SEO for SaaS: A 90-Day Plan for Search and AI Visibility

Diagnose indexing and demand first, strengthen the pages closest to revenue, and measure Google clicks, AI answers, human referrals, and qualified pipeline as separate outcomes.

At three months in, low traffic alone does not prove SEO is failing. If your SaaS is getting almost no impressions, check whether Google can find and index the important pages first. Then focus on a small set of buyer questions that connect to a real product decision.

The short answer

A strong SaaS SEO plan starts with diagnosis, builds the smallest useful set of high-intent pages, and measures qualified demand. Search Console shows whether Google can discover and serve those pages. AI monitoring shows whether answer engines retrieve, cite, and recommend them. The first 90 days should create a reliable baseline and a repeatable improvement loop.

Start with diagnosis before publishing more content

Google Search Console separates problems that look identical in an analytics dashboard. The Page Indexing and URL Inspection tools show whether Google can find and read a page. The Performance report shows impressions, clicks, CTR, and average position by query and page. Those signals should decide the next action.

A new SaaS with zero indexed product pages has a technical problem. An indexed product page with zero relevant impressions may have a demand or relevance problem. A page gaining impressions around the right query but no clicks has a listing or intent-match problem. Treating all three as a request for more blog posts wastes the first 90 days.

Observed signal	What it suggests	Next action
Important pages are not indexed	Google cannot serve a page that it has not indexed.	Inspect status codes, robots rules, canonical tags, server-rendered content, crawlable links, and the sitemap.
Indexed pages have almost no impressions	The page may miss real demand, use language buyers do not use, or lack enough relevance for the query.	Verify the target question, study the results Google selects, and improve the page that should own that intent.
Impressions arrive but clicks do not	Google sees the page, while the listing or intent match is weak.	Compare the title, description, visible answer, and page format with the current result set.
Clicks arrive but qualified leads do not	Traffic quality, proof, product fit, or the conversion path needs work.	Inspect the landing page, offer, proof, next step, and the queries that produced the visits.
AI crawlers fetch the page but engines do not cite it	Access exists. Retrieval, extractability, relevance, or independent authority may still be missing.	Compare the page with the cited sources for the same buyer question and close the exact evidence gap.
AI citations arrive but referrals do not	The answer may satisfy the user without a click, or the citation may not create enough reason to visit.	Protect the citation, then improve the page with data, tools, or detail that rewards the click.

Two Foglift datasets show why the evidence layers stay separate

Foglift analyzed 854 Technical Audits across 194 distinct domains. The median SEO score was 90/100, while the median AI Readiness Score was 60/100. Among the 165 sites with an SEO score of at least 70, 37 still had AI Readiness below 50. Strong conventional SEO health therefore did not guarantee clear entity facts, extractable answers, structured page signals, or citation readiness.

A separate 28-day Sensor study covered 16 anonymized sites, 311 observed site-days, 212,012 AI crawler fetches, and 1,011 recognized AI referrals. The daily crawler-fetch rate had a Pearson correlation of 0.04 with the AI-referral rate. Live user-fetch rate had a correlation of -0.09. Crawler access and human visits were different signals in that cohort.

194 domains

Latest eligible Technical Audit per site

90 vs 60

Median SEO and AI Readiness scores

16 sites

Sensor cohort with active crawler data

0.04

Crawler-rate to referral-rate Pearson correlation

The practical implication is simple. Use Technical Audits to diagnose access and page clarity. Use Search Console to measure Google discovery and clicks. Use answer-level monitoring to measure mentions and citations by engine. Use Sensor data to measure crawler behavior and recognized AI referrals. Use signups and pipeline to judge commercial value.

Read the full methods and limitations in Foglift's 194-domain AI Readiness benchmark and the 16-site crawler and referral study.

The 90-day SEO plan for an early-stage SaaS

This plan assumes a small team with a working product, a new or underperforming site, and limited publishing capacity. It favors complete buying pages and measured iteration over a large content calendar.

