Abstract
What are the most common reasons plugin users leave one-star reviews on WordPress.org? This study codes 1,539 recent one-star reviews from 77 popular plugins under an equal-N design. Across the full corpus, reliability failures lead: the plugin did not do its job, or it broke the site. Monetization pressure is the second major cluster. Support failures remain common but are not the primary driver for most reviews.
Roots & Fruit. “One-star reviews of popular WordPress.org plugins.” Dataset 2026-08-08-wporg-top100-1star, reviews fielded 26 July 2026, aspect-coded 8 August 2026. Up to 20 recent one-star reviews per plugin among the WordPress.org top 100 with at least 20 lifetime one-stars (1,539 reviews across 77 plugins).
1. Executive summary
This study aspect-codes every review in an equal-N sample of recent one-star WordPress.org topics from popular plugins. Reliability is the lead primary cluster (640 of 1,539 reviews, 42%). Monetization is second (439 reviews, 29%). The single largest primary facet is “plugin failed its job” (264 reviews), ahead of crippled free tiers (170) and site breakage (131). A separate paid-signal overlay finds 44 reviews (~2.9%) where the reviewer says they paid for or use a pro or premium tier.
2. Research question
Primary question. Among the most popular WordPress.org plugins, what complaint facets recur as the main anger driver in recent one-star review topics when each included plugin contributes the same maximum number of reviews?
Out of scope. This study does not estimate the share of all directory one-star reviews, does not score vendor quality overall, and does not evaluate private support tickets or churn data.
3. Methodology
3.1 Population and inclusion
The starting population is the WordPress.org plugins API list of the 100 most popular plugins by active installs. Plugins with fewer than 20 lifetime one-star ratings were excluded, leaving 77 plugins in the analytical sample.
3.2 Sampling
For each included plugin, up to the 20 most recent one-star review topics were collected from that plugin’s filtered review listing (?filter=1). Seventy-six plugins contributed a full 20 reviews. One plugin contributed 19. The design is equal-N across popular plugins. It is not a chronological sample of the global reviews forum.
3.3 Coding
Every review was coded against a fixed facet codebook (codebook v3). Coders assigned one or more facets and a single primary_facet for the main anger driver, plus an intensity label. Headline percentages use primary facet shares over the full corpus (N=1,539). Multi-label facet lists are published in the data package for secondary analysis.
3.4 Paid-signal subset (overlay)
As a derived overlay on the same corpus, reviews were scanned for first-person ownership or usage of a pro, premium, or paid tier (for example “I paid,” “we purchased the pro,” “using the Pro version,” “paying customer”). Mentions that only urge others to buy, or that describe aspirational upgrades, were excluded. This overlay does not change the main primary-facet ranking. Join file: paid_user_subset.json.
3.5 Field period
Reviews were fetched on 26 July 2026. Aspect coding was completed on 8 August 2026. Package generation timestamp: 2026-08-08T15:27:34Z.
| Parameter | Value |
|---|---|
| Popularity list | WordPress.org API browse=popular, top 100 |
| Star filter | 1 (one-star listing filter) |
| Lifetime minimum | ≥20 lifetime one-star ratings |
| Max reviews per plugin | 20 |
| Plugins included | 77 |
| Reviews collected | 1,539 |
| Reviews coded | 1,539 / 1,539 |
| Headline unit | primary_facet (one per review) |
4. Findings
The chart ranks primary facets (one main driver per review). Cluster cards below roll up those primaries. Multi-label co-occurrence counts ship in themes.json but are not the headline ranking.
Reliability leads when every review gets a primary facet
Primary drivers in this cluster include failed core jobs (264), site breakage (131), updates that broke the site (90), performance/bloat (70), regressions, lockouts, and harmful defaults.
Monetization is the second major cluster
Crippled free tiers (170) outrank ads and nags (102) as primary drivers. Billing pain, deceptive packaging, account lock-in, and wizard cross-sell complete the cluster.
“It doesn’t do the job” is the largest single primary facet
Reviewers describe a promised capability that fails without necessarily taking the whole site down: spam not blocked, mail not sent, restore failed, sync broken, cache useless.
