GrowthTechnical SEO

Most technical SEO problems are engineering problems. Pages a search engine can't render. Templates that generate thousands of near-duplicates. Redirects that chain. Navigation that hides the pages that matter. Speed lost to scripts nobody needs. An audit can describe them; we can also fix them, in your codebase, and measure whether the fix worked.

Search strategy sits on top of that foundation: understanding what people search for, mapping it to page types and a site structure, and designing content models so good pages can be published consistently and at scale.

What we do

What we work on

  1. 01

    Technical audits that end in changes

    Crawl, render and log analysis, prioritised by impact and delivered as code, not a spreadsheet.

  2. 02

    Migrations and replatforms

    URL mapping, redirect testing, and monitoring through and after launch.

    Web development
  3. 03

    Search strategy and site architecture

    Demand research mapped to page types, URL structure and internal linking.

  4. 04

    Content architecture

    Content models and templates for publishing useful pages at scale, without thin or duplicate output.

  5. 05

    Ecommerce SEO

    Collections, faceted navigation, product data and structured data for large catalogues.

    Ecommerce
  6. 06

    Performance

    Core Web Vitals improved in the code: rendering, images, fonts, scripts and caching.

    Web development

What we actually look at

Not a hundred-line audit template. These are the areas where we find the problems that matter, and where the fix is usually a change to code or configuration.

Crawling

  • Can every important page be reached by links?
  • Is crawl spent on pages that should exist?
  • Do sitemaps list only canonical, indexable URLs?

Rendering

  • Is the content in the HTML, or only after JavaScript runs?
  • Are links real <a href> elements?
  • Does lazy-loading hide content from crawlers?

Indexing

  • One canonical URL per piece of content
  • Parameters and facets controlled
  • Noindex where it belongs, and nowhere else

Architecture

  • Important pages within a few clicks of home
  • Topics grouped into a structure people and crawlers can follow
  • Breadcrumbs and internal links that match it

Migrations

  • Every old URL mapped to its closest new one
  • Single-hop 301s, tested before launch
  • Rankings and errors watched for weeks after

Performance

  • LCP, INP and CLS from real-user data
  • Render-blocking resources removed
  • Images and fonts sized and prioritised
/services/shopify-development  /services/ecommerce/             301
/services/ai-platforms         /services/ai-automation/         301
/services/platform-support     /services/technical-partnership/ 301
/expertise/                    /services/                       301
/card                          /                                302
Redirects from this site's own migration: each old address goes to its nearest new page in one hop.

Capabilities

Underneath

  • Search Console
  • Crawlers and log files
  • Lighthouse and CrUX
  • Schema validation
  • CI checks

Discovery

  • Robots, sitemaps and crawl paths
  • Index coverage and canonical consolidation
  • Server log analysis

Rendering

  • Server and static rendering for content
  • JavaScript SEO: real links, hydration, lazy loading
  • Internationalisation and hreflang

Structure

  • Information architecture and URL design
  • Internal linking and breadcrumbs
  • Structured data that describes real content

Measurement

  • Search Console and crawl monitoring
  • Release checks that catch SEO regressions in CI
  • Organic performance tied to conversions, not just rankings

Approach

How it runs

  1. 01

    Measure

    Crawl, render and log data alongside search and analytics, so the baseline is real.

  2. 02

    Prioritise

    Rank issues by likely impact and effort. A few changes usually matter far more than the rest.

  3. 03

    Implement

    Changes made in the codebase, reviewed and tested like any other change.

  4. 04

    Guard

    Automated checks in the release process so the same problems can't quietly return.

  5. 05

    Measure again

    Compare against the baseline and decide what's next.

Why Northcode

Strategy and implementation are the same team. A finding becomes a pull request rather than a ticket waiting in someone else's backlog, and we can tell you whether a recommendation is a one-line change or a month of work before you commit to it.

What we avoid

  • Location pages and doorway content written for crawlers, not people
  • Structured data that claims things the page doesn't show
  • Reports that rank issues by count rather than by what they cost you

Contact

Tell us where your software is today.

Whether it’s already live, half-built, fragile, or still an idea, we’ll give you an honest view of what it needs and what can wait.