Topical Authority 2026: How to Plan Topic Clusters for a Blog

Forty posts on one subject, and Google ranks three of them. Sound familiar?

Topical authority comes from covering one topic completely and linking the pages together, not from publishing more. Plan each topic as one pillar page plus 6 to 15 cluster pages, one per search intent, linked both ways. A spreadsheet is enough to start.

Our own SEO section failed that test, for the reason most topic clusters fail.

We earn a commission from the Semrush, Surfer and Scalenut links. It never decides what we recommend.

What does topical authority actually mean?

Topical authority is the trust a site earns on a subject by answering most of its questions on pages that clearly belong together.

It isn’t a score you can look up. Third-party tools show their own estimates, but Google doesn’t publish one.

Here’s the part most guides skip. Google’s only system officially called topic authority is a news feature. Google described it in May 2023 as a way to surface expert publishers on news searches about health, politics or finance. A blog about SEO tools isn’t the target.

So what helps a blog? Two things Google does document. Its ranking systems read how pages link to each other to work out what each page is about. Site-wide signals also feed into how single pages are judged.

Cover a topic well and link the pages sensibly, and both work in your favour.

Our take: treat topical authority as a planning method, not a ranking factor. The output is a site where every page has one job. A reader gets from the broad guide to the exact answer in one click.

Pillar, cluster, link: what each page has to do

A topic cluster is one broad pillar page, several narrow cluster pages, and the links that tie them together.

Notice the money page. On a review blog, that’s where the cluster pays off. A reader lands on a tutorial, trusts it, and clicks through to the review when they’re ready to buy.

How big should a cluster be? Big enough that a reader never leaves your site for the next obvious question. For most blog topics, that’s 6 to 15 pages.

A cluster of two is a post with a friend. A cluster of 60 usually means one intent got split into five near-identical posts.

How to plan a topic cluster in six steps

You can plan a cluster in an afternoon with a spreadsheet. The tools in the next section speed up steps 2 and 3, but they don’t change the method.

Step 3 is where most plans go wrong. Two queries that look different, like topic cluster and pillar page, often show the same results. Write two posts and they compete. Our content gap analysis guide is the place to find the missing queries before you group them.

We audited our own SEO cluster. Here’s what broke

We ran this check on DroidCrunch before telling you to run it on your site. On 30 October 2026 we counted every in-body internal link across our 596 published posts. Then we looked closely at the 35 filed under SEO.

The short version: we built a lot of good pages and forgot to connect them.

Look at the Semrush review. Eleven posts send readers to it. It sends them nowhere. A reader who finishes the review and wants the pricing breakdown or the keyword research tutorial has to go back to Google.

Then the category problem. Our SEO category held a Reply.io review and a QuillBot review, while our main keyword research tools roundup sat under Blogging.

A WordPress category is a menu. A cluster is made of links.

The script we used to check one cluster

Database queries are fine if you have them. For everyone else, this script reads the live pages and checks the two links that matter: pillar to cluster page, and back. It needs only Python 3.

"""Check that a pillar page and its cluster pages link to each other."""
import re, sys, urllib.request
from urllib.parse import urljoin, urlparse

START = ('class="entry-content', 'class="content-inner', "<main")
END = ("jeg_share_bottom", "jnews_prev_next", "jnews_related", "<footer", "</main>")
UA = {"User-Agent": "Mozilla/5.0 (cluster-audit)"}

def norm(url):
    p = urlparse(url)
    return (p.netloc.lower().removeprefix("www.") + p.path.rstrip("/")).lower()

def links_on(url):
    req = urllib.request.Request(url, headers=UA)
    html = urllib.request.urlopen(req, timeout=20).read().decode("utf-8", "ignore")
    # Count links in the post body only, not menus, sidebars or related posts.
    # Adjust the markers to your theme (these fit most WordPress themes).
    starts = [i for i in (html.find(m) for m in START) if i >= 0]
    start = min(starts) if starts else 0
    ends = [i for i in (html.find(m, start) for m in END) if i > 0]
    body = html[start:min(ends)] if ends else html[start:]
    hrefs = re.findall(r'href="([^"#?]+)', body)
    return {norm(urljoin(url, h)) for h in hrefs}

def main(pillar, spokes):
    pillar_links = links_on(pillar)
    missing = 0
    print(f"{'cluster page':45} pillar->page  page->pillar")
    for s in spokes:
        down = norm(s) in pillar_links
        up = norm(pillar) in links_on(s)
        missing += (not down) + (not up)
        print(f"{urlparse(s).path:45} {'yes' if down else 'NO':12}  {'yes' if up else 'NO'}")
    print(f"\n{missing} missing links out of {len(spokes) * 2}")

if __name__ == "__main__":
    main(sys.argv[1], sys.argv[2:])

Run it with the pillar first, then the cluster pages: python3 cluster_audit.py https://yoursite.com/pillar/ https://yoursite.com/page-1/ https://yoursite.com/page-2/. The START and END markers fit most WordPress themes. If every row says NO, view your page source and change them to whatever wraps your post body.

