⚡ What Does SEO for a New Website Actually Require in 2026?
The 90-day playbook breaks into three phases: get your technical foundations right (days 1–30), publish targeted content (days 31–60), then build authority and AI visibility (days 61–90). The mistake I see over and over on new launches is jumping straight to content before the site can be properly crawled or indexed — dozens of articles going up on a site Google can barely read. Sort the foundation first, then publish. Most new sites start showing up in Google Search Console within 30–60 days of doing this correctly. Real organic traffic takes longer — months 3 to 6 is realistic. In Ahrefs' study of 2 million new pages, only 5.7% reached the top 10 within a year, so the keywords you pick in these first 90 days matter more than almost anything else.
I've personally run SEO for new site launches across B2B SaaS, e-commerce, professional services, and publishing — including taking a SaaS from nothing to 8,400 monthly organic visitors in 11 months, getting a professional services firm onto page one for 14 service-area queries within 4 months of launch, and building first-ranking traction for multiple e-commerce stores from scratch. This 90-day sequence is what I've settled on after trying a lot of different approaches across different verticals and site types.
1. The Realistic SEO Timeline for New Websites — What the Data Says
Before you build your 90-day plan, get clear on what SEO can actually deliver in three months — and what it can't. More SEO programmes get abandoned because of bad expectations than bad execution. Teams do the right things, don't see results by week eight, and call it. The work wasn't the problem.
| Research Source | Key Finding | Implication for New Sites |
|---|---|---|
| Ahrefs — How Long Does SEO Take? | In a 2-million-page study, only 5.7% of new pages reached the top 10 within 12 months. The ones that did typically took 2 to 6 months, and keyword difficulty drove most of the variance. | Use the first 90 days primarily to target low-competition keywords, while selectively testing moderate-difficulty terms where the SERP looks genuinely winnable |
| Backlinko — Google CTR Statistics 2025 | Position 1 pulls 27.6% of clicks; position 2 gets 15.8%; position 10 gets 2.4%. A steep dropoff that makes position matter more than volume. Exact figures vary by study, and AI Overviews pull informational clicks down further. | Target keywords where you can hit the top two positions — even at lower volume — rather than chasing competitive terms and landing at position 8 |
📈 What to expect by day 90
A well-run new site SEO programme should have: (1) all important, indexable pages submitted through the sitemap and checked for indexing issues, with most priority pages indexed (Google doesn't guarantee indexing), (2) impressions growing week-on-week on target keywords, (3) at least 3 to 5 pages ranking in positions 1–20 for low-competition long-tail queries, and (4) a content cluster of 10 to 15 pages with internal linking in place. Full traffic momentum typically builds between months 4 and 6.
The timeline question is the hardest conversation on every new site launch I work on. The pressure to show results quickly is almost always present, and the honest answer — that meaningful organic visibility typically takes three to six months at minimum — is not what anyone wants to hear in week one.
The reframe that works is shifting from rankings to indexation signals. Week one success is Googlebot crawling and indexing your pages. Week four success is appearing in GSC for long-tail impressions. Week eight success is seeing click-through on those queries, even at low volume. These are real, measurable milestones that tell you the foundation is working — even before a single page one ranking appears. Founders who understand what the leading indicators look like are much less likely to abandon a sound strategy at the six-week mark because they can't yet see the end result.
2. Why Sequencing Beats Speed
New site SEO fails in a predictable pattern when you skip steps. These are the four failure modes I document most often in audits:
Publishing pages onto a site with robots.txt issues, JavaScript rendering problems, or noindex tags left over from staging means none of those pages can get indexed. The content investment is wasted until the technical gate is cleared — and you can't get that time back. A crawl audit before publishing anything takes 2 to 3 hours and prevents weeks of wasted effort (the technical SEO guide covers what to check).
Chasing backlinks when you have fewer than 10 to 15 solid indexed pages gives link prospects no reason to link — and gives Google too little topical signal to understand what your site is actually about. Links to a thin site carry less ranking value than the same links pointing to a topically established one. Build the content cluster first, then run the link campaign.
A new domain with little or no authority is unlikely to rank for highly competitive terms like "best CRM software" or "SEO agency" within 90 days, however good the content is. Those terms usually need a year or more of authority building. Start with ultra-low-competition long-tail queries, pick up early ranking wins, generate real traffic signals, and build the domain authority that competitive terms will eventually need.
Google's own documentation says AI Overviews and AI Mode have no requirements beyond normal Search eligibility. In Ahrefs' data, 76% of pages cited in AI Overviews also rank in Google's top 10. A site that skips crawlability and keyword fit in favour of AI formatting tricks loses in both places.
3. The Three-Phase Overview
The 90-day playbook runs in three phases. Each phase has clear entry conditions — things that need to be true before you move forward — and exit milestones that confirm the phase is done.
Phase 1 · Days 1–30
Technical Foundations
- HTTPS + crawlability confirmed
- Google Search Console + GA4 set up
- XML sitemap submitted
- Core Web Vitals passing on mobile
- Schema markup implemented
- Keyword research completed
- Site architecture planned
Phase 2 · Days 31–60
Content & On-Page Launch
- Pillar pages + cluster articles live
- On-page SEO optimised (titles, H1s, meta)
- Internal linking structure built
- FAQ sections where readers need them
- Author pages with credentials
- 8–12 articles targeting low-KD keywords
- GSC indexing monitored weekly
Phase 3 · Days 61–90
Authority & AI Visibility
- Link-earning campaign started
- Brand mentions + PR outreach begun
- Google Business Profile (if local)
- AI citation manual sampling started
- First content refresh based on GSC data
- 2–3 deeper pillar articles published
- 90-day performance review completed
4. What to Skip in Your First 90 Days
Most founders have more ideas than hours, so the cut list matters as much as the to-do list. Before the plan starts, take these off the table:
- Thirty generic blog posts. Ten real answers to buyer questions will do more.
