An SEO marketing plan is a written commitment to a specific set of pages, a specific technical baseline, and a specific measurement cadence over a defined period. It is not a keyword list, and it is not a slide deck of best practices. If your plan does not say what gets published in week six and who publishes it, it is a wish.
The plan below runs 90 days, fits on two pages, and is built for a site owner or a small team rather than an enterprise program office. It assumes honest timelines: meaningful movement on competitive terms takes four to eight months, and anyone promising faster is selling something.
Start with one number tied to revenue
Every plan needs a single primary goal expressed in business terms, not SEO terms. “Increase organic demo requests from 12 to 25 per month by day 180” is a goal. “Improve rankings” is not.
Work backward from it. If your site converts organic visitors to demos at 1.5%, 25 demos means roughly 1,700 monthly organic sessions. If your average commercial page earns 120 sessions a month once mature, that is about 14 pages of the right kind. Now you have a build quantity, and the plan stops being abstract. The arithmetic will be wrong at the edges — every model is — but it forces the conversation about whether the target is reachable at all with the resources you have.
Audit before you plan the content
Publishing onto a broken foundation wastes the content. Spend the first two weeks establishing what you are working with.
- Crawl the site. Find missing and duplicate titles, broken internal links, orphaned pages, accidental noindex directives, and canonicals pointing at redirects or 404s. Insist on evidence — actual titles, URLs, and H1 text — because a count of issues is not a work queue.
- Check Core Web Vitals with field data, not lab scores. Real-user measurements are what Google uses.
- Connect Search Console and GA4 and export 12 months of query and landing-page data. This is your baseline and your ground truth.
- Benchmark two or three real competitors — domain rating, referring domains, organic keyword count, top pages. Note the weakest of them; that is your realistic 12-month target, not the category leader.
Build the keyword plan around clusters
Scattered pages do not compound. Pick three to five topic clusters that map to what you sell, and build each one properly before starting the next.
For each cluster you want one substantial hub page covering the topic broadly, and five to ten supporting pages answering specific sub-questions. Source the sub-questions from the People also ask and related searches blocks on your target queries, then attach volume, difficulty, CPC, and SERP-feature data so you can sequence them. A multi-seed explorer that returns up to 150 ideas per search from a country-specific index turns this from a week of work into an afternoon; SEO Rocket saves the shortlist to a project keyword pool that later feeds both the writer and the rank tracker, which keeps research and measurement pointed at the same terms.
The 90-day roadmap
Days 1–30: foundation
Fix every technical issue that blocks indexing or wastes crawl budget. Rewrite titles and meta descriptions for your top 20 pages by impressions — this is the fastest measurable win available, and it often shows in click-through rate within three weeks. Finalize the cluster plan and publish the first two supporting pages.
Days 31–60: production
Publish six to eight pages. Start with the terms where an existing page already sits in positions 8–20, because improving a page with age and links behind it beats starting from zero. Build internal links from the hub to each supporting page and back. Begin outreach for the three or four links your weakest competitor’s top pages tell you the category expects.
Days 61–90: compound and correct
Publish another six to eight pages, finish the first cluster, and run the first honest review. Pull Search Console data for everything published in days 1–45 and sort by impressions. Pages with impressions but no clicks need snippet work. Pages with neither need to be reassessed for intent match — you probably built the wrong format.
Set expectations you can defend
The single biggest cause of abandoned SEO programs is a timeline nobody agreed to. Put these in writing at the start:
- Weeks 1–4: technical fixes and title rewrites can produce visible click-through changes. Nothing else will move.
- Weeks 6–12: new pages get indexed and begin picking up long-tail impressions. Positions will be unstable.
- Months 4–8: competitive terms start settling. This is where the compounding shows up.
- Month 12: the cluster either owns its topic or it does not, and you have enough data to decide whether to double down or redeploy.
Also state the caveats. Daily rank movement of two or three positions is ordinary noise. Third-party volume, difficulty, and position figures are modeled estimates from periodic crawls — directional, not exact. Links are necessary but not sufficient, and thin content loses even with good links behind it.
Report on four metrics, not forty
Monthly reporting should fit on one screen. Track organic sessions and conversions from GA4, total impressions and clicks from Search Console, average position for your tracked keyword set as a trend line, and pages published against plan. That last one is the leading indicator: when publishing slips, everything else follows two months later.
Keep third-party rank data beside your own, not instead of it. Trackers show you competitors and direction over time, which Search Console cannot; Search Console shows you the truth about your own site, which no tracker can. Reporting both, clearly labeled, is what stops the quarterly argument about whose numbers are right. A shareable read-only dashboard is worth setting up if a client or an executive is reading the results — it removes the monthly PDF from your workload entirely.
Resource it honestly
A plan that assumes 20 pages a month from a team that has produced four in the last year will fail on schedule. Decide who writes, who reviews, and who publishes, and cut the target until it matches. Two well-built pages a week, sustained for six months, beats a burst of fifteen followed by silence — search rewards consistency because it is evidence the site is maintained.
Budget the tooling to match the scale. A small site can run this plan on Search Console, a spreadsheet, and a browser. Once you are tracking a few hundred keywords, comparing several competitors, and publishing weekly, a single workspace saves real time — SEO Rocket runs research, competitor and content-gap analysis, the writer with hard validation gates, rank tracking, and full-site crawls at a flat $50 per month. Whatever you use, the plan is the thing that matters: one revenue goal, a clean technical baseline, three clusters, and a publishing cadence you can actually hold.
