\n

So here is the same measurement at every cut, and you can take whichever one you think is fair:<\/p>\n\n\n\n\n\n\n\n\n\n
Published and never updated since<\/th>\nPlugins<\/th>\nShare of directory<\/th>\n<\/tr>\n<\/thead>\n
any age (the raw figure)<\/td>\n15,637<\/td>\n22.4%<\/td>\n<\/tr>\n
published more than 1 year ago<\/td>\n10,061<\/td>\n14.4%<\/td>\n<\/tr>\n
published more than 2 years ago<\/strong><\/td>\n8,851<\/strong><\/td>\n12.7%<\/strong><\/td>\n<\/tr>\n
published more than 3 years ago<\/td>\n8,546<\/td>\n12.2%<\/td>\n<\/tr>\n
published more than 5 years ago<\/td>\n7,628<\/td>\n10.9%<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

The curve flattens hard after the two-year mark, and that’s the interesting part. Once a plugin has gone two years without its first update, it almost certainly never gets one. Between the three-year and five-year cuts we only lose about 900 plugins. These aren’t projects that are running late. They’re finished, in the sense that a headstone is finished.<\/p>\n

How we know they were never updated<\/h2>\n

We should be upfront that this is a proxy. The API gives an added<\/code> date and a last_updated<\/code> timestamp, and we’re calling a plugin never-updated when both land on the same calendar day. That could in principle catch plugins whose update history simply isn’t reflected in those fields, so we tested it three ways before believing it.<\/p>\n

The first test was version strings. If these plugins really never shipped again, they should be sitting on a first release. 70.3% of the never-updated set is on a version like 1.0.0, 1.0, 0.1 or 1.0.1.<\/strong> Across the directory as a whole, only 31.1% of plugins sit on such a version. That’s a 2.3x enrichment, which is about as clean a signal as you get from this kind of data. The single most common version in the set is 1.0.0<\/code>, held by 5,003 plugins.<\/p>\n

The second test was reading the actual listings. We pulled a handful of the set up in a browser and checked them against what the API told us. They matched exactly, including the ones we’d have bet against.<\/p>\n

The third was simply looking at the oldest entries, which turn out to go back further than we expected. protect-old<\/code> was added on 26 December 2004 at version 0.1. wysiwyg<\/code> landed in March 2005 and still has 70 sites running it. Both are still installable today.<\/p>\n

The ones with actual users<\/h2>\n

Most of these plugins are invisible. 13,640 of the 15,637 have ten or fewer active installs, which is what you’d expect: somebody scratched an itch, published it, moved on, and the world never noticed.<\/p>\n

But 419 of the two-year-plus set have 100 or more active installations<\/strong>, and a handful are genuinely widely used. The one that stopped us was acf-unique-id-field<\/code>: published on 5 December 2014, still on version 1.0.0, last touched twelve years ago<\/strong>, and currently running on 3,000+ live WordPress sites<\/strong>. vlibras-widget<\/code> is on 5,000+ sites at version 1.0, untouched since April 2020.<\/p>\n

We want to be careful here, because “not updated” is not the same as “broken”. A small plugin that does one thing against a stable API may genuinely need no changes for a decade, and WordPress’s backward compatibility is good enough that plenty of them still work fine. The honest framing is narrower: these are plugins where, if something did break or a vulnerability were found, there is no evidence anyone is home.<\/p>\n

While we were in there: the 59% figure is out of date<\/h2>\n

There’s a statistic that has been travelling around the WordPress world for about a year now: 59% of WordPress plugins, more than 34,000 of them, haven’t been updated in over two years.<\/em> We went looking for where it came from. The earliest version we could find is a September 2025 article that gives it as “34,000+ plugins (59.3%)” and attributes it to “WordPress.org and WPExperts” with no link and no method. It was re-quoted in February 2026 as “research published in 2025”, naming no study.<\/p>\n

It was almost certainly correct when it was written. The directory held around 58,000 plugins in 2025, so 34,000 of them is indeed 59%. Here’s the same measurement today:<\/p>\n\n\n\n\n\n\n\n
<\/th>\nSeptember 2025 (as quoted)<\/th>\nAugust 2026 (this census)<\/th>\n<\/tr>\n<\/thead>\n
Plugins not updated in 2+ years<\/td>\n34,000+<\/td>\n35,671<\/strong><\/td>\n<\/tr>\n
Total plugins in the directory<\/td>\n~58,000<\/td>\n69,838<\/strong><\/td>\n<\/tr>\n
Share<\/td>\n59.3%<\/td>\n51.1%<\/strong><\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

So the percentage everyone is quoting has fallen by eight points. It would be easy, and wrong, to read that as improvement. The number of unmaintained plugins went up by roughly 1,700.<\/strong> The share fell only because the directory added about 11,800 new plugins underneath it. Nothing got cleaned up. The denominator just grew faster than the problem.<\/p>\n

