90 day SaaS SEO plan

Most first quarters of SaaS SEO produce a deck. Somebody audits for six weeks, presents 140 findings in a spreadsheet with a red-amber-green column, and the first page goes live in month four. By then the person who approved the budget has stopped asking, and the programme spends the rest of the year defending itself.

The plan below is built the other way round. Audit for two weeks, publish in week three, and accept that some of what you ship will need fixing later. Pages that exist can be improved. Pages in a backlog cannot.

Every block has named owners and a deliverable you can point at. Dates assume a Monday start and a team of one marketer, part of one engineer, and a subject matter expert who can give you two hours a week.

Weeks 1 and 2: audit, map, baseline

Two weeks, three artefacts, nothing else. The temptation to keep looking is the enemy here, and the audit is finished when the calendar says so rather than when the list stops growing.

Build the baseline prompt set the same week. Pick 20 buying questions a prospect would ask an AI assistant, run them across ChatGPT, Perplexity and Google AI Mode, and record which domains get cited. Re-run the identical set monthly. It takes 40 minutes and it is the only version of AI visibility measurement that survives a model update, because the prompts stay fixed even when the answers change.

Keep the map in one place rather than in three tools. Our SaaS keyword map template has the columns for intent, estimated demo rate and owner, which is what makes the prioritisation argument in week two a five-minute conversation instead of a workshop.

Weeks 3 to 6: ship five bottom of funnel pages while the audit is still open

Publishing starts on day 15 whether or not the technical list is closed. This is the single highest-value decision in the plan and the one most teams get wrong.

Five pages, all bottom of funnel, chosen from four patterns:

Write comparison and alternatives pages honestly or do not write them. A page claiming you beat the leader on every row is read as marketing and converts like marketing. Name two things the competitor does better and the whole page becomes credible. The mechanics are in SaaS alternatives pages and the drafting sequence in the ship five bottom of funnel pages module.

Integration pages are the fastest win available in week three, because the query is specific, the competition is usually a forum thread, and you already have the technical detail in your docs. If you have more than 15 integrations, treat them as a template family with a real brief rather than as 15 individual projects, using the programmatic page brief template and the structural guidance in SaaS integration pages.

Run the technical work in the same four weeks. Close the six to twelve blocking issues, then take the highest-impact items from bucket two: internal linking, title rewrites on pages already sitting in positions four to fifteen, and any orphaned page with links pointing at it. One engineering sprint covers most of it.

Finish week six with an internal linking pass. Every new page should receive at least three contextual links from existing content, and should link out to the product pages it supports. This is 90 minutes of work that changes indexation speed noticeably on younger domains.

Weeks 7 to 10: one asset worth linking to, plus the comparison set

By now you have pages that convert and no links. Weeks 7 to 10 fix the second problem while widening the first.

Build one asset that a person outside your company would link to without being asked. In practice that means a free tool, a calculator, or a dataset from your own product usage. A calculator built in a week outperforms six months of outreach for most SaaS companies, because it gives journalists and bloggers a reason to cite you that has nothing to do with your marketing.

Pick the asset by asking one question: what number do our customers currently work out in a spreadsheet? Payback period, seat cost, storage overage, compliance deadline, headcount ratio. Build the tool that replaces that spreadsheet, put the method on the page so it can be checked, and skip the email gate for the first quarter. A gated tool collects addresses and earns no links, which defeats the purpose of shipping it in week seven. If you have product usage data nobody else holds, a small dataset with a stated sample size is the stronger option, and it takes an analyst about three days.

Then expand the page set. Four to six more bottom and middle of funnel pages, chosen from the same keyword map, plus the second tranche of comparison pages if the first ones are showing impressions. Add schema in the same sprint: Organization, Product, FAQPage where genuinely applicable, and BreadcrumbList. Schema will not rank you, but it makes your entity easier for answer engines to resolve, which matters when 82 percent of B2B technology queries now surface an AI Overview.

