There’s a specific kind of frustration that comes with running a WordPress site. You publish something genuinely useful, wait, and watch it settle onto page four. So you install another plugin. You rewrite the title tag. Nothing moves.
Here’s the part most advice skips: a WordPress site usually fails to rank because Google can’t see, index, or understand the page. Not because the writing is bad. Rewriting your headings before you’ve confirmed the page is even in the index is like repainting a shop that nobody can find. My read is that the order matters more than the tactics: indexation first, technical health second, content and links third. Content is where the biggest upside lives, but it’s the last thing to fix, not the first.
WordPress runs 40.2% of all websites, according to W3Techs’ September 2026 survey, which means every problem you’re about to read about is shared by millions of other site owners. The good news is that the most common causes are also the cheapest to rule out.
Check whether your site is indexable before anything else
Start with the checkbox people are embarrassed to admit they missed. In WordPress, Settings → Reading contains an option labelled “Search engine visibility” (older versions phrase it “Discourage search engines from indexing this site”). Tick it and WordPress injects a site-wide noindex directive. It exists for staging sites, and it gets left on after a launch far more often than anyone likes to say.
Google’s own documentation on blocking search indexing is blunt about the consequence: a page carrying that directive gets dropped from results “regardless of whether other sites link to it.” No amount of quality writing or link building overrides it. I’ve watched people spend months on a content strategy for a site Google had been explicitly told to ignore.
Plugins add their own versions of the same switch. Yoast, Rank Math, and the rest let you noindex individual posts, categories, tag archives, and author pages. A migration, a theme switch, or a cloned staging database can flip those without anyone noticing.
Two checks settle it in minutes. Search Console’s URL Inspection tool tells you whether a specific URL is indexed and, if not, which reason Google logged. The Page Indexing report shows the same pattern across the site. As a rough cross-check, search site:yourdomain.com: if the count is nowhere near your number of real pages, or your homepage is missing entirely, stop here and fix it before touching anything else.
“Discovered” is not the same as “indexed”
Pages that pass the noindex check still sometimes fail to appear. Search Console labels them Crawled – currently not indexed or Discovered – currently not indexed, and those two labels mean very different things.
- Discovered, not indexed: Google knows the URL exists but hasn’t spent crawl budget fetching it. Usually a sign the site is too large, too slow, or too low-authority to earn a visit.
- Crawled, not indexed: Google fetched the page and still chose to leave it out. This is the more interesting failure, because the page loaded fine, but Google judged it a near-duplicate, too thin, or less useful than what already ranks.
That second one is where most “my content is good but it won’t rank” cases actually live. If you have ten blog posts answering the same question slightly differently, you’re competing with yourself, and Google will pick one and ignore the rest. Merging them into one strong page and redirecting the leftovers usually does more than publishing an eleventh.
The version Google ranks is your mobile version
Since mobile-first indexing became Google’s default, it uses the mobile version of a page, crawled with its smartphone agent, for indexing and ranking. Your desktop version can be richer, but it isn’t the one being assessed.
This bites WordPress sites in a predictable way. Page builders like Elementor, used on 31.5% of WordPress sites per W3Techs, often tuck content inside tabs, accordions, or sliders on mobile. Google generally won’t load content that requires a swipe or a click to reveal. If half your service page is hidden behind an accordion on a phone, half your service page may as well not exist. The fix isn’t deletion. Keep the content in the HTML and let CSS collapse it visually, so it stays in the markup even when it’s hidden on the screen.
Speed is real, but it isn’t the lever you think
Core Web Vitals matter. They’re just not the first domino. A fast site that isn’t indexed ranks for nothing, and a slow site with genuinely useful content can still rank in plenty of niches. The order is what trips people up.
| Metric | What it measures | “Good” threshold |
|---|---|---|
| LCP | Loading | 2.5 seconds or less |
| INP | Interactivity | 200 milliseconds or less |
| CLS | Visual stability | 0.1 or less |
Those Core Web Vitals thresholds apply at the 75th percentile of real page loads, measured separately on mobile and desktop, so one clean lab test doesn’t mean you pass. One more detail worth knowing: INP replaced the old First Input Delay metric in 2024, and a surprising amount of WordPress performance advice still references FID, which no longer exists as a Core Web Vital.