- Head terms like "project management software". They need links you don't have yet.
- Paid links and link packages. Google's spam policies name them directly.
- Chasing DA or DR. Those are tool-vendor scores, not Google metrics.
- One page per keyword variation. Cover the topic once, well, before splitting it.
- Five overlapping tools. Search Console, GA4 and one paid suite are plenty.
- Waiting for perfect copy. Publish, read Search Console, then improve.
- Reporting traffic alone. Put leads and signups beside it from week 1.
- AI-written pages made to raise the post count. Volume for its own sake is the pattern Google's scaled-content rules target.
5. Phase 1 (Days 1–30): Technical Foundations
Technical SEO is the gatekeeper for everything else. A slow mobile site, crawling problems, or missing structured data will quietly undermine every piece of content and every link you build on top of it. This phase isn't glamorous, but there's no higher-leverage work in the entire 90-day programme.
Week 1 — Day 1–7: Crawlability, Indexing, and Analytics Setup
- Install HTTPS (SSL certificate) — Google treats it as a ranking signal, and Chrome flags non-secure sites with a warning that kills trust immediately
- Verify and submit your XML sitemap to Google Search Console, then use URL Inspection to request indexing for a handful of priority pages (homepage, core product or service pages). Google notes that repeat requests don't speed crawling up, so don't hammer the button
- Audit
robots.txt— confirm it's not blocking Googlebot from your content, CSS, or JavaScript files CRITICAL - Check every page for accidental noindex tags carried over from staging — this is the most common launch-day technical error I find
- Set up Google Search Console with your verified property and connect it to Google Analytics 4
- Configure GA4 conversion events: form submissions, phone clicks, email clicks, and any transactional actions (see the GA4 guide)
- Add proper consent management to GA4 (required for GDPR compliance in EU markets)
- Confirm Googlebot can actually render your pages — run URL Inspection on 3 to 5 representative pages and check what Google sees
- Pick your preferred domain (www vs non-www) and set up a 301 redirect from the non-preferred version
Week 2 — Day 8–14: Core Web Vitals, Mobile, and Site Speed
- Run PageSpeed Insights on your homepage and 3 key landing pages — record your LCP, INP, and CLS scores on mobile CRITICAL
- Target LCP under 2.5s, INP under 200ms, and CLS under 0.1 on mobile — those are Google's "Good" thresholds
- Compress and convert all images to WebP — unoptimised images are the most common cause of slow LCP on new sites
- Enable lazy loading for below-the-fold images — add
loading="lazy"to any img tag that isn't in the initial viewport - Test your mobile layout on at least three device sizes in Chrome DevTools
- Check tap targets — buttons and links need to be at least 48×48px on mobile
- Enable browser caching and GZIP or Brotli compression at the server level
- On WordPress: install WP Rocket or W3 Total Cache and connect a CDN
- Aim for a fast mobile load above the fold. Slow pages lose visitors before they read a word, and the Core Web Vitals guide walks through the fixes in depth
Week 3 — Day 15–21: Schema Markup, Site Architecture, and On-Page Foundations
- Add Organization (or LocalBusiness) structured data to your homepage so your business name, logo and official profiles are stated clearly. It helps Google connect the entity, but it doesn't guarantee a Knowledge Panel
- Add BreadcrumbList schema to all pages with visible breadcrumb navigation
- Lock in your URL structure before publishing anything — short, keyword-containing slugs (e.g.
/keyword-phrase, not/p=12345) - Keep site architecture flat — no important page should be more than 3 clicks from the homepage
- Build your About page with named team members, their credentials, and links to verifiable external profiles — this is your E-E-A-T foundation
- Publish a privacy policy and terms of service — Google's quality raters look for these as basic trust signals on new sites
- Set a consistent internal linking rule: every new article links to at least 2 existing pages and has at least 1 existing page linking to it (the internal linking guide has the full method)
- Create a 404 page with navigation links — stops crawl dead-ends and keeps users from bouncing when they hit a broken URL
Week 4 — Day 22–30: Keyword Research and Content Architecture Planning
- Pick 1 to 3 core topic clusters — the specific subject areas you'll build authority on in the first 90 days. Sites that try to cover 8 to 10 topics at once build shallow authority across all of them; focusing on 2 to 3 builds real depth STRATEGIC
- Use Ahrefs or Semrush to pull 30 to 50 candidate keywords within your clusters. A KD filter (roughly under 20 in Ahrefs) is only a first screen; the manual SERP check in Section 8 decides which ones stay. Favour long-tail queries of 3 to 5 words
- Manually check the SERP for each keyword. What format dominates the top 3 results? Yours needs to match what's already ranking (the search intent guide goes deeper)
- Assign each keyword to one specific page URL and one only — targeting the same keyword on two pages creates cannibalization CRITICAL
- Pick 2 to 3 competitor sites ranking for your target keywords — look at their content structure, word count, internal linking, and schema
- Build a 90-day content calendar: publishing dates, target keywords, assigned author, and word count targets for each article
- Map your pillar page structure: 1 comprehensive pillar per cluster (1,800 to 2,500 words) supported by 4 to 6 cluster articles targeting sub-queries
- Decide your page order before you write: homepage, core product or service pages and use-case pages come ahead of most blog posts (Section 10)
6. Phase 2 (Days 31–60): Keyword-Targeted Content Launch
Phase 2 is where Phase 1 pays off. You now have a technically sound, indexable site with a mapped keyword architecture. The goal here is simple: publish enough high-quality, well-optimised content to start building topical authority in your chosen clusters and generate first ranking signals in Google Search Console.
