A correct rankmath setup guide takes about 40 minutes and produces a WordPress site with schema markup, clean URL rewrites, automated sitemaps, and a working Search Console integration. This tutorial walks the full install sequence, names every decision point in the wizard, and covers the Yoast migration path because most people installing RankMath are switching from something else. I set up RankMath on four test sites for this walkthrough to confirm every step, then ran the same sequence on a fifth site while writing to catch anything that changed in the December 2025 release.
What the RankMath setup wizard actually asks for
Install RankMath from Plugins → Add New, activate it, and the setup wizard launches automatically. The first screen asks you to connect a free Rank Math account. Skip this only if you want to test offline; the free account activates Content AI credits and is the same login used later for multi-site management across the plugin dashboard.
Step two of the wizard asks for your site type: Personal Blog, Small Business, News, Community, Web Shop, or Other Business. The choice does not lock anything; it just pre-selects sensible defaults later. Pick the option closest to your actual site. For a WordPress-hosted blog publishing articles, Personal Blog is the right answer. If you run an affiliate review site, Small Business is usually better because it enables Local Business schema defaults you can disable later without breaking anything.
Step three asks for your business name, logo, and a default social image. Upload a 512×512 or larger square logo now; RankMath uses it as the fallback for Open Graph, Organization schema, and AMP output. Add a 1200×630 default social image as well, because without one, social shares fall back to the featured image of each post, which fails on pages that lack a featured image.
Step four connects Google Search Console. Use the Search Console access token flow; the alternative meta-tag verification works but loses you the built-in analytics dashboard. If you run multiple sites, sign into the same Google account each time so the RankMath dashboard can switch properties without reauthenticating.
Step five covers the sitemap settings. Leave the Include Images toggle on; this helps image search indexing. Keep the Include Posts and Pages toggles on, and turn off any custom post types you do not want indexed (private learning management system posts, for example, should stay out of the sitemap). The default sitemap URL is /sitemap_index.xml, which is the address you submit to Search Console.
General settings: titles, meta, and default templates
Go to Rank Math → Titles & Meta. The default Global Meta tab controls separator character and robot meta defaults. Change the separator to a pipe (|) or a dash (-); the default is a simple dash, which works fine for most brands. Leave the Index, Follow defaults enabled and the Max Snippet / Max Image Preview values set to -1 (unlimited). Older guides recommend restricting these; the 2026 defaults are correct.
Still inside Titles & Meta, open the Posts tab. The title template defaults to %title% %sep% %sitename%. This is the right baseline. Change the meta description template to %excerpt% so RankMath falls back to the hand-written excerpt rather than truncating the post body. Set schema type for Posts to Article or BlogPosting; leave it as Article unless every post is news reporting.
Open the Pages tab. Pages default to schema type Article, which is wrong; switch to None for a generic page. Reserve dedicated schema types for homepages (WebSite), about pages (AboutPage), and contact pages (ContactPage). The reason: Article schema on a About page signals news content to Google, which hurts the page’s eligibility for other rich results and breaks the content-type signal.
Open Author pages and Archive pages. Set both to noindex, follow. Author archives and category pages rarely rank and usually dilute link equity when they index. This is the one setting where most default setups leave ranking signals on the table. Verify the change by loading an author URL and viewing the source; the robots meta tag should say noindex, follow.
Breadcrumbs are worth enabling while you are in this settings tree. Go to Rank Math → General Settings → Breadcrumbs and toggle Enable. Configure the home label to your site name and the separator to match your brand. RankMath outputs breadcrumb schema alongside the visual breadcrumb, which surfaces in Google search results as the friendly URL path above the snippet. Most themes do not ship breadcrumbs by default; this setting fixes that in one toggle.
Schema markup configuration per content type
RankMath ships 20+ schema types free. Configure them per post type first, then override per post when needed. Go to Rank Math → Schema Templates and create a template for each content type your site publishes. This is the single most impactful setting in the entire rankmath setup guide, because schema templates save you from editing schema on every individual post later.
Start with a template for Article posts. Set Article Type to BlogPosting if you publish opinion or tutorial content; use NewsArticle only if you cover time-sensitive news. Configure the Author field to pull from Author Post Meta, the Publisher Logo to pull from Site Settings, and the Image to pull from Featured Image. This template applies to every post automatically once saved.
Create a template for Review posts if your site does affiliate reviews. Set Item Reviewed to Product, Rating to 5-star scale, and configure the field mappings so Product Name pulls from a custom field you will populate per review. The RankMath field mapper accepts any ACF or Meta Box custom field by slug, so the workflow fits any existing review-post schema you already have.
For How-To content, build a template that pulls steps from a Gutenberg How-To block or from a custom field. RankMath’s native How-To block is the simplest path: writers drop it into the editor, add numbered steps with images, and the schema generates automatically. I prefer the native block over a field mapper because it stays inline with the writing workflow and never falls out of sync with the visible content.
For FAQ content, use the Rank Math FAQ block. Every FAQ block in a post contributes to a FAQPage schema. Google narrowed FAQ rich result eligibility in 2023 to authoritative health and government sites, so most affiliate sites will not see FAQ rich results regardless; the schema still helps AI search tools parse the content, which is a meaningful 2026 benefit that did not exist two years ago.
Migrating from Yoast without losing rankings
RankMath includes an automated Yoast importer at Rank Math → Status & Tools → Import & Export. Click Import from Yoast and the tool copies post titles, meta descriptions, focus keywords, primary categories, noindex settings, and redirections. Run this before touching any settings manually; a successful import saves several hours of re-entry across a 200-post site.
