You hit publish. You paste the URL into Slack, LinkedIn, or a group chat. The preview shows the wrong title, a random image from the sidebar, or nothing. Then you scramble through Yoast, Rank Math, and the theme header while the post is already live.
Meta tags are small. The embarrassment is not. A meta tag checker, and a habit of looking at the social preview before you share, saves that. Nicxro’s checker is for reading what the page actually emits, not what the editor sidebar promised.
This is how to inspect titles, descriptions, canonicals, robots, and Open Graph tags like they are part of the page, because they are.
What you are checking, in plain terms
Search engines and social platforms do not “see” your WordPress preview. They fetch HTML and read tags in the head.
Common ones:
title is the browser tab and a default search title.
meta name="description" is a suggestion for the search snippet. Google may ignore it.
link rel="canonical" says which URL is the real one.
meta name="robots" can noindex a page you meant to rank.
Open Graph (og:title, og:description, og:image, og:url) drives many social previews.
Twitter/X has its own tags and often falls back to Open Graph.
If those tags are missing, duplicated, or still say “Hello world,” the preview will look like a stub. If og:image is too small, some networks skip it. If the image is blocked by robots or login, they skip it.
Do not trust the editor. Trust the HTML.
SEO plugins show a preview based on their fields. Themes override. Caching plugins serve old head tags. A CDN caches the HTML. Multilingual plugins emit two canonicals.
View the live URL. Nicxro’s meta tag checker should fetch the public page. Compare:
- Title in the plugin vs title in the HTML
- Description vs what you wrote
- Canonical vs the URL in the address bar
- og:url vs canonical
- og:image as an absolute URL, not
/wp-content/uploads/x.jpgonly, if the platform needs absolute. Most OG images should be fullhttps://...URLs.
If the checker and “view source” disagree with the plugin, the plugin is not winning. Fix the HTML.
Titles that work in a tab and in a feed
A title that is a full thesis will truncate. Front-load the useful words. Put the brand at the end if you include it: How to Format JSON | Nicxro, not Nicxro - Tools - Blog - JSON.
Do not write ALL CAPS. Do not stuff keywords until it reads like a tag cloud.
Every important URL needs its own title. If pagination, filters, or UTM copies share one title, search results look duplicated even when canonicals are correct.
Descriptions that earn the click
The description is an ad for the page. It should say what the reader will get. It should not repeat the title. It should not be empty. Empty descriptions let the engine pick a random sentence, often a cookie banner or a menu.
Length is a guideline, not a scoring bonus. Write a complete thought. Cut fluff.
If you localize, translate descriptions. A English description on a Spanish page looks careless in previews.
Open Graph images are a product decision
Pick an image that still works at a 1.91:1 crop. Faces in the center. Text on the image should be large. Tiny blog screenshots become gray blobs.
Recommended sizes change, but a 1200×630 PNG or JPEG is still a workhorse. Compress it so the fetch is fast. A 5MB OG image can time out.
The image URL must be publicly fetchable. No login. No hotlink blocking that treats Facebook’s crawler as a thief if you actually want Facebook to show the image. og:image:width and height help some platforms.
After you change the image, social networks cache. Facebook’s debugger, LinkedIn’s inspector, and similar tools can refresh. Nicxro shows what you serve now. The chat app may show yesterday. Plan for cache when you fix a bad preview on a post that already spread.
Canonicals and the duplicate URL swamp
https://nicxro.online/page and https://nicxro.online/page/ and https://www.nicxro.online/page can be three URLs. Pick one host and one slash policy. Canonical should match.
UTM parameters should not create indexed copies. Canonical to the clean URL. Do not noindex the page just because ads use UTMs; noindex would drop the good URL if you put it on the template by mistake.
Facets and search-result pages often should be noindex. Money pages should not. A checker that shows noindex on a launch URL is a stop-the-press moment.
robots.txt vs meta robots
robots.txt can block crawlers from a path. Meta robots noindex a page they can fetch. If you block in robots.txt, they may not see the noindex. If you wanted to remove a page from search, that combination is messy.
The meta checker reads the page you gave it. Also glance at /robots.txt. A disallow on /blog has happened on sites that then wondered why articles did not rank.
Local tools, staging, and “why is the preview my login page”
If the URL redirects to login, the crawler gets the login HTML. og:title becomes “Log in.” Staging sites with basic auth do the same.
Check production URLs with the checker. For staging, allow the crawler or test tags in source while logged out in a private window.
If a page is client-rendered and tags are injected late by JavaScript, some crawlers will miss them. WordPress normally prints tags in the first HTML. SPAs need server-rendered tags or prerender. If Nicxro’s fetch shows no OG tags and “view source” also has none, your React app never put them in the document. Fix the server template.
A pre-publish checklist
- Final URL, not a preview nonce URL.
- Run Nicxro’s meta tag checker.
- Confirm title, description, canonical, robots.
- Confirm og:title, og:description, og:image, og:url.
- Open the og:image URL in a private window.
- Share to a private Slack channel or a test tweet if the post is important.
- Then share publicly.
For a homepage and a few key landing pages, save this as a monthly check. Plugins update. Themes update. Tags vanish.
Common WordPress-specific failures
Two SEO plugins at once. Duplicate titles, duplicate canonicals.
The homepage title left as the site tagline plus “Just another WordPress site.”
A default og:image that is the site logo at 200×60, too small for feeds.
Pagination tags missing on /page/2, or worse, page 2 canonicalized to page 1 when you wanted unique archives. Know what you want.
WooCommerce products inheriting the shop description.
Title and description that fight each other
If the title already says “JSON formatter,” the description should not open with “JSON formatter” again. Use the description for the job: format, validate, copy, no account. Search snippets are a conversation. Repeating the title wastes the only extra sentence you get.
If the description is a slogan (“The best tools for winners”), you will lose to a competitor who named the task. Nicxro should describe the task. Checkers help you see when a plugin defaulted back to a slogan after a theme update.
hreflang, language, and the wrong preview locale
og:locale and html lang should agree with the content. A French article with English OG tags will look wrong in a French Facebook feed. Multilingual plugins sometimes emit every locale’s URL as canonical. The checker will show a pile of hreflang links. If they point at 404s, fix the plugin before you chase rankings.
Why this belongs on a tools site
You already format JSON and check responsive layouts. Meta tags are the same class of problem: the machine-readable layer of a page. Humans see the design. Crawlers see the head.
Nicxro is for looking at that layer before you announce the URL. The five minutes you spend checking tags are cheaper than the day you spend explaining why LinkedIn shows a stock photo of a handshake on a technical article.