Where WordPress sites actually lose speed is mundane. Ten plugins each loading their own CSS and JavaScript. A hero image exported at 4,000 pixels wide. A caching plugin fighting a CDN that’s also trying to cache. I’d audit and remove before I’d optimize: every plugin that isn’t earning its keep is a render-blocking request you didn’t need in the first place.
The uncomfortable truth about content
Google’s guidance on creating helpful, reliable, people-first content asks a question worth sitting with: does the page provide original information, reporting, research, or analysis? A page that restates the top five results with better grammar fails that test, and no amount of on-page optimization rescues it. The same documentation states plainly that Google has no preferred word count, which is worth remembering the next time someone tells you to hit 2,000 words.
The WordPress-specific trap is the service page. Many businesses pour energy into blog posts while their actual money pages are three generic sentences and a contact form. If you sell one thing, the page about that thing should be the most complete resource on the internet for it, not an afterthought behind the blog.
Authority is the part WordPress can’t fix for you
Everything above is about being eligible to rank. Ranking itself is a competition, and the deciding factor is usually who else on the web thinks you’re worth citing. WordPress neither helps nor hurts here. Internal linking does more than most people realize, because it spreads authority and teaches Google how your site is structured, but external signals arrive through links, mentions, and branded searches that no plugin can manufacture.
The order I’d actually work in
- Confirm the site is indexable and that no page is noindexed by accident.
- Read the Page Indexing report and deal with the “crawled / discovered, not indexed” pages.
- Make sure mobile and desktop carry the same content.
- Trim plugins, compress images, and fix the caching stack.
- Only then rewrite the thin pages that actually deserve the traffic.
- Build internal links, then earn external ones.
The order matters because steps one and two can be fixed in an afternoon, and they make everything else irrelevant when they’re broken. Step five takes months. Collect the cheap wins first.
When to bring someone in
The first three steps are doable for anyone comfortable in Search Console. People get stuck in the middle instead: crawling problems that only show up in server logs, canonical tags that contradict each other, a plugin conflict quietly loading two copies of jQuery. If that’s where you keep losing weekends, handing it off is reasonable. A specialist doing seo for wordpress work will start with crawl and indexation data rather than a content rewrite, which is the same order this article recommends. If a proposal opens with “we’ll write 20 blog posts,” ask to see the indexation picture first.
Frequently asked questions
Why did my WordPress site rank and then suddenly stop?
Check the dates against the drop. A ranking fall that lines up with a Google core update, a plugin update, a host migration, or a theme change usually has a traceable cause. Search Console’s performance report will show whether impressions fell across the whole site or only in one topic area. A site-wide drop points to something technical; a topic-specific drop points to content or new competition.
Does WordPress have SEO built in?
Partially. WordPress produces clean URLs, supports titles and meta descriptions, and handles canonical tags on archives reasonably well. It does not evaluate content quality, control your caching, or stop a plugin from adding noindex. It gives you a solid foundation and no guarantees.
Do I need an SEO plugin?
One is useful. Several are not. A single plugin handles titles, meta descriptions, sitemaps, and schema. Installing two that both manage the same tags is a common source of contradictory canonicals and duplicated meta tags.
How long until a fix shows results?
A technical fix can show movement within days if indexation was the problem, because you changed what Google can see rather than how it judges you. Content and authority changes typically take weeks to months and often land with the next core update. Anyone promising a two-week turnaround on a competitive keyword is guessing.
Is a slow WordPress theme killing my rankings?
Probably not on its own. Theme weight contributes to LCP and INP, especially with page builders, but it sits behind indexation and content quality as a cause. If you’re indexed, mobile matches desktop, and your content is strong, then theme performance can be the tiebreaker. Rarely before that.
How this article was put together
I wrote this against Google’s current Search Central documentation on mobile-first indexing, noindex, and people-first content, plus the Chrome team’s Core Web Vitals thresholds, all checked in September 2026. The WordPress and page-builder market figures come from W3Techs’ live survey on 25 September 2026. Core Web Vitals are a stable metric set reviewed roughly annually, so the thresholds above should hold through 2026; recheck them if a new metric is announced. Where I’ve given an opinion about the order to fix things, that’s judgment, not a Google statement.