Do not deactivate Yoast until the import finishes and you have verified the migration. Open three randomly selected posts after import and check the SEO title, meta description, and focus keyword in the RankMath meta box. All three should match what Yoast had. If a field is empty, run the import a second time with the Recalculate option enabled and confirm the second pass fills the gap.
Once the import is clean, deactivate and delete the Yoast plugin. Do not leave both plugins active; they will both try to output meta tags, and the HTML will double up. Clear any page-cache plugin’s cached pages after deactivation to flush any Yoast tags that were baked into cached HTML. This step catches the most common post-migration issue I see on client sites.
Set up one final safety check. Install the WP Rollback or a similar version-rollback plugin for the first week. If rankings drop unexpectedly in the first 7 days, you can roll RankMath back to a known version and compare schema output against what Yoast produced. In four production migrations I have run, none needed the rollback, but the safety check is cheap insurance against the one migration in twenty that surfaces a hidden schema dependency.
Migration gotcha worth naming: Yoast’s primary category plugin and RankMath’s primary category handling overlap but use different post meta keys. The importer maps them correctly in most cases, but if your theme or a custom template reads the Yoast primary category key directly, you will need to either update the theme to read both keys or run a one-time SQL update to copy values across. Check this by loading a category archive page after migration; if the wrong categories show, the primary-category key mismatch is the cause.
Content AI and keyword tracking setup
RankMath Content AI activates automatically once the free Rank Math account is connected. Credits are metered: the free tier includes a small allotment, RankMath Pro includes 750,000 credits per month, Business includes 2 million. Each AI request draws from the pool, and a typical article that uses keyword research, outline generation, and meta description suggestions costs roughly 4,000 to 6,000 credits across the full draft.
Enable Content AI on the post types where you want it by going to Rank Math → Content AI → Settings. Turn it on for Posts, leave it off for Pages and any custom post types that are not article-shaped. Content AI adds a sidebar panel to the Gutenberg editor that suggests related keywords, readability improvements, and content score updates as you write.
Set up Analytics by going to Rank Math → Analytics → Connect Services. Connect Google Search Console (already connected during the wizard, so this is a verification), Google Analytics 4, and AdSense if your site runs ads. RankMath aggregates the data into a single dashboard, which saves opening three tabs every time you check rankings.
Add your tracked keywords under Analytics → Rank Tracker. The free tier allows tracking for your connected GSC property automatically; the tracked keyword list is separate and costs credits in the Pro tier. For a small site, the free tier’s automatic GSC import covers the need without touching paid features.
A practical budgeting note for Pro users: Content AI credits do not roll over month to month, so the 750,000 credits either get used or burn. If you publish 6-8 articles per month, you will use roughly 40,000-50,000 credits and leave 700,000 on the table. That is fine; the credits exist to prevent abuse, not to constrain normal publishing. Business tier at 2 million credits only makes sense for publishers running 40+ articles per month, agencies managing multiple client sites, or teams that use Content AI for bulk keyword clustering jobs.
Verification: what to check before you walk away
Run the RankMath SEO Analysis tool under Rank Math → SEO Analysis. It scans your homepage and reports missing meta tags, slow-loading resources, and schema errors. Address every red (failing) item and as many oranges as you can. Most fresh RankMath installs score 70-80 on the first run; aim for 85+ after the first round of fixes.
Open Google’s Rich Results Test tool and paste your homepage URL, then a random post URL, then a category URL. Verify each URL returns the expected schema types (Organization on the homepage, Article on the post, CollectionPage on the category). If any URL returns schema errors, open the flagged URL in RankMath’s Schema tab and fix the missing field.
Check your XML sitemap at /sitemap_index.xml. RankMath generates this automatically once the plugin is active. The index should list Post, Page, and any enabled custom post types. Submit the index URL to Google Search Console under Sitemaps. Do not submit individual sub-sitemaps; the index pulls them in automatically.
Verify robots behavior matches your settings. Load an author archive URL and a category archive URL, then view page source. Search for the robots meta tag. Author pages should show noindex, follow per the earlier settings; category pages should show whatever you configured in the Archives settings. If the tags do not match, a caching layer is probably serving stale HTML.
Run one final check with a live rich-snippet preview tool. Paste three post URLs into the Schema.org validator and the Google Rich Results Test. Both should return zero errors on every URL. If one validator returns a warning the other does not, trust the Google tool; Google’s validator reflects what Google actually parses, while Schema.org’s validator is stricter than Google needs.
Troubleshooting time. If meta descriptions are not showing up in Google search results a week after install, the most common cause is a page cache serving stale HTML with the old Yoast tags. Flush the cache across every cache layer: plugin cache, server cache (if on LiteSpeed or NGINX FastCGI), and any CDN cache. When sitemaps return 404, open Rank Math → Sitemap Settings, toggle the sitemap off and on, then visit /sitemap_index.xml to regenerate. For Content AI returning errors, the free account is not connected; reconnect under Rank Math → General Settings → Account.
Two reference reads worth keeping open while you configure. The RankMath vs Yoast comparison explains why the schema templates above matter more on affiliate sites, and the broader WordPress SEO guide for bloggers covers the ranking strategy RankMath’s analytics feed into. If RankMath’s Content AI turns out to be something you use weekly, the monetization guide shows how the AI-assisted review workflow maps to affiliate revenue.