Week 5–6 — Day 31–45: Core Pages and First Cluster Articles
- Publish your homepage, primary service/product pages, and About page with full on-page optimisation — title tags under 60 characters, meta descriptions under 155, H1 containing the primary keyword. These money pages go live before the blog (Section 10)
- Publish your first pillar article for cluster #1 — target a moderately competitive head term (KD roughly 15 to 25, if the SERP check passes), 1,800 to 2,500 words, with an FAQ section built from real People Also Ask questions MILESTONE
- Publish 3 to 4 cluster articles supporting the pillar — each targeting a specific long-tail sub-query (KD under 15, 800 to 1,200 words), each linking back to the pillar and to one another
- Write image alt text that accurately describes what the image shows — descriptive and keyword-contextual, not stuffed
- Add a short FAQ where readers genuinely have follow-up questions, and write it for them. Google ended FAQ rich results on 7 May 2026 and says AI features need no special markup, so FAQ schema is optional. Keeping it does no harm
- Internal link every new article: linked to from at least one existing page, and linking out to at least two relevant internal pages
- Link every new URL from a relevant page and keep it in the sitemap. Save URL Inspection requests for the pages that matter most
- Set up rank tracking in Semrush or Ahrefs for your 30 to 50 target keywords — record baseline positions on day 31 before any content is live
Week 7–8 — Day 46–60: Second Cluster, Content Depth, and GSC Monitoring
- Publish the pillar article for cluster #2, using the same structure as cluster #1, with cross-links between related articles across both clusters
- Publish 3 to 4 cluster articles supporting pillar #2
- Check Google Search Console — by day 45 to 50, your earliest pages should be generating impressions. If they're not, re-inspect those URLs and dig into indexing issues MILESTONE
- Look for queries generating impressions but no clicks in GSC — these are pages Google has indexed and is testing. Rewrite their title tags and meta descriptions to improve CTR
- Add E-E-A-T signals to every published article: named author with a linked credential page, at least one first-person observation or specific case outcome, at least one cited statistic with a linked primary source
- Create a dedicated author bio page for every named contributor — link to it from every article they've written, marked up with author schema
- Revisit internal linking across your growing content library — link new articles from relevant older ones, and older articles to relevant new ones
- Mid-phase content audit: any article published in weeks 5 to 6 with zero impressions by week 8 needs a keyword difficulty and intent check
The GSC impressions check at day 45 to 50 is something I run on every new site launch. It tells me three things at once: whether Googlebot is actively crawling the site, whether the content is being understood correctly, and whether the keyword difficulty assessments from Phase 1 were actually right.
On a software client launched in mid-2024, we had healthy impression growth on 8 of the 11 articles published in Phase 2 — but three of them had zero impressions. Manual investigation found the problem: all three were targeting queries where Google was serving video results in positions 1 and 2, with no text articles in the top 5. We'd published long-form text content for queries where Google had decided video was the right format. We reformatted two of the three as structured list posts with embedded video, and both started generating impressions within 10 days of republishing. One diagnostic check — not just keyword difficulty but what format dominates the SERP — stopped three articles from being dead weight for the rest of the programme.
7. Phase 3 (Days 61–90): Authority, Links, and AI Visibility
By day 60, you should have 10 to 15 indexed, optimised articles across 2 topic clusters with growing GSC impressions on target keywords. Phase 3 is about building the authority layer — the external validation signals (backlinks, brand mentions, AI citations) that turn early impressions into durable top-10 positions.
Week 9–10 — Day 61–75: First Link Building Campaign and Brand Authority
- Find 20 to 30 link prospects in Ahrefs Link Intersect — sites linking to 2 or more of your competitors but not to you, filtered for topical relevance to your cluster (treat DR as a rough sanity check, not a target)
- Create one high-value linkable asset per cluster — original data, a proprietary framework, a truly comprehensive guide, or a practical tool. Generic "best practices" content isn't a linkable asset. Original research is MILESTONE
- Earn links where an editor chooses to link: expert quotes in industry roundups, partner and customer pages, and contributions to publications you'd be happy to appear in. Skip link swaps, paid placements and guest posts written mainly for the anchor text; Google's spam policies treat those as manipulation
- Claim the profiles your buyers actually use: industry directories, chamber listings, review sites. They rarely move rankings, but they confirm the business is real and keep your details consistent across the web
- If you have a professional network, email 5 to 10 contacts with a specific, personal ask to link to your best resource — not a blast, a targeted request to people who already know you
- Set up Google Business Profile if you serve customers locally or in a defined service area. It needs a legitimate local presence, so most pure SaaS products can skip it
- Start building external social proof: LinkedIn company page with consistent posts, product/content mentions on forums and niche communities, industry association profiles
Week 11–13 — Day 76–90: AI Visibility, Content Refresh, and 90-Day Review
- Run your first manual AI citation check: test 10 to 15 target queries in ChatGPT Search, Perplexity, and Google AI Overviews — note which queries trigger AI answers and whether your site is cited
- For queries where AI is answering but not citing you, look at which pages are cited and what they have that yours doesn't: original numbers, a clearer opening answer, a bigger footprint on third-party sites. Don't copy formatting patterns on the assumption that they cause citations
- Refresh Phase 2 content based on GSC data: find articles sitting in positions 11 to 20 and update them with expanded FAQ sections, additional cited data, and stronger E-E-A-T signals — these are your closest shots at page one MILESTONE
- Check that Article markup on your key pages carries a real author, publish date and modified date, and that it matches what's visible on the page. Skip HowTo markup; Google retired those rich results
- Publish a third topic cluster pillar if you have capacity — or deepen existing clusters with 2 to 3 sub-topic articles targeting keyword gaps you've spotted in GSC impression data
- Complete your 90-day review: document GSC impressions trend, pages generating clicks, ranking positions for all tracked keywords, and referring domains acquired
- Set 30/60/90-day targets for months 4 to 6 based on what the data from phases 1 to 3 has shown you about your competitive landscape
A project management SaaS launched in January 2024 with no domain history, no backlinks, and a single product landing page. During Phase 1, we caught a JavaScript rendering issue that had been preventing Googlebot from reading their dynamically loaded content — found in week 1 via URL Inspection, fixed by week 2. Phase 2 targeted two clusters: "project management for remote teams" and "agile sprint planning templates" — both with KD under 20 and clear long-tail sub-queries. Eleven articles across those two clusters by day 60.