Weeks 1 to 2

Establish the technical and demand baseline

Outcome: A measured starting point and a short blocker list

Verify Search Console and analytics. Record the exact 28-day window you will use as the baseline.
Inspect the homepage, product, pricing, integration, comparison, and key use-case pages for indexing.
Run a Technical Audit on each core page. Fix status, robots, canonical, rendering, heading, and internal-link issues first.
Record Google impressions, clicks, CTR, and average position by page. Keep branded and non-branded demand separate.
Record AI mentions, citations, referrals, and crawler requests as separate evidence layers. Never treat a fetch as a visit.

Weeks 3 to 4

Choose a narrow buyer-intent wedge

Outcome: One page owner for each priority question

Collect the words used in sales calls, onboarding, support, churn notes, community threads, and Search Console queries.
Choose five to ten questions tied to a problem, use case, integration, comparison, price boundary, or switching decision.
Map each question to one existing page before creating anything new. Strengthen the current owner when the intent already exists.
Classify the current winners. Publisher roundups require an earned mention; strong vendor pages provide a content benchmark.
Write the sentence a buyer, Google result, or AI answer should lift from each page. Include the audience, job, evidence, and limit.

Weeks 5 to 8

Improve the pages closest to revenue

Outcome: A small set of complete, verifiable buying pages

Improve product, pricing, integration, use-case, comparison, security, and documentation pages before expanding the blog.
Put a direct answer near the top. Add current prices, plan boundaries, supported workflows, screenshots, examples, and limitations.
Publish original evidence that only your company can provide, such as benchmark data, implementation lessons, or observed workflow results.
Add descriptive internal links between related product and evidence pages. Keep important facts available in server-rendered HTML.
Distribute the strongest evidence through customers, partners, communities, and publications where the intended buyers already spend time.

Weeks 9 to 12

Measure, improve, and decide what earns another cycle

Outcome: A page-level decision log tied to qualified demand

Compare the same Google page and query metrics with the baseline. Diagnose each page from its actual signal pattern.
Compare the same buyer questions across each monitored AI engine. Report mentions and citations per engine.
Count recognized AI referrals by landing page, then connect Google and AI visits to qualified signups, demos, trials, or pipeline.
Improve pages gaining relevant impressions before creating adjacent pages. Protect pages already earning clicks or AI referrals.
Choose the next 90-day cycle from measured gaps. Archive ideas that have no evidence of buyer demand or product fit.

Build a small page set around buyer decisions

Broad category terms such as “project management” or “AI platform” are weak starting points for a new SaaS. Use the product's specific buyer, problem, workflow, integration, constraint, and alternative. The page should help a buyer make a decision even if that person arrived directly.

Start with the homepage, one clear product page, pricing, core integrations, high-intent use cases, documentation, security, and honest comparisons. Publish an article when the question needs explanation, original research, or a reusable workflow. Every distinct intent gets one owner page, and several keyword variations can map to that same owner.

Weak starting target

“project management software”

The buyer, workflow, constraint, and reason to choose the product are missing.

Focused buyer question

“project planning for a five-person agency with client approvals”

The page can answer a specific job, show the workflow, state plan limits, and demonstrate fit.

Google's people-first guidance asks whether a page contains original information, substantial analysis, clear sourcing, first-hand expertise, and enough detail to help the reader achieve a goal. For a SaaS company, the strongest material often comes from the product itself: implementation data, anonymized workflow patterns, support questions, benchmark results, change logs, API examples, and clearly stated plan boundaries.

Measure every page across five evidence layers

A single blended visibility score hides the next action. Keep each layer intact, and judge the page on the job it is supposed to perform.

Layer	Measure	Question answered
Google demand	Search Console queries, pages, impressions, clicks, CTR, and average position	Can Google find the page, and does the listing earn a visit?
AI answers	Brand mentions, recommendation position, sentiment, competitors, and cited URLs by engine	Does an answer engine select and describe the SaaS accurately?
Human visits	Recognized AI referrals and Google clicks by landing page	Did a person reach the site from either discovery channel?
Technical access	Index status, crawl rules, rendered content, canonicals, internal links, and crawler requests	Can the relevant system reach and interpret the page?
Business outcome	Qualified signups, demos, trials, opportunities, and revenue attributed with a stated method	Did the visit produce a commercially useful next step?

