SEO Is Not a Marketing Problem. It's an Engineering Problem.

Every week, a company hires a "content strategist" to fix their SEO. They brief an agency, commission ten blog posts, build a backlink spreadsheet, and wait. Six months later, rankings are flat. The agency blames the algorithm. The client blames the agency. Nobody blames the actual problem: the website.

Here's the opinion that will make some marketers uncomfortable — most SEO failures are engineering failures. The keyword research was fine. The content wasn't bad. But the site took 4 seconds to load, the HTML was a single-page app with no semantic structure, the headings were divs with classes, and Google's crawler bounced before it finished rendering the page.

No amount of content fixes that.

Google doesn't rank content. It ranks pages. And pages are built by engineers, not copywriters.

The illusion of SEO as a content game

The idea that SEO is primarily about content comes from an era when it was true. In 2012, publishing more words on a topic reliably moved rankings. That signal still matters — but it's now one layer in a much deeper stack.

Today, Core Web Vitals are a ranking factor. Server response time affects crawl budget. JavaScript-rendered pages can take 5–10 seconds to be indexed, if they're indexed at all. Structured data (schema markup) directly influences how your result is displayed in the SERP. None of this is a content problem. All of it is a code problem.

When Google says "page experience," they mean it literally. The experience of loading your page, navigating it, and reading it — all measurable, all technical, all fixable only at the build level.

What a 100 Lighthouse score actually signals

We often mention that 100 Lighthouse is our floor, not our goal. People sometimes read that as bragging. It's not. It's a statement about process.

A perfect Lighthouse score means your images are optimised and sized correctly, your server responds fast, your layout doesn't shift as it loads, your JavaScript doesn't block rendering, and your HTML is semantically correct. These are not aesthetic choices. They are table stakes for ranking in 2026.

A site that scores 52 on Performance isn't just slow — it's sending Google a signal that the page isn't worth prioritising. Meanwhile, a competitor with equivalent content but a clean, fast build will consistently outrank it over time.

Who actually owns SEO?

In most organisations, SEO is owned by marketing. That's the wrong home for it.

Marketing should define the keyword strategy and own the content. But the technical conditions that make SEO work — performance budgets, semantic HTML, meta structures, schema markup, crawlability, canonical tags, redirect logic — those belong in engineering. Or, better, in a team that does both.

The reason agencies struggle is structural: they're hired to write, not to build. They can audit the technical issues and write a report, but they can't fix the codebase. That fix requires a developer. Often, it requires the original developer who understands the architecture. By the time the audit lands on a dev's desk, it's ticket 47 in the backlog.

If your SEO strategy can be executed entirely without touching the codebase, it's probably not the strategy that's going to move rankings.

The practical implication

If you're building a new site — or rebuilding one — the SEO conversation needs to happen before a single line of code is written. Not after launch as a retro-fit. Not as an agency engagement six months in. From the start.

That means choosing a rendering strategy that Google can actually parse. It means designing a URL structure that reflects your content hierarchy. It means writing components that output semantic HTML, not just visually correct layouts. It means image pipelines that generate next-gen formats automatically. It means a codebase where adding schema markup is a five-minute task, not a two-week project.

This is why we don't separate design, development, and SEO into different workstreams. When the person writing the component also understands how Google will read it, the output is fundamentally different — and the rankings reflect that.

The uncomfortable summary

Your SEO isn't underperforming because you haven't published enough content. It's underperforming because the foundation it's built on isn't giving Google a reason to trust it.

Fix the foundation first. The content will work much harder once you do.