By day 70, GSC was showing 1,200 monthly impressions on 8 of the 11 articles, and 3 had moved into positions 8 to 14 on their target queries. Phase 3 link building — 4 guest posts and 2 original data pieces — added 11 referring domains by day 90. By month 5, those same 3 articles had climbed to positions 3 to 7. By month 11: 8,400 monthly organic visitors across 34 ranking articles. Without the Phase 1 rendering fix, every one of those 11 articles would have been invisible to Googlebot for the entire programme. That one week-one audit saved the whole project.
8. Keyword Strategy for New Websites: How to Pick Winnable Targets
Keyword selection is the most important decision you'll make in the first 90 days. A new domain with no backlink history can't compete for competitive head terms. Picking keywords your site can actually reach the top 10 for within 90 days is what separates launches with early wins from ones where months of work produce nothing measurable.
The keyword difficulty threshold for new sites
On a brand-new domain with zero backlinks, start with low-difficulty keywords: roughly under 15 in Ahrefs, or under 40 in Semrush, for your first 60 days. Treat that as a screen, not a forecast. Tools calibrate difficulty differently. A KD of 8 can still be out of reach if every result comes from a heavyweight site, while a KD of 20 can be winnable when page one is thin or off-intent. Read the SERP before you commit.
| Keyword Type | Example | When to Target | Expected Ranking Timeline |
|---|---|---|---|
| Ultra-long-tail (5+ words) | "sprint planning template for remote engineering teams" | Days 1–60 (Phase 1 and 2) | 2–6 weeks if content quality is high and technical foundations are solid |
| Long-tail (3–4 words) | "agile sprint planning templates" | Days 30–90 (Phase 2 and 3) | 6–14 weeks; may fluctuate before stabilising |
| Mid-tail (2–3 words) | "sprint planning software" | Month 3–6; requires 5+ referring domains | 3–6 months; requires domain authority building |
| Competitive head terms (1–2 words) | "project management software" | Month 6–18; requires significant link authority | Often 12 months or more; rarely a realistic 90-day target for a new domain |
A five-minute SERP check before you commit to a keyword
Difficulty scores summarise backlinks, not the whole picture. For each shortlisted keyword, open the top five results and read them the way a competitor would: note the page type, how current and thorough each one is, how many sites link to it, and which questions it leaves unanswered. Then sort the keyword into one of the two columns below.
| Signal | Go after it | Wait or skip |
|---|---|---|
| Quality of top results | Thin, dated or only half on-topic | Current, thorough and exactly on-intent |
| Links pointing at them | A handful of referring domains | Dozens or more, from respected sites |
| Who is ranking | Small sites, forums or Reddit threads | Only established brands |
| Result format | Text articles in the top five | Video, shopping or tool results dominate |
| Fit with your offer | Connects to something you sell or can demonstrate | Pure curiosity query with no next step |
Use the "go" keywords for your first clusters, park the "wait" ones until month 4 or later, and drop the rest. It's the same check that caught the video-dominated queries in the Day 45 box earlier.
Question-based keywords: a double opportunity
In 2026, question-based keywords deserve more than a volume number. Ahrefs' analysis of 146 million SERPs found AI Overviews on 57.9% of question queries and 46.4% of queries with seven or more words. Those long phrases also tend to be less competitive, so a parallel list of question-style targets gives a new site a realistic shot at both a classic ranking and an AI citation. Write each one as a direct answer first, with the detail underneath, and use the featured snippets guide for the formatting details.
9. Content Strategy: Quality, Clusters, and E-E-A-T from Day One
In 2026, content quality means something specific: content that matches search intent precisely, demonstrates real expertise, and couldn't have been produced by dumping a prompt into an AI tool. Every competitive SERP already has AI-assisted content in it, often with high NLP scores. What actually differentiates is the layer above that: original data, first-hand observations, expert perspective that can't be assembled from what's already on the internet.
The 90-day content target
For most new sites, 10 to 15 articles in 90 days is the right target — not 30 to 50. Each piece should genuinely satisfy the search intent for its specific query, not hit a word count. The right question for every article isn't "is this long enough?" — it's "does this fully answer what someone searching this query actually wants?"
The split: 2 to 3 pillar articles (1,800 to 2,500 words) targeting cluster head terms, and 8 to 12 cluster articles (800 to 1,400 words) targeting specific long-tail sub-queries. Every cluster article links to its pillar. Every pillar links to all its cluster articles. This internal linking structure is how you signal topical authority to Google's crawlers.
10. Money Pages Before Blog Posts
The cluster model above can quietly turn a startup into a small publishing house. Your first rankings should come from pages that sit close to a sale or signup, with articles feeding them rather than the other way round. A visitor comparing options is worth more than one reading for fun.
For most startups, this is the order that works:
- Homepage. Says what you do and who it's for on the first screen.
- Core product or service page. One page per main offer, written around the phrase buyers actually search.
- Use-case and audience pages. "Project management for remote teams", "for agencies".
- Integration pages (software). One per tool you connect to, where people search for the pairing.
- Comparison and alternative pages. "X alternative" or "X vs Y". Be fair about who each option suits.
- Supporting articles. Answers to buyer questions that link into the pages above.