Start the citation work in week eight rather than week twelve. Pitch the asset to three or four industry newsletters, get listed in the obvious directories and roundups for your category, and answer the questions in your space where practitioners actually ask them. The approach is covered in earn links and AI citations, and the honest version is that you should expect 5 to 15 referring domains in this window, not 50.

Weeks 11 to 13: refresh, report, plan the next quarter

The last block has almost no new publishing in it. Three jobs: improve what exists, build the reporting so it runs without you, and write the quarter two plan with evidence instead of assumptions.

Run a refresh pass over the existing library. Pull every page with impressions but a position between five and twenty, and fix the title, the opening 60 words and the internal links pointing at it. On an established domain this produces more traffic in six weeks than any new page you could have written instead.

Build the reporting view your executive team can open themselves. Non-brand organic sessions, signups or demos from organic split by page, pipeline created, indexed page count, referring domains, and the AI citation rate from your fixed prompt set. Weekly refresh, one link, no request required. Reporting that requires a human to assemble it stops happening in month five.

Then write the quarter two plan against the actual curve rather than the forecast. Feed the real month one to three numbers into the organic traffic forecast calculator and produce a revised projection. If your page-level conversion rate came in at 1.2 percent rather than the 3 percent you assumed, everything downstream changes and it is better to say so in week thirteen than in month nine.

Who owns what, and the hours it really takes

A 90 day plan with no named owners is a wish list. Here is the realistic allocation for a small team.

Total internal load lands between 12 and 18 hours a week outside the lead’s own time. Teams that plan for less than that discover in week five that nobody can review a comparison page, and the schedule slips a fortnight.

Three things are explicitly out of scope for this quarter and should be written down as such: a site redesign, a domain migration, and any international rollout. Each one will consume the entire quarter and push every measurable outcome into month seven. International SaaS SEO is a quarter three project at the earliest, and it is a good one, just not now.

What day 90 actually looks like, by domain strength

Expectations set badly in week one are the reason good programmes get cancelled. Here is what a competent quarter produces, by starting position.

Starting point	Rankings at day 90	Traffic at day 90	Pipeline at day 90
New domain, under DR 20, under 12 months old	Indexed, first impressions, a few long-tail positions past page 3	Under 500 non-brand sessions a month	Effectively none. Do not promise any
DR 20 to 40	Page 2 to 3 on target bottom of funnel terms, page 1 on long tail	500 to 2,000 non-brand sessions	2 to 8 signups or demos a month
DR 40 to 60	Page 1 on several low-competition bottom of funnel terms	2,000 to 8,000 non-brand sessions	10 to 30 signups or demos a month
DR 60 plus, established library	Top 5 on target terms, refresh pass producing most of the gain	8,000 plus non-brand sessions	Attributable pipeline, comparable to a paid channel

Say the first row out loud to your executive team in week one. A new domain will not produce pipeline in 90 days and anyone promising otherwise is either selling something or has never run the first quarter on a young site. What you can promise is indexation, impressions on the right queries, a working measurement system and pages that will produce in quarter two. If you are at that starting point, the sequencing in seed stage SaaS SEO is a better fit than this plan’s default assumptions.

The failure that kills most first quarters

It is not bad keyword selection or weak writing. It is the six-week audit, followed by a strategy document, followed by a stakeholder alignment round, followed by the first page going live in month four.

By month four you have burned a third of the runway you were given, produced nothing indexable, and spent your political capital on a document. The technical issues you catalogued so carefully were mostly cosmetic, and the two that mattered could have been found in three days.

Ship in week three. Some of those five pages will be wrong and you will rewrite two of them in month five. That is a much cheaper mistake than an empty site with a beautiful spreadsheet.

Day 91

Quarter two is where the compounding starts, and it looks nothing like quarter one. Publishing volume goes up, the refresh cadence becomes monthly, and link and citation work moves from an afterthought to a third of the effort.

Before you plan it, do one thing: re-read the baseline document from week one next to today’s numbers. That comparison is the argument for the next budget, and it only exists because somebody spent an hour in week one writing down where you started.

The full programme this quarter fits into, including the parts deliberately deferred, sits in the SaaS SEO pillar.