What I Actually Changed After Running Sage SEO’s Own Marketing Pages Through Sage SEO

We built an SEO audit product for the better part of two years before anyone on our team pointed it at our own homepage. That’s not a typo. We ran it against client sites, demo sites, test sites we spun up on a Tuesday afternoon — but never our own marketing pages. When someone finally asked, half-joking, in a Thursday standup, “wait, has anyone actually scored sageseo.ai with Sage SEO?” the room went quiet in a way that told us the answer, and told us we needed to fix it fast.

This is the post-mortem from that exercise. Not a highlight reel, not a demo dressed up as a Sage SEO case study, but an honest account of what our own tool flagged on our own pages, what we argued about, and what we shipped — and didn’t ship — before the next release.

Why We Finally Audited Our Own Marketing Pages With Our Own Tool

We ran the audit because we were tired of answering “do you use this on yourselves?” with an awkward pause. Curiosity mattered too — would the scoring engine we’d spent months tuning actually find something real, or would it just confirm what we already believed about our own copy? That’s a harder question than it sounds, because nobody is more attached to a homepage than the people who wrote it.

We picked three pages that do very different jobs, scored them cold, and treated every flag as a hypothesis to argue about rather than a verdict to accept.

The Pages We Picked and Why Each One Is a Different Kind of Test

We chose pages that stress-test different parts of the funnel, not three variations of the same problem. A homepage lives or dies on brand clarity and broad relevance. A pricing page lives or dies on conversion copy that sales actually wants kept intact. A cornerstone post lives or dies on topical depth, and ours had been edited piecemeal by whoever happened to own the content calendar that quarter.

That variety was the point. If the content audit only caught issues on one type of page, that would tell us something uncomfortable about how narrow its judgment actually was.

What the Homepage Audit Caught Us Off Guard With

The homepage failed on things we’d stopped seeing entirely, not things we’d never considered. We knew, abstractly, that the page had a lot of hands on it over the years. What we didn’t expect was how much that history had diluted the actual keyword focus — the H1 said one thing, the meta description implied another, and three separate H2s were each quietly competing to be the “main point” of the page.

Where the Team Disagreed With the Tool

Not every flag survived the argument. A couple of heading-structure warnings felt like edge cases — our H3 nesting broke a strict hierarchy rule in a spot where, visually and semantically, it made sense to us as humans. We left those alone. But the meta description mismatch was real: it described a feature set we’d quietly deprioritized two product cycles earlier, which is exactly the kind of staleness that’s invisible to people who read the page every day and never actually read it.

• Heading hierarchy that technically broke rules but worked for human readers
• Keyword focus split across competing H2s instead of one clear throughline
• A meta description describing a product we no longer led with
• Thin alt text on the one hero graphic that mattered most

We rewrote the meta description, consolidated the competing H2s into a cleaner hierarchy, and left the heading-nesting warning as a known, accepted exception. The score moved meaningfully after the changes — not because we chased a number, but because the fixes addressed things a cold reader would genuinely have found confusing.

Is the Pricing Page Also an SEO Asset? We Treated It Like One for the First Time

Yes — and we’d been treating ours purely as a conversion tool for years, which meant we’d never looked at it through an SEO lens at all. Pricing pages get obsessive attention from sales and product marketing, and almost none from whoever owns organic search. Ours was no different. The tool flagged thin body content, almost no semantic variation around our core terms, and — the one that actually stung — a near-total absence of internal links back to the blog posts or feature pages that would give a search engine any sense of what sat around it in our site’s topic map.

Sales Copy vs. SEO Recommendations

This is where the internal debate got real. Sales-adjacent copy wants short, punchy, scannable — SEO wants enough contextual depth that a crawler (and a cautious buyer doing research) understands what the page is actually about. We didn’t fully resolve that tension; we compromised. We added two sentences of plain-language context above the fold and linked out to our pricing page’s natural companions — a features explainer and a relevant blog post — without touching the short, punchy plan-comparison table that converts well. We deliberately ignored a recommendation to lengthen the FAQ accordions at the bottom; it would have helped topical depth marginally and hurt scannability a lot, and for a conversion-critical page, that trade wasn’t worth it.

Why the Cornerstone Blog Post Was Quietly Competing With Our Own Homepage

Our oldest cornerstone post and our homepage were both chasing the same core phrase, and neither of us had noticed. This is keyword cannibalization in its most common form: two pages, same site, overlapping intent, and a search engine left to guess which one actually deserves to rank. The post in question — our audit framework piece — had been written early, then patched by three different authors chasing three different content priorities over a few years. Nobody had stepped back and asked whether it still matched the job the page was originally built to do.

Deciding Which Page Should Win

We worked through this in roughly the same order every time cannibalization shows up:

1. Pull both pages’ current rankings and see which one Google is already favoring for the overlapping term.
2. Decide, based on search intent, which page a new visitor actually wants to land on first.
3. Narrow the loser’s targeting to a more specific, adjacent angle instead of competing head-on.
4. Rebuild the internal links so the winning page gets reinforced, not undermined.

The homepage won the broad term, as it should — it’s the brand anchor. We refocused the blog post on the more specific, practitioner-level angle it was always better suited to, rewrote its intro to make that distinction obvious to both readers and crawlers, and added internal links pointing from the post up to the homepage instead of the two pages quietly competing sideways. That same discipline is basically what we describe in our piece on rebuilding topical authority on older sites through internal linking — it turns out we needed our own advice.

What Surprised the Team Most About Auditing Our Own Site

The biggest surprise wasn’t anything the tool found — it was how blind we’d gone to pages we look at every single day. Familiarity is the enemy of a fresh read. You stop seeing the meta description because you wrote it eighteen months ago and haven’t glanced at it since. You stop noticing the pricing page has no internal links because you only ever land on it from a demo call, never from organic search.

We also disagreed, more than once, about how literally to follow the tool’s suggestions. One flag recommended breaking the pricing page’s hero paragraph into a longer, more keyword-rich block. We chose not to. It would have nudged the score up and made the page read worse to an actual human trying to decide whether to buy. That’s the judgment call every team using a scoring tool has to be willing to make — understanding how the score is actually built matters more than chasing the number itself.

What We Changed Before Shipping the Next Release

Across all three pages, the pattern was the same: small, specific fixes beat sweeping rewrites. We didn’t tear anything down. We tightened headings, clarified meta descriptions, added internal links where none existed, and redistributed keyword ownership between two pages that had been quietly stepping on each other.

The exercise also surfaced a scoring edge case worth fixing on our end: the hierarchy rule that flagged our homepage’s H3 nesting was too rigid for pages with legitimately nested sub-sections. We logged that for the product roadmap. Dogfooding isn’t a one-time stunt for us anymore — it’s becoming a quarterly habit, the same way we’d tell a client to revisit their own performance data instead of assuming last year’s optimization still holds.

Our Own Pages Weren’t Special — That’s the Whole Point

Every flaw our tool surfaced on sageseo.ai is a flaw we’ve seen on dozens of client sites: diluted focus, forgotten meta descriptions, a pricing page nobody thought to link, two pages fighting over one keyword. We weren’t exempt because we built the thing. If anything, we were more exposed to it, because nobody reads their own homepage with fresh eyes. If you’re evaluating whether this kind of audit is worth running on your own site, don’t look for a score that validates you — look for the flag that makes you a little uncomfortable. That’s usually the real one.