A project management SaaS might publish /project-management-software, /project-management-for-remote-teams and /asana-alternative first, then write sprint-planning articles that link into them. Built in that order, the blog adds to the sales pages instead of competing with them. The internal linking guide shows how to wire the two together.
| Business model | Build these first |
|---|---|
| B2B SaaS | Use-case pages and comparison or alternative pages, then integrations |
| Developer tool or API | Documentation, integration pages and worked use cases |
| E-commerce | Category pages and best-selling product pages before any buying guides |
| Local or service business | Service pages for each area you cover, plus a Google Business Profile |
| Marketplace | Category and location pages that have real listings behind them |
| Agency or consultancy | Service pages, industry pages and a few detailed case studies |
11. Building E-E-A-T on a New Domain
E-E-A-T (Experience, Expertise, Authoritativeness, Trustworthiness) isn't a single ranking factor. Google describes it as the set of qualities its systems try to recognise when judging whether content is helpful and reliable, with trust at the centre (see its guidance on helpful content). For a new domain, that means giving readers and Google visible proof of who is behind the content and why they can be believed. The E-E-A-T guide goes deeper on each signal.
| E-E-A-T Signal | Implementation for New Sites | Priority |
|---|---|---|
| Named authorship | Every article needs a named author with a linked credential page. The author page should include job title, years of experience, specific areas of expertise, and at least one external link to a verifiable profile (LinkedIn, professional association, or published work). | 🔴 Critical — Day 1 |
| First-hand experience signals | Each article needs at least one specific observation from direct experience — a named client outcome, a measurement, a real decision in a real situation. "In my experience" without any specifics doesn't count. | 🔴 Critical — Every article |
| Author schema markup | Add Person schema to every author bio page with name, jobTitle, knowsAbout, and url. Reference the author @id in Article schema on every piece they write. This gives Google a machine-readable entity connection between author and content. | 🟡 High — Week 3 |
| Cited primary sources | Every statistic, data claim, or research finding links to the original primary source — not a secondary aggregator. "According to studies" or unnamed research is a red flag for quality evaluators. | 🟡 High — Every article |
| Organization schema and entity establishment | Organization schema on the homepage with name, url, logo, contactPoint, and sameAs properties pointing to verified external profiles. sameAs should include your LinkedIn company page, Crunchbase, Google Business Profile, and any verified directory listings. | 🟡 High — Week 1 |
| External validation mentions | Any press mention, directory listing, podcast appearance, or award citation gets a "Featured in" or "As seen in" reference on the homepage or About page — with a link. Even small external mentions contribute to the prominence dimension of E-E-A-T. | 🟢 Medium — Phase 3 |
12. Making Startup Content Easy for AI Search to Cite
Google's documentation on AI features says a page needs to be indexed and eligible to show a snippet, and that there are no additional technical requirements for AI Overviews or AI Mode. There's no secret layer to bolt on. What you can do is write pages that are easy to understand and worth citing, and build the off-site footprint that AI answers draw on. The wider move towards answers on the results page is covered in the zero-click searches guide.
Six habits that make a new site easier to cite
Where people genuinely ask a question (People Also Ask and AlsoAsked show which), use it as an H3 and follow it with a 2 to 4 sentence direct answer before the detail. It helps readers first, and it makes the passage easy for any system to lift.
A named source, a date and a figure beat a vague claim. "Ahrefs, 2026: AI assistants sent 0.33% of site traffic against 34% from search" is more useful to a reader, and to an AI system, than "most traffic still comes from search".
Name the tools, standards and organisations you're discussing, and keep your own name, one-line description and founder profiles identical across your site, LinkedIn, Crunchbase and G2. Ahrefs points out that conflicting details are how AI answers end up wrong about a brand. The semantic SEO and entity guide covers the mechanics.
In Ahrefs' analysis of 75,000 brands, YouTube and general web mentions correlated with AI visibility more strongly than backlinks. Claim profiles on the review sites your buyers use (G2, Capterra, Trustpilot, Clutch). SE Ranking found domains with those profiles were cited by ChatGPT about three times as often. It's a correlation, but a cheap box to tick.
A short FAQ is worth having when readers really do ask follow-ups. Google no longer shows FAQ rich results and doesn't ask for special markup for AI features, so don't add FAQPage schema as an AI tactic. If your CMS adds it automatically, leave it; it's harmless.
Ahrefs found AI-cited pages are about 25.7% fresher than pages ranking organically, and SE Ranking saw pages updated within three months cited more often. Update top pages with new data or corrected steps. Bumping the date without changing anything doesn't count.
13. Recommended Tool Stack by Budget
| Budget Tier | Tool | Cost | What It Covers |
|---|---|---|---|
| 🪙 Bootstrap $0–$50/month |
Google Search Console | Free | Indexing status, keyword impressions, Core Web Vitals, URL inspection — the most important tool on this list, at any budget |
| Google Analytics 4 | Free | Traffic sources, user behaviour, conversion events — non-negotiable from day 1 | |
| PageSpeed Insights | Free | Core Web Vitals diagnosis with specific, actionable recommendations for each issue | |
| Ahrefs Webmaster Tools | Free (verified domains) | Backlink monitoring and basic keyword data for your own domain | |
| Screaming Frog SEO Spider | Free (≤500 URLs) | Technical crawl to catch indexing issues, broken links, and on-page problems before they compound | |
| AlsoAsked.com | Limited free | Question-based keyword discovery for FAQ sections and People Also Ask targeting | |
| Google Rich Results Test | Free | Checks eligible structured data before it goes live (FAQ support was removed from the tool in June 2026) | |
| Claude Pro or ChatGPT Plus | ~$20/month | Schema JSON-LD generation, batch title tags, meta description drafts, content briefs — probably the best-value $20 in this stack for a new site | |
| 📈 Early-stage startup $100–$250/month |
Ahrefs Starter or Semrush Pro | $29–$139/month | Keyword research, competitor analysis, backlink monitoring. Ahrefs for backlink quality; Semrush for broader keyword strategy tools |
| Screaming Frog (paid) | $259/year (~$22/month) | Full site crawls without the 500-URL cap, plus bulk AI meta generation via the GPT API for title and description drafts | |
| Frase or NeuronWriter | $23–$45/month | Content brief generation and NLP scoring — cuts article research and on-page optimisation from 3 to 4 hours down to under 1 | |
| Google Looker Studio | Free + GA4/GSC connectors | A solid performance dashboard for weekly monitoring without any extra spend | |
| 🚀 Funded startup $300–$600/month |
Semrush Pro or Ahrefs Standard | $139–$249/month | Full keyword research, clustering, competitor gap analysis, and position tracking |
| Surfer SEO | $99/month | Real-time NLP content scoring against the SERP top 20 while you write — worth it once you're producing 6+ articles a month | |
| Screaming Frog + ContentKing Starter | ~$80/month total | Monthly crawl audits plus real-time monitoring to catch technical regressions the moment they happen | |
| Hunter.io Starter | $49/month | Contact discovery and email verification for Phase 3 link outreach | |
| Otterly | $99/month | Google AI Overview citation tracking — a good entry-level GEO monitoring tool if your team is starting to measure AI visibility alongside organic |
What your week can actually afford
Money isn't the only limit, hours are too. This is a planning example for a founder with about 8 hours a week, roughly 100 hours across the 90 days. It's a budgeting aid rather than a benchmark, so swap in your own numbers.
