[POST/blog]

How to Rank on Google in 2026

Free SEO guide for founders and indie hackers, built from the Google leak, DOJ testimony, patents, and 100+ websites shipped.

Field notesBy Daniel CastellaniApril 29, 2026Updated April 29, 202648 min read

seogoogle rankingtopical authorityai overviewstechnical seoindie hackers

Free web edition. Share it, link to it, and use the printable companion at /seo-guide/checklist if you want the skim-first version.

Need the shorter supporting pieces?

Why This Guide Exists

In May 2024, a Google engineer accidentally pushed 2,596 internal API modules to a public GitHub repo. 14,014 attributes. Confirmed under DOJ antitrust testimony.

Google had denied the existence of these ranking signals for over a decade. The leak proved them wrong.

Most SEO guides you've read were written before this happened. They're recycling the same 2018-era advice (write good content, build backlinks, optimize meta tags) without knowing what actually moves rankings now.

This guide is different. Every claim here is backed by:

The May 2024 Google API leak (analyzed by Mike King, iPullRank)
DOJ v. Google antitrust testimony (sworn statements from Eric Lehman, 17-year Google Distinguished Engineer)
Patent filings (U.S. patent numbers cited)
Documented case studies (with verified traffic data)
100+ websites I've personally shipped at Publishd

If you've been told "just write quality content and the rankings will come" — that advice is incomplete and outdated. There are mechanical ranking systems Google uses that you can specifically optimize for, once you know they exist.

This guide tells you what they are.

Who this is for:

Founders shipping apps and websites
Indie hackers who don't have a $5K/month SEO retainer
Small business owners doing it themselves
Anyone who's tired of generic "SEO 101" guides

What you'll learn:

How Google actually ranks pages (the leaked architecture)
Why behavioral signals matter more than backlinks for new sites
How to build topical authority (0 → 128K visitors in 123 days, documented case)
AI Overview citation strategy (the new top-of-funnel)
Programmatic SEO for indie shippers (the unfair advantage)
A 90-day action plan with weekly milestones

If your bigger problem is that the site itself is holding you back, not just the SEO plan, start with my website design and development page. If you want the long version first, keep reading.

The 2024 Google Leak: What Changed Everything
How Google Actually Ranks (The Real Architecture)
Behavioral Signals: Confirmed Under Oath
Topical Authority: The Indie Shipper's Cheat Code
Entity SEO and N-E-E-A-T
Technical SEO That Actually Matters
Content Strategy: Passage Indexing + Freshness
AI Overviews and LLM Citation Strategy
Programmatic SEO: The Indie Hacker Unfair Advantage
Free Tools as Link Magnets
"Alternatives to X" Plays for SaaS
Link Building Without a Budget
The 90-Day Ranking Plan
When to Stop DIY and Hire Help
The 24-Point Audit Checklist

Chapter 1: The 2024 Google Leak — What Changed Everything

What Happened

May 5, 2024. A Google engineer accidentally pushed internal API documentation to a public GitHub repository. 2,596 modules. 14,014 attributes. Former Google engineer Erfan Azimi spotted it. Rand Fishkin (SparkToro) and Mike King (iPullRank) verified and published the analysis.

The technical breakdown is at ipullrank.com/google-algo-leak. I'll summarize the four findings that change how you should think about SEO.

Finding #1: NavBoost Exists (And Uses 13 Months of Click Data)

NavBoost is a ranking system that ingests user click behavior. Specifically, it tracks per query-URL pair:

goodClicks — clicks where the user found what they wanted
badClicks — clicks where the user bounced quickly
lastLongestClicks — the click that ended the search session (i.e., they got their answer)

Eric Lehman, a 17-year Google Distinguished Engineer, confirmed NavBoost exists during DOJ antitrust testimony. Under oath, he confirmed Google employees were instructed not to discuss it publicly because SEOs would game it.

Why this matters: Click-through rate, dwell time, and pogo-sticking (clicking, bouncing, clicking another result) are direct ranking inputs — not correlations.

Finding #2: siteAuthority Is Real

Google publicly denied for years that anything like "domain authority" existed in their algorithm. The leak found siteAuthority — a real, persistent, site-wide quality score stored in CompressedQualitySignals, fed into an internal system called Q*.

Why this matters: A new site with the same content as an established site will not rank as well. Site-level authority is real and accumulates over time.

Finding #3: hostAge Confirms the Sandbox

For over a decade, Google denied the existence of a "new site sandbox." The leak found hostAge, an attribute documented to "sandbox fresh spam in serving time."

Why this matters: New domains are held back. If you launched 6 weeks ago and aren't ranking, that's why. Plan for 3-6 months minimum before you see meaningful organic movement on a new domain.

Finding #4: chromeInTotal Tracks Browser Data

The leak revealed chromeInTotal, an attribute in the page quality module tracking site-level Chrome browser views. Google had repeatedly stated Chrome data was not used in ranking. It is.

Why this matters: Real user traffic to your site (not just from search) feeds into Google's quality signals. Building any kind of audience that visits your site directly — newsletter, social, podcast listeners — compounds your search rankings over time.

What Most Guides Get Wrong (And You Now Know)

Most SEO guides written before mid-2024 are based on what Google said publicly. The leak proved Google said one thing publicly and did another internally for years.

The implication: ignore Google's public statements when they conflict with documented behavior. Follow the evidence.

Sources for this chapter:

Chapter 2: How Google Actually Ranks (The Real Architecture)

Most guides describe Google as a single algorithm. It isn't. Based on the leak, court testimony, and patent filings, here's the actual architecture.

