Comparison Matrices for SEO: Decide What to Write, Update, or Kill

There’s this moment every SEO person hits.

You open Search Console, flip through your content list, and realize you have… too many pages. Some are doing great. Some are doing nothing. Some are doing that annoying thing where they kind of rank, kind of get traffic, but never convert, and you keep telling yourself you’ll “fix them next week”.

And then next week becomes six months.

Comparison matrices are how you stop guessing.

Not the fancy MBA ones. Just a simple, brutally honest grid that tells you, page by page, whether you should:

• write something new
• update what you already have
• merge pages
• redirect and kill the dead weight

And yes. You can do this in a spreadsheet. Or you can automate most of it. But first, you need the logic.

Let’s build the matrix.

What a comparison matrix actually does (in SEO terms)

A comparison matrix is basically a decision system.

Instead of “I feel like we should update this blog post”, you’re saying:

• This page has impressions, low CTR, decent position. That is a snippet or title problem. Update.
• This page has traffic, but no conversions and overlaps with a better page. Merge.
• This page has zero impressions for 3 months and no links and no obvious business value. Kill.

It’s not magic, but it’s consistent. And consistency is what most content teams don’t have.

The hidden bonus: once you run this process once, you can run it every month. Suddenly content pruning and content planning stop being dramatic projects and become maintenance.

The 3 buckets: write, update, kill (plus the ones nobody talks about)

Most people do “write, update, kill”. That’s a good start.

But in real life you also need:

• Merge: two or more pages compete or overlap.
• Hold: page is fine, not urgent, maybe seasonal, maybe still climbing.
• Reposition: same page, different intent. The content is okay but aimed at the wrong query.

So the practical buckets are:

1. Write
2. Update
3. Merge
4. Hold
5. Kill (redirect or 410, depending)

Your matrix helps you put every URL into one bucket without arguing about it for an hour.

The inputs you need (keep it simple)

You can build a matrix with a ton of metrics. Don’t. You’ll never finish.

Start with these. They cover 90 percent of cases.

From Google Search Console

• Impressions (last 28 or 90 days)
• Clicks
• CTR
• Average position
• Top queries (and whether they match intent)

From analytics (GA4, Plausible, whatever)

• Sessions
• Engagement (time, scroll, engaged sessions)
• Conversions (or assisted conversions)
• Landing page value, if you track it

From your site/content inventory

• URL
• Content type (blog, landing page, comparison page, docs)
• Publish date / last updated
• Primary keyword (if you have it)
• Topic cluster / category
• Notes: outdated? thin? off brand? legally risky? (yep, it happens)

Optional but useful if you have it:

• Referring domains / backlinks
• Internal links count
• Indexation status / canonical issues

If you want a quick way to sanity check on page quality, an on page audit tool helps. Something like an on-page SEO checker can catch obvious issues you just won’t see in a spreadsheet, like missing headings, weak internal linking, title problems, and so on.

Step 1: Build your content inventory (without overthinking it)

Make a sheet with one row per indexable URL.

Columns like:

• URL
• Page type
• Topic
• Target keyword
• Last updated
• Impressions (90d)
• Clicks (90d)
• CTR
• Avg position
• Sessions (90d)
• Conversions (90d)
• Backlinks (Y/N or number)
• Notes
• Decision (Write/Update/Merge/Hold/Kill)
• Priority (High/Med/Low)

That’s it. Seriously.

If you have 2,000 URLs, don’t try to do all at once. Start with your top 200 by impressions, or top 200 by business importance. You’ll usually find the biggest wins there anyway.

Step 2: The core matrix (the one you’ll actually use)

Here’s a practical comparison matrix you can copy.

Matrix A: Performance vs potential

Think of two axes:

• Performance now: traffic, impressions, rankings, conversions
• Potential: keyword opportunity, business value, ability to improve

Now the decisions:

This matrix is simple, but it forces the real question: even if a page is doing badly, is it worth saving?

A lot of pages aren’t. And that’s okay.

Matrix B: GSC pattern matrix (fast decisions)

This one is my favorite because it’s based on patterns you can spot in 30 seconds.

If you want to go deeper on fixing the page itself once you’ve flagged it, run it through a workflow like an AI SEO editor or a manual content brief. The point is, the matrix tells you what to fix, not just that it “needs work”.

Step 3: Add an “overlap” check (so you stop cannibalizing yourself)

Cannibalization is often self inflicted. You publish a new post because “we need more content”, and suddenly you have 4 pages about basically the same thing.

So add one more mini matrix: overlap.

Quick overlap signals:

• Similar titles and H1s
• Same primary keyword in your doc
• Same top queries in GSC
• Users bounce because the page doesn’t answer what they expected

When overlap is high, don’t “update both”. Pick a winner.

Here’s a practical rule set:

• Keep the page that already has backlinks, better rankings, or better conversions.
• Merge the unique bits from the weaker page into the stronger one.
• 301 redirect weaker URL to the stronger URL.
• Update internal links to point to the winner.

You can also make a “canonical cluster” choice if you have a reason to keep both, but most sites do not. Most sites just end up with clutter.

Step 4: Decide what to kill (without fear)

Killing content feels scary because it feels like deleting work.