| Workstream | Hours over 90 days | What it buys |
|---|---|---|
| Technical setup and fixes | 12–15 | A crawlable, indexable site with Search Console and GA4 working |
| Keyword and SERP research | 10–12 | 20–30 "go after it" keywords mapped to pages |
| Money pages | ~20 | Homepage, core offer pages and 2–3 use-case pages |
| Supporting articles | 35–40 | 6–8 articles at roughly 5 hours each |
| Links and outreach | 8–10 | One linkable asset and a few real relationships |
| Reporting and refresh | 8–10 | A weekly Search Console check and one refresh cycle |
At this budget you won't publish 30 posts, run a big outreach push and produce original research all at once. Pick two and do them properly.
14. Milestones and What to Measure at Each Stage
| Timeframe | Target Milestone | What to Measure | If Milestone Is Missed |
|---|---|---|---|
| Day 7 | GSC property verified; sitemap submitted; robots.txt confirmed not blocking Googlebot | GSC Pages (indexing) report shows pages submitted; no "Blocked by robots.txt" errors present | Pause all content work. Fix the technical issues first — there's no point publishing onto a site Google can't read |
| Day 21 | Core Web Vitals passing on mobile; schema on homepage and key pages validated | PageSpeed Insights showing "Good" for LCP, INP, CLS on mobile; structured data validating without errors | Sort image optimisation and caching before moving to Phase 2 content publishing |
| Day 45 | First GSC impressions appearing for target keywords; 6–8 articles indexed | GSC Performance report showing impressions on target keywords; URL Inspection checked on every priority page, with most showing "Indexed" | Check non-indexed pages for noindex tags, crawl errors, or JavaScript rendering issues |
| Day 60 | 2–3 articles ranking in positions 1–20 for target long-tail queries; impressions growing week-on-week | Rank tracking showing positions for 5+ target keywords; GSC impressions trending upward over the past 4 weeks | Revisit your keyword difficulty assessments — you may have targeted queries that are too competitive for the current domain |
| Day 90 | 5+ referring domains; 10+ pages generating GSC impressions; at least 1 page generating organic clicks | Ahrefs or Semrush referring domain count; GSC pages with clicks; rank tracking showing improving position trends | Audit content quality and E-E-A-T signals on your top-impression, zero-click pages; push link building harder |
| Month 6 | Measurable organic traffic growth; multiple pages in positions 1–10 for long-tail queries; a growing set of relevant referring domains (15+ is an example target, not a rule; the right number depends on your SERPs) | GA4 organic sessions trend; conversion events from organic traffic; total ranking keywords; referring domain velocity | Run a full content quality audit — find which articles have impressions but no clicks and refresh them with stronger titles and E-E-A-T signals |
The business numbers behind the SEO numbers
The milestones above track search health. Founders also need to see whether search is producing demand, so keep these beside them.
| Metric | Day 30 | Day 60 | Day 90 |
|---|---|---|---|
| Priority pages indexed | Baseline recorded | Live pages submitted and checked for issues | Most priority pages indexed; blockers on the rest diagnosed |
| Queries with impressions | Baseline | First target queries appearing | Growing week on week |
| Queries ranking 11–30 | Baseline | A few | Your refresh list for the next phase |
| Organic clicks | Likely near zero | First clicks | A steady upward trend |
| Organic leads or signups | Tracking confirmed | First signals | Measurable, even if small |
| Referring domains | Baseline | First earned links | 5 or more |
| AI citations (manual sample) | Baseline test | Skip | Re-test and note changes |
Rankings and clicks are leading indicators. What pays the bills is qualified organic demand: demo requests, signups, quote requests. Set up conversion events in week 1 and read them in the same weekly check as Search Console.