Support, trust, and product experience are smaller primary clusters
Support failures are the primary driver in 83 reviews. Trust and safety (security fear, privacy, data loss) account for 103. Product experience (UX complexity, missing features, forced AI) accounts for 104. Residual keys (thin vents, plugin-local bugs, other) account for 170.
4.1 Subset: one-star reviewers who signal they paid
The main ranking above describes one-star reviewers in general. A smaller overlay asks what changes when the reviewer also says they paid for or use a pro or premium tier.
Paid disclosers are a thin slice, but they write as breached customers
Within those 44 reviews, the most common agent primary facet is Support failures (13). Disclosure bias applies: this is not a rate of paid users among all one-star reviewers, only among those who say so in the public review.
Free users often punish how you sell. Paid disclosers punish the promise after purchase.
One-star anger in general still includes strong monetization pressure, especially crippled free tiers. Paid-signal one-stars more often sound like “you already have my money” when support or the paid capability fails. The equal-N corpus still shows reliability as the modal story overall.
5. Evidence exhibits
Verbatim excerpts with public WordPress.org topic URLs. Exhibits illustrate primary facets. They are not a random sample of the corpus.
| Theme | Plugin | Topic | Excerpt |
|---|---|---|---|
| Plugin failed its job | Fluent Forms | Not Worth the Cost, Multi-Step Forms Don’t Work, Alternatives are Better | “We recently bought a subscription to the pro version . We’re switching from Forminator because it is missing things like confirmation emails (aka double opt-in) and built in support for custom field…” |
| Crippled free tier | CookieYes | This plugin went from useful to useless | “CookieYes plugin unfortunately went from being a very useful plugin to useless. A basic functionality such as translation is a must, so if you have a multilingual website and you are not willing to…” |
| Site breakage / WSOD | All-In-One Security (AIOS) | Don’t ever download this thing. You will regret it 100% | “If I could give this plugin a zero, I would. This plugin shut my website for weeks. It made me almost have a bad relationship with my hosting company. I was blaming them thinking they are the ones…” |
| Ads and nags in wp-admin | Solid Security | 183 emails in 12 hours – and they keep coming! | “I installed the plugin, and set up to receive newsletters related to the plugin. First day, 256 emails . I said ok, they sent me some to catch up. second day 324 Third day, in the first 12 hours,…” |
| An update broke the site | Spectra | Full of bugs – avoid | “Back and in a day I bought Astra’s Growth bundle which granted me access to all of their existing and future products (which did not prove to be the case) and what a mistake I made. After spending…” |
| Support failures | Rank Math SEO | Tech support for pro version is not good anymore | “I have been using rank math pro for a couple of years. The experience started out as ok but it is degarding quickly. First of all lasy year when I renewed my pro plan I got enrolled in content Ai…” |
| Feels like a scam / deceptive | Limit Login Attempts Reloaded | Fake brute force login lockout attempts | “I setup a basic WordPress website for my nephew using my web hosting service. It’s a one page partially created webpage that was just sitting around for three years unfinished. Last week I started…” |
| Performance and bloat | Royal Addons for Elementor | DO NOT BUY! DO NOT USE FREE VERSIO | “The plugin do not deserve even one star for performance, i was very excited because the visuals are very good, the header look so nice and you can make really good aesthetic websites BUT…. there is…” |
| Security fear / alerts | WPCode | It’s hacked. Virus infected, don’t use!! | “I read all the other similar reviews and the answers by the developers. To users : Don’t use this plugin! My website was infected by a virus that redirected to .top websites (space-robot), a known…” |
5.1 Paid-signal subset exhibits
These one-star reviews include first-person language that the reviewer paid for or uses a pro or premium tier. They illustrate the subset in §4.1. They are not folded into the main primary ranking above.
