Search performance, fixed at the source.
Technical SEO and search architecture from engineers who can change the code: crawling, rendering, information architecture, internal linking, structured data, performance and migrations.
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
- 01
Technical audits that end in changes
Crawl, render and log analysis, prioritised by impact and delivered as code, not a spreadsheet.
- 02
Migrations and replatforms
URL mapping, redirect testing, and monitoring through and after launch.
Web development - 03
Search strategy and site architecture
Demand research mapped to page types, URL structure and internal linking.
- 04
Content architecture
Content models and templates for publishing useful pages at scale, without thin or duplicate output.
- 05
Ecommerce SEO
Collections, faceted navigation, product data and structured data for large catalogues.
Ecommerce - 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 / 302Capabilities
Underneath
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
- 01
Measure
Crawl, render and log data alongside search and analytics, so the baseline is real.
- 02
Prioritise
Rank issues by likely impact and effort. A few changes usually matter far more than the rest.
- 03
Implement
Changes made in the codebase, reviewed and tested like any other change.
- 04
Guard
Automated checks in the release process so the same problems can't quietly return.
- 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.