15. The 8 Most Common Mistakes on New Site SEO Launches
| # | Mistake | Why It Happens | How to Avoid It |
|---|---|---|---|
| 1 | Publishing content before technical audit | Pressure to ship something; underestimating how often staging-to-live technical errors slip through | Hard rule: crawlability, indexing, robots.txt, and sitemap are all confirmed clean before the first article goes live |
| 2 | Targeting competitive keywords immediately | Head terms have higher volume and feel more valuable; going after long-tail queries feels like settling | Use KD as a first screen (around 15 in Ahrefs for the first 60 days), then read the SERP. Competitive terms belong on a month 6+ roadmap, not month 1 |
| 3 | Anonymous or generic authorship | Discomfort with putting names on content; assuming "the company" counts as an author; no time to build author pages | Every article needs a named author with a linked credential page from day 1. This is core E-E-A-T infrastructure, not a nice-to-have |
| 4 | Spreading too thin — too many clusters | Wanting to cover the whole market; fear of leaving keyword opportunities on the table | Stick to 2 clusters in the first 90 days. Five to six articles per cluster builds real depth. Twelve articles across twelve topics builds nothing |
| 5 | Ignoring the questions buyers actually ask | Writers start from what the company wants to say rather than from what searchers type | Where readers have natural follow-ups, add a short FAQ with H3 questions and direct 2-to-4 sentence answers. Schema is optional now that Google has ended FAQ rich results |
| 6 | Starting link building in the first 30 days | Knowing backlinks matter; impatience with the content-first approach | Links pointing at a 3-page site have minimal effect. Wait until 10+ articles are live and indexed before any outreach begins |
| 7 | Treating SEO as a one-time setup | Wanting to "finish" SEO; not accounting for the ongoing work as algorithm updates, indexing issues, and competitive gaps emerge | Build a recurring weekly routine: GSC impressions review (30 min), one new article, one existing article refreshed based on data |
| 8 | Building a blog before the money pages | Articles feel easier to ship than product, use-case and comparison pages, and the content-cluster model makes blogging look like the whole job | Put the homepage, core offer pages and use-case pages live first, then write articles that link into them (Section 10) |
16. 90-Day SEO Launch Checklist
Phase 1: Technical Foundations (Days 1–30)
- HTTPS installed and all pages serving on secure protocol
- Google Search Console property verified and connected
- Google Analytics 4 installed with conversion events configured
- XML sitemap generated and submitted to GSC
- robots.txt checked — Googlebot not blocked from content, CSS, or JS
- Every page checked for accidental noindex tags (critical if migrated from staging)
- URL Inspection run on 5 representative pages, indexing status checked and any issues logged
- Core Web Vitals passing on mobile (LCP <2.5s, INP <200ms, CLS <0.1)
- All images compressed and served in WebP format
- Mobile layout tested on at least 3 device sizes
- Organization or LocalBusiness schema on homepage
- BreadcrumbList schema on all pages with visible breadcrumb navigation
- About page live with named team members, credentials, and external profile links
- Privacy policy and terms of service published
- URL structure locked in — short, keyword-containing slugs
- 404 page created with navigation links
- Keyword research done — 30 to 50 candidates screened by KD and a manual SERP check, with the "go after it" list mapped to pages
- 2 topic clusters defined with pillar + cluster keyword architecture mapped
- Page order decided: homepage, core offer pages and use-case pages ahead of most articles
- 90-day content calendar built with publishing dates and assigned keywords
Phase 2: Content Launch (Days 31–60)
- Homepage, core product or service page and 2–3 use-case pages live
- Pillar article for cluster #1 published (1,800–2,500 words, KD roughly 15–25)
- 4–6 cluster articles supporting pillar #1 published (KD under 15)
- Pillar article for cluster #2 published
- 3–4 cluster articles supporting pillar #2 published
- Every article has a named author with a linked credential page
- Articles with natural follow-up questions carry a short FAQ (H3 questions, direct answers); FAQ schema optional
- Every article has at least one cited statistic with a primary source link
- Every article has at least one first-person experience observation
- New URLs linked internally and included in the sitemap; URL Inspection used for the priority ones
- Rank tracking set up for all 30–50 target keywords
- Internal linking checked: every article links to ≥2 internal pages; ≥1 existing page links to it
- GSC impressions checked at day 45–50 — impressions showing on target keywords
- SERP format checked for any articles not generating impressions — content format matches what's ranking
Phase 3: Authority and AI Visibility (Days 61–90)
- One high-value linkable asset created per cluster (original data, proprietary framework, or practical tool)
- 20–30 link prospects identified using Ahrefs Link Intersect
- A few links earned from relevant sources (expert quotes, partners, original data), with no paid or exchanged links
- Industry directory and Chamber of Commerce submissions done
- Google Business Profile created and verified (if applicable)
- Trustpilot, G2, Clutch, or sector-specific review platform profiles created
- Organization schema sameAs properties updated with all new directory URLs
- First AI citation check done: 10–15 target queries tested in ChatGPT, Perplexity, and Google AI Overviews
- Phase 2 articles sitting in positions 11–20 refreshed with expanded FAQ, stronger E-E-A-T signals, and updated data
- 90-day review done: GSC impressions trend, referring domains, ranking positions, and organic sessions documented
- Months 4–6 targets set based on 90-day data — competitive keywords moved into the active targeting roadmap
- Founder KPI view set up: leads or signups tracked beside rankings and clicks
- Treat competitive head terms (KD 30+) as a later-stage target. Move them into active targeting only after your site has built enough topical authority, relevant content and referring-domain strength to compete with the current SERP
- Never publish articles with no named author. Anonymous or generic authorship makes it harder to demonstrate experience, expertise and accountability on a new domain
📚 Sources and Primary Research References
| Source | Key Data Used |
|---|---|
| Ahrefs — 96.55% of Content Gets No Traffic From Google | About 14 billion pages studied; 96.55% get no organic Google traffic. Used in Section 9. |
| Ahrefs — How Long Does It Take to Rank in Google? | 2 million new pages studied; 5.7% reached the top 10 within a year, typically after 2 to 6 months. Older data, used as a directional guide. |
| Backlinko — Google CTR Statistics | Clicks fall steeply by position (27.6% at position 1, 2.4% at position 10 in Backlinko's figures). Varies by study and by AI Overview presence. |
| Ahrefs — SEO for Startups (updated April 2026) | 3 to 6 months to meaningful progress; AI assistants sent 0.33% of site traffic vs 34% from search; 76% of AI Overview citations also rank top 10; hourly cost figures. |
| Ahrefs — AI Overview triggers | AI Overviews show on roughly 21% of US keywords in 2026. Other trackers report different shares. |
| Ahrefs — How to Rank in AI Overviews | 146 million SERPs: AI Overviews on 57.9% of question queries and 46.4% of queries with 7+ words. |
| Ahrefs — AI brand visibility correlations | 75,000 brands: YouTube and web mentions track AI visibility more closely than backlinks or DR. |
| SE Ranking — ChatGPT citation factors | 129,000 domains: citations rise with referring domains, recent updates and review-platform profiles (about 3x). Correlational. |
| Google Search Central — AI features and your website | Indexed, snippet-eligible pages can appear in AI Overviews and AI Mode; no extra technical requirements or markup. |
| Google Search Central — Helpful, reliable, people-first content | E-E-A-T describes qualities Google's systems look for, with trust most important; it is not a single ranking factor. |
| Google Search Central — Spam policies | Link spam (paid links, excessive link exchanges) and scaled content abuse. |
| Google Search Central — Ask Google to recrawl your URLs | Repeat URL Inspection requests don't speed up crawling; indexing isn't guaranteed. |
| Search Engine Journal — Google drops FAQ rich results | FAQ rich results stopped on 7 May 2026; FAQ markup can stay without harm. |
| Ahrefs — SEO pricing study | Hourly rates for agencies, consultants and freelancers, plus in-house salary benchmarks. |
Frequently Asked Questions
How long does SEO take for a new website in 2026?