Our results were blunt. With the Semrush review as the pillar and five Semrush-related guides as the cluster, 7 of 10 links were missing. With the keyword research roundup as the pillar and six related posts, 10 of 12 were missing. In both, the pillar linked down to zero pages.

The fix is cheap. Add a short “where to go next” passage to each pillar, then a link up from every cluster page that lacks one. It’s an hour of editing per cluster, and it’s the first job on our own list.

Which tool should build your topical map?

You don’t need a tool to plan one cluster. You want one when you’re planning five at once or need search volume on every row. Every price below was checked on 30 October 2026. We haven’t run these three on a live project for this guide, so the workflows come from each vendor’s own help docs.

Semrush Keyword Strategy Builder

Give it up to five seed keywords and it groups up to 10,000 related keywords into topics, pillar pages and subpages. Each page shows its keywords, difficulty, volume and the sites already ranking. You can export the lot to a spreadsheet.

It comes with the SEO plan, which runs about $140 monthly ($117.33 if you pay for the year). That plan gets 10 structuring runs a month, plenty for one site. Agencies on Pro+ ($299 a month) get 30.

Pick Semrush if you already use it for keyword research, because the clusters land next to the rest of your data. Our Semrush keyword research tutorial covers the seed-keyword part.

Surfer Topical Map

Surfer’s Topical Map starts from your domain, not a seed list. Connect Search Console and it shows which clusters you already cover, which neighbouring clusters you don’t, and how a competitor’s coverage compares. It refreshes every 14 days.

The catch is the plan. Topical Map needs Pro, at $182 a month billed yearly, which is about $2,184 upfront. Discovery and Standard don’t include it. Surfer also said in October 2026 that a rebuilt Topical Map is in testing, so the screens may change.

Pick Surfer if you’ll write inside its Content Editor anyway, since one subscription covers planning and writing. Our Surfer SEO review covers the editor side.

Scalenut Keyword Planner

Scalenut’s Keyword Planner turns one keyword, or an upload of 10 to 200 keywords, into AI-made clusters. Each cluster lists its primary keyword, intent, volume and difficulty, with a button to start an article from it.

It’s the cheapest way in. Starter costs $59 a month and builds 5 clusters a month. Plus, at $89, raises that to 30 and adds internal link suggestions. Scalenut was also running a 60% off offer when we checked, so look at the checkout price.

Pick Scalenut if you’re a solo blogger planning a cluster or two a month on a small budget. Our Scalenut review goes through the writing tools.

Our pick for most bloggers: plan the first cluster by hand. You’ll learn what a good grouping looks like, and you’ll judge any tool’s output better afterwards.

Five mistakes that quietly undo a topic cluster

Most clusters don’t fail because of a bad plan. They fail in the year after it, one small shortcut at a time.

• One post per keyword. If pillar page examples and what is a pillar page share a results page, two posts split the clicks. Merge them and redirect the weaker URL.
• Mass-produced cluster pages. Fifty thin AI drafts in a week is what Google’s scaled content abuse policy targets. Publish fewer pages you’d defend line by line.
• A pillar that only receives links. Our Semrush review shows how this looks. The pillar has to send readers down, or it’s a dead end.
• Treating categories as clusters. A category archive is a list, not a map. Clusters live in in-body links.
• Topics with no business behind them. Traffic on a subject that never leads to a review, product or signup is a cost you pay to host.

AI search raises the stakes on the first two. Answer engines quote one clear page per question, so a cluster with one page per intent is easier to cite. Our guide to generative engine optimization covers what those systems look for.

How do you know a cluster is working?

Track the cluster as one unit in Search Console, not post by post.

Filter the Performance report by page with a regex that matches the cluster’s URLs, such as semrush|keyword-research. Then compare the last three months with the three before.

Watch three numbers: total impressions, how many queries the cluster shows up for, and how many of those rank in the top 10.

Expect impressions to move first and clicks last. A query count that grows while positions sit at 8 to 20 means Google is testing the cluster. That’s the moment to improve the pages, not to start a new cluster.

Seeing your pages pulled into AI answers too? The method in our guide to ranking in Google AI Overviews works page by page inside a cluster you’ve already built.

Frequently asked questions

Map one cluster this week

Topical authority isn’t something you buy with volume. It’s what a reader feels when every next question already has a page, and the link to it is right there. Plan one topic cluster: a pillar, the strongest three pages, and links in both directions.

Use a spreadsheet for the first one, and bring in Semrush, Surfer or Scalenut when you plan several at once. One warning: don’t skip the links, or you’ll build a pillar like our Semrush review.

Your next step? Run the script on your biggest pillar page tonight.