Technical SEO process
A technical SEO methodology built for development teams.
Four steps, one feedback loop. Every engagement starts from evidence, turns it into specification, ships inside your release process — and compounds with every cycle.
sdn-loop.yml
Diagnose
→ ranked issue register
Architect
→ specs & tickets
Implement
→ merged changes + CI checks
Compound
→ release impact report
↺ loop back to 01 with better data
The problem
Why most SEO recommendations never ship.
Most developers don’t care about SEO, and most SEOs don’t understand code. The result is predictable: recommendations that can’t be implemented, and implementations that ignore search.
A typical SEO deliverable is a long document of issues sorted by a tool’s severity score. It doesn’t say which template causes the problem, how much traffic is at stake, or what “done” looks like. Developers can’t estimate it, product managers can’t prioritise it, and it slowly dies in the backlog.
Our methodology closes that gap. We treat technical SEO as an engineering discipline: diagnose from data, specify like a software requirement, implement inside your workflow, and measure every release. It’s the same process behind every technical SEO service we offer.
Operating principles
Four rules behind every engagement.
Evidence over opinion
Every recommendation traces back to data — server logs, crawl output, rendering tests or Search Console. If it can’t be measured, it isn’t prioritised.
Specification, not suggestion
Fixes are written like software requirements: the component affected, the expected behaviour and acceptance criteria a developer can test.
Inside the workflow
We work in your tools — Jira or Linear, GitHub or GitLab, your sprint cadence and your release process. No parallel universe of SEO documents.
Measured in releases
Impact is reported per release, not per month, so you can see exactly which change moved crawl, indexation, rankings and revenue.
The four-step loop
From raw data to compounding growth.
01
Diagnose
We reconcile server logs, crawl data, rendering tests, Search Console and analytics into one model of your site — segmented by template, so problems are traced to the component that causes them.
Every issue is scored by revenue impact and engineering effort. You get a short list of root causes instead of thousands of warnings.
02
Architect
We design the fix as a specification: URL rules, template contracts, internal linking models, rendering strategy and JSON-LD schemas — documented the way your engineers document software.
Each specification becomes a ticket with context, evidence and acceptance criteria, ready to be estimated in your next planning session.
03
Implement
We work inside your sprints: refining tickets, answering questions, reviewing pull requests for SEO-critical changes and validating on staging before anything reaches production.
Where possible we add automated checks to your CI pipeline, so canonicals, robots rules, structured data and performance budgets can’t silently regress.
04
Compound
Every release is measured against its baseline: crawl behaviour, indexation, rankings, Core Web Vitals and organic revenue. Wins are documented; regressions are caught within days, not quarters.
Those results feed the next diagnosis. Each cycle starts from better data than the last — which is how organic growth stops being a campaign and becomes infrastructure.
Dev ↔ marketing alignment
One backlog. Two teams that agree on it.
Engineering and marketing want different things from the same work. We translate between them, so SEO stops being a negotiation and becomes part of how the product ships.
What engineering gets
Scoped tickets with technical specs and code examples
Acceptance criteria and test cases
Effort estimates aligned to your sprint cadence
CI checks that catch SEO regressions before deploy
SDN
SHARED KPIs
What marketing gets
Priorities sized by traffic and revenue impact
Clear timelines tied to real release dates
Release-level reporting on organic performance
A shared language with the development team
Measuring impact
Reported per release, not per month.
Monthly SEO reports blend a hundred variables into one line. We annotate every release and measure it against its own baseline.
That makes wins attributable, regressions visible within days, and the business case for the next round of SEO work obvious to everyone — from the engineers who shipped it to the stakeholders who funded it.
release-impact.log · illustrative
v4.12
Canonicalised filtered category URLs
crawl waste −38%
v4.13
Server-rendered pagination links
indexed +2,140
v4.15
Hero image priority + font preload
LCP p75 3.8s → 2.1s
v4.16
Template change removed breadcrumbs
regression caught in CI
4 releases · 30 days
→ monthly review
A typical engagement
What the first 90 days look like.
WEEK 0
Scoping call
30 minutes on your platform, goals and constraints. You receive a fixed-scope proposal within days.
WEEKS 1–3
Diagnose & architect
Data access, audit, root-cause analysis and a ranked backlog of specified tickets.
WEEKS 4–10
Implement
Sprint-by-sprint delivery with PR reviews, staging QA and CI guardrails.
WEEK 12+
Compound
Release impact review, next priorities, and an ongoing loop if you need one.
Start the loop
Ready to make SEO part of how you ship?
Tell us about your stack, your team and what you’re trying to grow. We’ll show you where the loop starts.