The rest of the snapshot<\/h2>\n

Since we had all 69,838 rows sitting there, here is everything else worth reporting. Every percentage uses the same denominator and the same snapshot.<\/p>\n\n\n\n\n\n\n\n\n\n\n\n\n
Measure<\/th>\nPlugins<\/th>\nShare<\/th>\n<\/tr>\n<\/thead>\n
Not updated in over 1 year<\/td>\n41,510<\/td>\n59.4%<\/td>\n<\/tr>\n
Not updated in over 2 years<\/td>\n35,671<\/td>\n51.1%<\/td>\n<\/tr>\n
Not updated in over 5 years<\/td>\n28,153<\/td>\n40.3%<\/td>\n<\/tr>\n
10 or fewer active installs<\/td>\n41,916<\/td>\n60.0%<\/td>\n<\/tr>\n
No ratings at all<\/td>\n39,383<\/td>\n56.4%<\/td>\n<\/tr>\n
Declares no minimum PHP version<\/td>\n31,673<\/td>\n45.4%<\/td>\n<\/tr>\n
Declares PHP 8.0 or higher<\/td>\n4,253<\/td>\n6.1%<\/td>\n<\/tr>\n
No support thread in the trailing two months<\/td>\n66,442<\/td>\n95.1%<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

Two of those deserve a sentence. 45.4% of plugins declare no minimum PHP version at all<\/strong>, and only 6.1% declare PHP 8.0 or higher, even though PHP 7.4 has been end-of-life for years. The most common single declared value is 7.4, on 16,013 plugins.<\/p>\n

And the support number is not what it looks like at first glance. wordpress.org’s thread counts cover a rolling two-month window<\/em>, not all time \u2014 the listings say “issues resolved in last two months” in as many words. So the correct reading of 95.1% is that most plugins are quiet, not that most plugins ignore their users. Among the 3,396 plugins that did get a question in those two months, 8,731 threads were raised and 5,346 marked resolved, a pooled resolution rate of 61.2%.<\/p>\n

The concentration is worth noting too: the twenty largest plugins account for 32.6% of every active install we counted. If you’ve ever wondered how the directory decides what to show you<\/a>, this is the raw material it’s working with.<\/p>\n

How we counted<\/h2>\n

Everything here comes from one public endpoint, api.wordpress.org\/plugins\/info\/1.2\/<\/code> with action=query_plugins<\/code>, walked 280 pages deep at 250 results per page on 17 August 2026. We keyed every row by slug, so duplicates can’t inflate anything. The run returned 69,838 unique slugs against the API’s own reported total of 69,838, with no failed pages and no unparseable dates.<\/p>\n

Four things we’d want a careful reader to know, because they’d change how you quote this:<\/p>\n

    \n
  • The directory has no official total.<\/strong> Ask the API for the popular<\/code> listing and it reports 65,739 plugins. Ask for new<\/code> and it reports 69,838. That’s a 4,099 gap, and wordpress.org\/plugins\/ itself publishes no number at all. We counted the new<\/code> listing because it’s the more complete of the two, and because its ordering is stable enough to paginate to the end. The popular<\/code> listing is sorted by install count, and so many plugins tie at the bottom that deep pages re-serve results you’ve already seen.<\/li>\n
  • Active install counts are rounded into buckets<\/strong> by WordPress.org \u2014 10, 20, 30, and so on up to 10 million. Anything we’ve derived from them is approximate by construction.<\/li>\n
  • Only open plugins appear.<\/strong> Closed and removed plugins vanish from the API entirely, which mechanically lowers every “stale” share we’ve reported. The true picture is worse than these numbers, not better.<\/li>\n
  • Support thread counts are a two-month window<\/strong>, as described above. They are not lifetime totals.<\/li>\n<\/ul>\n

    If you’d rather not take our word for any of it, the full snapshot is published below.<\/p>\n

    The data<\/h2>\n

    The complete census \u2014 one row per plugin, with the added date, last updated timestamp, version, declared requirements, install bucket, ratings and support counts \u2014 is published as a single download: wordpress-plugin-census-2026-08.zip<\/strong><\/a> (7.2 MB). It contains the data as both CSV and JSON, plus a README documenting every column and the limitations above. Unzipped it’s 38 MB, which seemed rude to make anyone download.<\/p>\n

    This is the exact snapshot the numbers above were computed from, not a summary of it. If you want to check a specific figure in this article, the row is in there.<\/p>\n

    To cite this: “WordPress.org Plugin Directory Census, JNK Plugins, 17 August 2026, n=69,838.”<\/em> A link back is appreciated but not required. If you rerun it and get something different from us, we’d genuinely like to know.<\/p>\n

    We’ll run it again in six months, which is when this stops being a snapshot and starts being a trend. In the meantime, if you’re weighing up what’s sitting in your own wp-admin, the related question of how many plugins is too many<\/a> is worth a read, and our notes on getting a plugin through wordpress.org review<\/a> cover what it takes to get one listed in the first place.<\/p>\n