Most new sites see their first Search Console impressions within 30 to 60 days of getting the technical basics and content right, and real organic traffic usually follows between months 3 and 6. In competitive industries, 6 to 12 months is more realistic. Ahrefs' study of 2 million newly published pages found only 5.7% reached the top 10 within a year, which is why keyword selection in the first 90 days matters so much.
What is the most important SEO task for a new website?
Technical foundations come first. A site Google can't crawl and index returns nothing on any content or link work. The day-one essentials are HTTPS, an XML sitemap submitted in Search Console, a robots.txt that isn't blocking Googlebot, and Core Web Vitals passing on mobile. After that, the highest-leverage task is picking keywords you can actually win: start with low-difficulty long-tail queries, then confirm by reading the SERP.
Should a startup focus on SEO or paid advertising first?
You don't have to pick one. SEO compounds over time but takes 3 to 6 months to build momentum. Paid search gives you conversion data fast, so you can learn which keyword clusters drive revenue while organic is still building. A sensible split for a budget-conscious startup is technical foundations and low-competition content in months 1 to 3, plus a modest paid budget to test commercial keywords in parallel (see our Google Ads guide for campaign setup). By months 4 to 6, organic can start reducing paid spend on the terms where it has gained traction.
Should a startup prioritise product pages or blog posts?
Product and service pages first. They target searchers closest to a purchase, and your articles should link into them. A sensible order is the homepage, core offer pages, use-case pages, comparison or alternative pages, and then supporting articles that answer buyer questions. Section 10 sets out the full order and what to build first for each business model.
How much does startup SEO cost?
It depends on who does the work. Ahrefs' pricing research puts agencies at about $135 an hour on average, SEO consultants at about $122, freelancers around $68, and an in-house specialist at roughly $71,000 a year in the US. Doing it yourself costs time rather than cash: Search Console and GA4 are free, and a paid keyword suite is usually the only essential tool in the first 90 days.
What is the Google Sandbox and does it still exist in 2026?
"Google Sandbox" is the informal name for something many new site owners notice: new domains often struggle to rank for competitive terms in their first few months, even when content and technical setup are solid. Google doesn't describe it as a separate mechanism, and it's better thought of as a trust-building window, since a new domain has no crawl history, no links and no track record. The practical response is to start with very low-competition, long-tail queries for the first 60 days, then move into mid-difficulty terms as your content earns impressions and links.
How many blog posts should a new website publish in the first 90 days?
Quality beats volume. Ten to twelve well-researched articles in one topic cluster usually outperform 30 thin posts spread across unrelated topics. Count your money pages separately: the homepage, core product or service page and a few use-case pages come before most blog posts. The target in this guide is 10 to 15 foundational articles across 2 topic clusters, plus 2 to 3 deeper pillar pieces in Phase 3.
Should a startup do SEO before product-market fit?
Often yes, if people already search for the problem you solve and you can name your buyer. A few money pages and the technical basics are cheap and start compounding straight away. Hold off on a large content programme if the category is brand new, nobody searches for the problem, or your positioning shifts every few weeks, because you'd be building pages around messaging you're about to rewrite.
How does AI search affect SEO strategy for new websites in 2026?
Google says AI Overviews and AI Mode need nothing beyond normal Search eligibility, with no special markup or files. For a startup, the practical effect is that AI answers can cut clicks on informational queries, while the pages they cite tend to be ones that already rank well (Ahrefs found 76% of AI Overview citations also rank in Google's top 10). So build the SEO basics, write clear direct answers backed by your own evidence, and grow your brand's footprint on third-party sites, YouTube and review platforms.
The full AI citation strategy referenced in Phase 3 — six universal citation signals, platform-specific optimisation for Google AI Overviews, Perplexity, and ChatGPT Search, and how to measure AI share of voice for a new domain.
Read the GEO & AEO guide →The underlying keyword research methodology behind this playbook's targeting approach — intent classification, question-based keyword discovery, cluster architecture, and conversational query research for AI platforms.
Read the keyword guide →The deep-dive into the E-E-A-T implementation this playbook introduces in Phase 1 — author schema, named authorship, entity establishment, digital PR as a trust signal, and the credibility signals that help both readers and AI systems trust a source.
Read the E-E-A-T guide →The master reference guide covering everything this playbook builds toward — the full signal landscape across technical, on-page, off-page, and AI visibility factors, including how topical authority, E-E-A-T, and GEO connect into one unified strategy.
Read the complete SEO guide →Campaign setup for Google, Meta, LinkedIn, and YouTube — for the modest paid budget this playbook recommends testing in parallel with SEO during months 1 to 3.
Browse Paid Ads guides →How to approach answer-engine and AI-citation work when your domain has no history, with a first-months plan that sits alongside this playbook.
Read the new-site AEO & GEO guide →The structure behind the two-cluster approach in Phase 2: how to choose a pillar, size the supporting pages, and link them so the topic reads as one body of work.
Read the pillar pages guide →A fuller walk-through of the crawl, indexing and rendering checks from Phase 1, useful when Search Console shows pages that refuse to get indexed.
Read the technical SEO guide →