But dead content has costs:

• It bloats your crawl budget (yes, even on smaller sites it matters over time)
• It weakens topical focus
• It creates internal link dilution
• It makes your site feel outdated

Here’s a simple kill criteria you can use:

Kill candidates often have:

• 0 impressions in 90 days
• no backlinks
• no conversions
• thin content or outdated info
• no strategic reason to exist (no product tie in, no brand value)

Then choose the kill method:

• 301 redirect if there is a clear relevant alternative page.
• 410 if it truly should not exist and there’s no replacement.
• Noindex is usually a temporary band aid. Fine in some cases, but don’t use it to avoid decisions forever.

Also, if the page does get some impressions but it’s for irrelevant queries, that’s not a “hold”. That’s a reposition or a kill.

Step 5: Prioritization (or you’ll end up with a list you never execute)

Once every URL has a decision, you need to prioritize. Otherwise you just built a really nice spreadsheet to feel productive.

I like a basic score:

Priority score = (Business value) + (Traffic potential) + (Ease of improvement)

Each 1 to 5.

• Business value: does this page drive signups, demos, purchases, qualified leads?
• Traffic potential: is the keyword opportunity real? Are competitors beatable?
• Ease: can you improve it in an hour, or does it need a full rewrite and new visuals?

High scoring updates go first. Always.

If you want a more tactical checklist for the “ease of improvement” side, a page level audit can help you see what’s missing quickly. This guide on how to improve page SEO is basically that, a practical list of levers you can pull once the matrix says “update”.

A real example matrix (with decisions)

Let’s say you have these five pages:

That’s the matrix doing its job.

No drama. No “I think we should”. It’s right there.

Where AI automation actually fits (and where it doesn’t)

AI helps most with the repetitive parts:

• pulling content inventories
• summarizing what a page is about
• clustering pages by topic
• identifying overlap and cannibalization patterns
• drafting refreshes, expanding missing subtopics, rewriting intros, updating examples

It does not replace:

• your business judgment (what actually matters to revenue)
• your unique insights and experience
• quality control, fact checking, and tone

If you’re trying to scale this beyond a spreadsheet and a few weekends, that’s where platforms like SEO software are useful. The whole point is hands off content marketing: scan your site, generate a topic plan, create articles, and publish them on schedule. Which is basically the “write” side of the matrix, automated.

And for the “update” side, it’s the same idea. You identify which pages should be refreshed, then run rewrites and improvements without turning it into a six week content ops project.

If you’re in the stage where you’re comparing tools, these breakdowns might help:

• SEO software vs Surfer SEO (more automation vs more manual optimization)
• SEO software vs Jasper (content generation with SEO workflow vs general AI writing)

Not saying you need to switch your whole stack today. Just saying, if you’re doing matrices because you’re drowning in content decisions, automation starts to look very attractive.

The "write" matrix: deciding what to create next

Okay, so far we've mostly talked about existing pages.

But a comparison matrix is also how you decide what to write next, without falling into the trap of random blogging.

Use a simple topic opportunity matrix:

Axes:

• Audience intent: high intent (buying, evaluating) vs low intent (learning)
• Difficulty / effort: easy win vs hard (needs expertise, links, brand authority)

And yes, comparison pages matter a lot right now. They're naturally aligned with what people search when they are close to taking action.

If your site sells an SEO product, you already know what pages tend to convert: "tool vs tool", "best AI SEO software", "alternative to X". Those are high intent. Put them on the calendar.

A quick workflow you can repeat monthly

Here's the process in a clean loop:

1. Export last 90 days data from GSC and analytics.
2. Update your inventory sheet.
3. Run the matrices: performance vs potential, GSC pattern matrix, and overlap check.
4. Assign decisions: write, update, merge, hold, kill.
5. Score priorities.
6. Execute: Update high priority pages first, merge and redirect cannibalized pages, kill true dead weight, and write new pages based on opportunity matrix.
7. Recheck after 30 days.

This is boring, which is kind of the point. SEO wins usually come from boring systems.

Common mistakes (so you don’t sabotage your own matrix)

A few things I see all the time.

Mistake 1: Saving everything

Not everything deserves an update. Some pages are just bad ideas from 2019. Let them go.

Mistake 2: Updating without fixing intent

If a page ranks for “how to” queries but you wrote it like a product page, no amount of keyword sprinkling will fix that.

Mistake 3: Ignoring internal linking

A page can be great but still stuck on page 2 because it has no internal support. Add links from relevant pages. Add it to navigation if it matters. Treat internal linking like a ranking factor because it is.

(And if you want an easy way to check obvious on-page issues before and after, tools like an on-page SEO checker make this less manual.)

Mistake 4: Not tracking what you changed

Add a “change log” column. Even a sentence. “Updated title and intro on Jan 12, added FAQ.” Later you’ll thank yourself.

Wrap up (and what to do next)

Comparison matrices are basically a way to stop running your content strategy on vibes. You build a clean inventory, pull a few key metrics, and let the matrix tell you what to do:

• Write when there’s a clear opportunity and you don’t already have the page.
• Update when the page has signals but needs better intent match, depth, or CTR.
• Merge when you’ve created overlap and your pages are competing.
• Kill when the page has no traction and no strategic purpose.
• Hold when it’s already working.

If you’re doing this at scale, and you want the “write and publish” parts to be more automatic, take a look at SEO software. It’s built for hands-off content marketing, from scanning your site and generating a strategy to drafting and scheduling SEO optimized articles.

Either way, start with the matrix. One spreadsheet. One afternoon. Then repeat it every month. That’s where the compounding happens.

In fact, implementing these strategies can significantly enhance your overall SEO content strategy blueprint, transforming it into a powerful engine that builds compounding traffic over time.