| Primary facet | Plugin | Topic | Excerpt |
|---|---|---|---|
| Support failures | WPForms Lite | Addon PDF doesn´t work | “Also they never answer the support tickets even i have a Pro version This topic was modified 7 months, 2 weeks ago by jaimesando .” |
| Support failures | WP Mail SMTP | Do Not Buy the PRO Version – Extremely Poor Support | “We purchased the PRO version of this SMTP plugin, followed all setup instructions correctly, and immediately began experiencing serious issues with email delivery, order…” |
| Site breakage / WSOD | All-in-One WP Migration | Bad product and customer service | “This used to be a good plugin, years ago. That’s when the good reviews come from. All the recent reviews show you it has become crap. Recently I purchased the pro version (the…” |
| An update broke the site | Spectra | Always Slow to Edit, Pro Version No Better | “We bought the pro version, and Spectra Pro is not what they promise. It is full of bugs. Be prepared for all kinds of new issues with your site in addition to the issues already…” |
| Forced AI features | Yoast SEO | Content marked by Yoast as “AI optimized”, but i do NOT use AI !! | “Lots of problems since the last updates.. Today I discovered surprisingly that Yoast adds this code to every headline or paragraph in Classic Editor: class=“ai-optimize” Mind…” |
6. Limitations
Results describe recent one-star reviews among popular plugins under an equal-N cap. They do not measure how common one-star reviews are across the entire Plugin Directory. Facets were assigned by agent coding against a frozen codebook, not by dual human annotation. Intensity and secondary facets are available in the data package but are not the headline ranking. The paid-signal overlay depends on reviewers disclosing purchase or pro use in public text, so it undercounts silent paid users.
| Limitation | Effect on interpretation |
|---|---|
| Equal-N per-plugin cap | Describes patterns among popular plugins. Does not estimate directory-wide prevalence. |
| Agent aspect coding | Labels follow codebook v3 with one primary per review. Not a human double-coded gold set. Prefer primary shares over multi-label floors unless stated. |
| English-leaning interpretation | Non-English reviews were still assigned facets when the driver was clear; residual non_english marks language when that is the main issue. |
| Public review text only | No private tickets, churn logs, or verified purchase context. |
| Paid-signal disclosure bias | The 44-review overlay is not a rate of paid users among one-star reviewers. Many paid users never say so. Do not treat ~2.9% as conversion or paid share. |
7. Data availability
The full coded corpus for this study is published with the report. Prefer the ZIP for a complete package, CSV for spreadsheets, and JSONL for scripts.
Full package (ZIP)
Corpus, codebook, sample frame, manifest, paid-signal subset, and README (CC BY 4.0).
Review corpus (CSV)
One row per review, including primary facet, multi-label themes, and public topic URL.
Paid-signal subset
Join file of reviews with first-person pro/premium/paid ownership language, plus subset insights.
License note: review text is public on wordpress.org. Coding, sample design, and this packaging are published by Roots & Fruit for reuse with attribution.
8. Facet codebook
Counts below are primary_facet assignments (one per review). Inclusion and exclusion notes ship in the data ZIP as codebook_v3.md; cluster labels also appear in themes.json.
| Key | Label | Cluster | Primary count |
|---|---|---|---|
| job_failed | Plugin failed its job | reliability | 264 |
| free_crippled | Crippled free tier | monetization | 170 |
| site_breakage | Site breakage / WSOD | reliability | 131 |
| upsell_nags | Ads and nags in wp-admin | monetization | 102 |
| update_broke | An update broke the site | reliability | 90 |
| plugin_specific_bug | Plugin-specific bug | residual | 87 |
| support_failed | Support failures | support | 83 |
| deceptive | Feels like a scam / deceptive | monetization | 75 |
| performance_bloat | Performance and bloat | reliability | 70 |
| security_fear | Security fear / alerts | trust_safety | 62 |
| ux_complexity | UX complexity | product_experience | 57 |
| thin_vent | Thin vent (little mechanism) | residual | 51 |
| missing_feature | Missing feature | product_experience | 38 |
| regression_used_to_work | Used to work / regression | reliability | 37 |
| wizard_crosssell | Wizard / setup cross-sell | monetization | 35 |
| refund_billing | Billing and renewals | monetization | 35 |
| data_loss | Data loss | trust_safety | 32 |
| other | Other / uncategorizable | residual | 32 |
| admin_lockout | Admin lockout | reliability | 29 |
| account_lockin | Forced account creation | monetization | 22 |
| default_on_harm | Harmful defaults | reliability | 19 |
| forced_ai | Forced AI features | product_experience | 9 |
| privacy_tracking | Privacy / tracking | trust_safety | 9 |