Where Foglift fits in the 90-day plan

Foglift connects the evidence layers without blending them. Unlimited single-page Technical Audits inspect SEO, AI Readiness, performance, security, and accessibility. Search Console integration exposes Google demand. AI Visibility monitoring records mentions, citations, sentiment, competitors, and answer trends. Sensor separates crawler requests from recognized AI referrals.

The permanent free layer includes unlimited single-page Technical Audits and active-use weekly Perplexity monitoring. Launch starts at $49 per month and adds ChatGPT, Claude, Perplexity, Gemini, and Google AI Overview monitoring for up to three brands, along with full-site audits, REST API, CLI, and MCP access. That lets a small SaaS run the diagnosis, answer, and measurement loop before adding a larger tool stack.

Start with the page closest to revenue

Run a free Technical Audit on your homepage, product page, or pricing page. Then use Search Console and answer-level evidence to decide whether the next task is access, relevance, listing quality, extractability, authority, or conversion.

Run a free Technical Audit See the Search Console workflow

Frequently asked questions

How long does SEO take for a new SaaS?

There is no universal timeline. A new SaaS should use the first 90 days to establish indexing, identify relevant impressions, improve core buying pages, and measure qualified outcomes. Almost no impressions calls for diagnosis. Relevant impressions without clicks call for a listing and intent review.

What should a SaaS company do first for SEO?

Start with Google Search Console and the pages closest to a buying decision. Confirm that the homepage, product, pricing, integration, comparison, and key use-case pages are indexed and readable. Then map a small set of buyer questions to one page each.

Should a SaaS startup publish blog posts every week?

Only when each article answers a real buyer question with original expertise or evidence. Core product and decision pages usually deserve attention first. Google explicitly recommends people-first, original content and warns against high-volume publishing built mainly to attract search visits.

How should SaaS teams measure SEO and AI visibility together?

Keep the layers separate. Search Console measures Google impressions and clicks. AI monitoring records mentions and provider-returned citations by engine. Sensor data records recognized AI referrals and crawler requests. Qualified signups and pipeline measure business impact.

Do AI crawler requests mean a SaaS will be cited?

No. A crawler request proves that an agent reached a page. It does not prove that an answer cited the page or that a person clicked through. In Foglift's 16-site study, total crawler-fetch rate had a Pearson correlation of 0.04 with recognized AI-referral rate.

What does Foglift provide for a small SaaS team?

Foglift provides unlimited single-page Technical Audits on every plan. Free accounts receive active-use weekly Perplexity monitoring. Launch starts at $49 per month and adds monitoring across ChatGPT, Claude, Perplexity, Gemini, and Google AI Overview for up to three brands, plus API, CLI, and MCP access.

Sources and methodology

Google Search Central, Get started with Search Console. Primary guidance for indexing, URL Inspection, sitemaps, and query and page performance measurement.
Google Search Central, Creating helpful, reliable, people-first content. Primary guidance for original information, substantial analysis, expertise, sourcing, and site focus.
Google Search Central, AI features and your website. Primary guidance for generative Search eligibility and the continued role of SEO fundamentals.
OpenAI, Overview of OpenAI crawlers. Primary documentation for purpose-separated crawler controls.
Foglift Research, AI Readiness Across 194 Websites. The study uses 854 Technical Audits, keeps the latest eligible result per distinct domain, and publishes methodology and limitations.
Foglift Research, AI Crawler Fetches and AI Referrals Measure Different Things. The study covers 16 anonymized sites and 311 observed site-days from September 13 through October 10, 2026.

Fundamentals: Learn about GEO (Generative Engine Optimization) and AEO (Answer Engine Optimization) (the two frameworks for optimizing your content for AI search engines).