A lot of teams start with the asset they know how to make. If the roadmap likes landing pages, everything becomes a landing page. If the content team is strong, every problem gets translated into a blog post. Search does not care about your org chart. It cares about what the person was trying to get done.
That sounds obvious, but it explains why some pages quietly win while prettier pages stall. The page shape matches the job. The story comes second.
A category term is often a lesson before it is a signup
Ahrefs' HubSpot case study makes the point cleanly. HubSpot struggled to win a broad category term with a product page, then did better with an informational companion page for a category head term that explained what CRM is and who it is for.
I like this pattern because it respects the hesitation inside the query. A lot of category searches are not buying searches yet. The reader still wants orientation. If your page tries to close before it explains, it often loses both the ranking and the trust.
Sometimes the page should just do the thing
The same HubSpot analysis points to the opposite lesson on tool queries. Its email signature generator worked because it behaved like an exact-intent free tool one-pager. The searcher wanted a tool, not an essay about signatures.
This is where a lot of product marketing gets in its own way. Teams add framing, brand talk, feature tours, and proof blocks before the useful action. On a tool query, that extra material is often just delay. The cleanest page is usually the one that gets out of the user's way fastest.
Reference pages get stronger when they stay in one place
There is another page shape that gets underrated because it looks boring: the annual state report on a stable URL for backlink compounding. Ahrefs showed how HubSpot kept its flagship State of Marketing report on the same canonical page so each new edition inherited the trust and links of the last one.
This is less glamorous than launching a fresh microsite every year, but it is much smarter. Reference pages win by becoming the place people cite without thinking twice. You only get that if the URL itself becomes familiar.
Utility often travels better than translated thought leadership
Medicai offers a useful correction for teams doing international SEO. Instead of treating expansion as a translation exercise, it leaned on localized free utility pages for global SEO such as imaging tools, then used localization and hreflang fixes to support those assets.
That makes sense to me. A useful tool survives language changes better than a generic article because the value is already concrete. If a market responds to the utility page, you have earned the right to translate the broader story later.
Messy head terms usually need a hub, not a monologue
Carwow did this well with a review hub architecture for mixed-intent head terms. Instead of one oversized review, it used a parent page plus subpages for the interior, specs, colors, and other buyer questions around the same model.
That structure works because broad queries usually hide several narrower questions. A hub lets the reader move toward the exact answer they wanted, while still giving the brand one coherent surface to own. It is better for rankings, but it is also just better manners.
Where this applies
For SaaS, this often means splitting category education from the core product page, and treating free tools as first-class acquisition surfaces. For AI products, it means proving one useful action on the page before explaining the system behind it. For marketplaces, it means building review or buyer hubs around the confusing head terms people use before they trust a platform. For creator tools, it usually means making the output page or utility itself the thing that earns distribution.
When a page underperforms, I would ask a simple question before rewriting the copy: what job did the searcher hire this page to do? A surprising amount of growth work gets clearer once you answer that honestly.