SEO strategy roadmap

Most SEO “roadmaps” you find online are either generic 12-step lists with no sequencing logic, or 80-page PDFs that nobody actually executes. What follows is the version a senior operator runs in their head when planning a 12-month programme: the four phases, what each one delivers, what comes first, what comes second, and what the right success metric is at each gate.

This is the closing article in the Fundamentals cluster. By the time you’ve read what is SEO, how search engines work, how the algorithm works, and the types of SEO, the natural next question is: where do I start, and in what order. Here’s the answer.

“A roadmap isn’t a list of tactics. It’s a sequencing argument — a defence of why this work, in this order, will produce a compounding outcome rather than scattered effort.”

The four-phase roadmap at a glance

The phases overlap. Foundation work doesn’t stop at month 2; content production doesn’t stop at month 6. What changes is the weight— where the team’s hours land — and what the gating success signal is for moving on.

Phase 1 — Foundation (months 1–2)

The single biggest mistake in SEO programmes is starting with content before the foundation is solid. A site with broken indexation, slow Core Web Vitals, missing schema, or a mangled hreflang configuration will produce content that can’t rank no matter how good it is. Phase 1 fixes the floor.

What gets done:

• Technical audit — crawl, indexation, Core Web Vitals, mobile rendering, JavaScript rendering, broken canonicals, broken hreflang, redirect chains, orphan pages, duplicate content. Every issue triaged: critical / important / nice-to-have.
• Site architecture review — URL structure, internal linking topology, depth from homepage, category hierarchy, breadcrumbs. Aim: every important page reachable in three clicks from the homepage.
• Schema implementation — Organisation, WebSite, Article (or Product), BreadcrumbList, FAQPage where applicable. Validates in Google’s structured data testing tool.
• Analytics + Search Console + Bing Webmaster setup — proper conversion goal configuration, segmentation, custom dashboards.
• Baseline keyword research and competitor benchmarking — what queries the buyer types, where competitors rank, where the gaps are.
• Topical map and pillar plan — the strategic argument for which clusters to build, in what order.

Exit signal: the site is fully crawlable, indexed, instrumented, and competitively benchmarked. The team has a written topical map naming the next 6 months of content. Without that exit signal, do not move to Phase 2.

Phase 2 — Content engine (months 3–6)

With the floor solid, the work shifts to building the content asset that will generate compounding traffic for years. The structural rule: build in clusters, not in isolated articles.

What gets done:

• Pillar pages for the 3–5 priority topical clusters identified in Phase 1. Each pillar is the canonical entry point for its topic — comprehensive, well-structured, links out to supporting articles.
• Supporting articles beneath each pillar, targeting specific informational and commercial-investigation queries. Aim: 6–12 supporting articles per pillar in the first 6 months.
• On-page optimisation for every article: title tag, meta description, H1, headings hierarchy, internal links, schema, image alt text, canonical, OG tags.
• Internal linking discipline — every supporting article links up to its pillar; pillars link to relevant siblings; cross-cluster links where the topical relationship is real, not forced.
• Early link earning — original research, expert quotes, data assets, tools/calculators that earn organic links rather than transactional placements.
• Conversion path mapping — every content asset has a defined “next step” toward signup / demo / contact, sized to the buyer’s journey stage.

Exit signal: first non-branded organic traffic flowing in; long-tail keywords ranking; at least 2–3 topical clusters live with full pillar + supporting article structure. Content production is operating on a sustainable cadence (typically 4–10 articles per month depending on team size).

Phase 3 — Authority compounding (months 6–9)

By month 6, the content engine produces consistent output and long-tail rankings are flowing. Phase 3 is where the programme starts moving up the difficulty curve — earning the authority signals that let the site rank for harder, more commercial keywords.

What gets done:

• Digital PR campaigns — pitching original research, data, expert commentary to relevant publications. Goal: high-authority links from publications buyers actually read.
• Expert content — long-form pieces with genuine first-hand expertise, original data, named experts. The content that earns citations from competitors and authoritative sites.
• Guest collaborations — co-authored articles, podcast guesting, expert quotes given to other publications. Builds entity signals and brand mentions on top of links.
• Deeper topical authority — extending each cluster with the harder, more comprehensive articles only an entity with established credibility can publish convincingly.
• Comparison + alternatives content — “X vs Y”, “alternatives to Z” pages targeting commercial-investigation queries. These are competitive but high-value once the site has the authority to rank for them.

Exit signal: domain-level authority signals (referring domains, branded search volume, citation flow) trending up; head-term commercial keywords starting to rank in pages 2–3 of Google; first competitor citations of your content; AI engine mentions starting to appear.

Phase 4 — Optimisation + AI engine work (months 9–12)

By month 9, the programme has produced a meaningful content asset, real authority signals, and traffic. Phase 4 is about extracting maximum value from the foundation that’s now in place.

What gets done:

• Content refresh cycles — pages that drifted out of date, missed ranking opportunities, or have improved competition get rewritten with new data and structure. Selective, not blanket.
• Conversion rate optimisation — high-traffic pages get tested for better CTAs, lead capture, and conversion paths. SEO produces visitors; CRO converts them.
• AI Overview optimisation — restructuring content for citation in Google’s AI Overviews: clear answers in early paragraphs, structured comparisons, schema-rich data, named entities.
• ChatGPT, Gemini, Perplexity, Claude citation work — engine-specific optimisation patterns, brand mention building, sameAs identity reinforcement, technical accessibility for AI crawlers.
• Programme review + Year 2 plan — what worked, what didn’t, where the next year’s investment goes.

Exit signal: mature compounding traffic curve; consistent AI engine citations across at least 2 major surfaces; clear revenue contribution traceable to organic; written Year 2 plan informed by Year 1 data.

What KPIs to track at each phase

The mistake most agency reports make is reporting only on visibility metrics (positions, impressions, clicks) and skipping the business layer. A senior operator weights the business layer most heavily — organic-attributed signups, leads, revenue, CAC, LTV — because that’s what determines whether the programme keeps getting funded.

Common roadmap mistakes

• Starting content production before the technical foundation is solid. Articles get published, don’t rank, team blames the writers. Real cause: site can’t be crawled or rendered correctly.
• Targeting head-term commercial keywords from day one. A new domain has no authority. Head-term commercial keywords require months of topical authority and link signals before they rank. Start with informational + long-tail.
• Treating SEO as content production only. Content without technical hygiene, internal linking, and authority signals is just blog posts. The compounding only happens when all three layers move together.
• Quitting at month 4–5. The compounding curve is back-loaded. Operators who quit at month 4 get the cost without the payoff. The decision to start an SEO programme is implicitly a 12+ month commitment.
• Reporting only on visibility metrics. Rankings without revenue is a vanity programme. Track to the business layer or the programme will be the first thing cut in the next budget review.
• No AI engine work in 2026. A roadmap built in 2022 that doesn’t treat AI Overview / ChatGPT / Perplexity citation as a first-class workstream is structurally behind. Add it at Phase 4 minimum, ideally earlier.

The bottom line

A real SEO roadmap is a sequencing argument: foundation first, content engine second, authority compounding third, optimisation + AI engine work fourth. Each phase has a defined exit signal, primary and secondary KPIs, and a weight of effort that shifts as the programme matures. Programmes that follow this sequence compound; programmes that skip phases or quit early don’t.

The Fundamentals cluster ends here. From here, the natural next move is into how the algorithm worksif you haven’t read it yet, or into the technical, content, on-page, off-page, local, and AI SEO clusters depending on which phase of the roadmap your programme is in right now.