The Stack (Simplified)

When you search Google, your query passes through multiple ranking systems in sequence:

1. QUERY UNDERSTANDING
   ↓
2. INITIAL RETRIEVAL (Indexing tier)
   ↓
3. RANKING (Multiple subsystems)
   - Mustang (primary scoring)
   - RankEmbed BERT (semantic understanding)
   - NavBoost (click signals)
   - Q* (quality signal aggregation)
   ↓
4. RE-RANKING (Twiddlers — small adjustments)
   ↓
5. SERP ASSEMBLY (AI Overviews, featured snippets, PAA, etc.)

Each system has its own inputs. You're not optimizing for "Google" — you're optimizing for ~30 distinct subsystems that each weight different signals.

The Three Pillars That Actually Matter

Cutting through the complexity, every signal Google uses falls into three categories:

Pillar 1: RELEVANCE (Does this match the query?)

Keyword presence (still matters, but less than 2010)
Semantic match (BERT/embeddings — does the meaning match?)
Entity presence (is the right "thing" being discussed?)
Passage relevance (does any section of the page directly answer the query?)

Pillar 2: AUTHORITY (Should we trust this source?)

Backlink profile (who links to you, with what context)
siteAuthority (the site-wide quality score, leaked)
Entity recognition (are you a known entity in the Knowledge Graph?)
Brand search volume (people searching for you by name)

Pillar 3: USER SATISFACTION (Did this actually help?)

NavBoost click signals (goodClicks, badClicks, lastLongestClicks)
Dwell time
Return-to-SERP rate (pogo-sticking)
Direct traffic (chromeInTotal — Chrome views)

The mental model: You can win Pillar 1 with content. You build Pillar 2 over time. You earn Pillar 3 by actually being useful.

A guide that only optimizes Pillar 1 (keywords, meta tags) will hit a ceiling at page 2-3. Cracking page 1 requires all three.

Why New Sites Struggle

The combination of hostAge (sandbox) + siteAuthority (slow to build) means new sites face a real disadvantage in the first 3-6 months. This isn't a conspiracy — it's a documented anti-spam mechanism.

The workaround: Topical authority (Chapter 4) lets you bypass site-wide authority constraints by becoming the obvious answer in a narrow niche. Generalists wait years for site authority. Specialists rank in months.

Chapter 3: Behavioral Signals — Confirmed Under Oath

The Court Testimony

The U.S. v. Google antitrust trial (2023-2024) produced sworn testimony that confirmed what SEOs had suspected for a decade.

Eric Lehman, 17-year Google engineer, testified that click data flows into ranking via three systems:

NavBoost (the system from the leak)
Glue (separate click-data system)
RankEmbed BERT (trained on real user interaction data)

He testified that Google deliberately avoided public acknowledgment of these systems.

Rand Fishkin's CTR Experiment

In 2014, before any of this was public knowledge, Rand Fishkin ran a live experiment at SEO conferences:

Direct an audience of 1,500+ to search a specific query
Have everyone click result #4
Have everyone immediately bounce from result #1 (the current top result)

Within 30 minutes, results #1 and #4 swapped positions. The effect lasted ~2 days.

He ran the experiment 11-13 times. 7-8 confirmed successes.

The leak provided the mechanism for what he was observing: NavBoost was reading the unusual cluster of goodClicks to result #4 and badClicks to result #1, and adjusting in real time.

What This Means for Your SEO

Behavioral signals are now first-class ranking inputs. You optimize for them like this:

Optimization 1: Title Tags as Ad Copy

Your title tag isn't just a label — it's the headline that determines whether someone clicks your result vs. a competitor's.

Brian Dean's analysis shows CTR-optimized title rewrites lift clicks 20-40% on existing rankings. That CTR lift then becomes a positive NavBoost signal that lifts your ranking position further.

The compounding effect: Better title → more clicks → NavBoost reads "this satisfies users" → ranking improves → more impressions → more clicks → more NavBoost signal.

Title tag patterns that win:

Specific numbers: "How to Rank in 2026: 14 Tactics That Worked for 100+ Sites"
Curiosity gaps: "What Google's Leaked Documents Reveal About Ranking"
Time bounds: "I Ranked This Site in 90 Days. Here's the Playbook."
Negative framing: "Why Most SEO Guides Are Wrong (And What Actually Works)"

What loses:

Generic: "SEO Tips for Beginners"
Vague: "How to Improve Your Website"
Keyword-stuffed: "SEO Guide SEO Tips SEO Strategy SEO 2026"

Optimization 2: First 3 Sentences = Stay/Leave Decision

When someone clicks your page, they decide within 3-5 seconds whether they're in the right place. If they bounce back to the SERP, that's a badClick.

The first 3 sentences of any page need to confirm:

The reader is in the right place (mentions their query/problem)
There's a payoff coming (specific value)
It's actually different from what they just saw on competitors

Bad opening:

"Search engine optimization is the practice of optimizing your website to rank higher in search engine results. SEO has been around since the early days of the internet..."

Good opening:

"Most SEO guides written before May 2024 are wrong. That's the month a Google engineer accidentally leaked the internal API. This guide rebuilds the playbook from what we now know."

The good version: confirms you're in the right place (SEO, ranking), promises payoff (real playbook), differentiates (post-leak knowledge).

Optimization 3: Reduce Pogo-Sticking with Layered Information

Pogo-sticking is when a user clicks your result, glances, returns to the SERP, clicks the next result. This is the worst behavioral signal — it tells Google your page didn't satisfy.

Layer your content:

Top: Direct answer in first 100 words (catches scanners)
Middle: Detailed explanation (catches readers)
Bottom: Expert depth (catches researchers)

If a scanner gets the answer fast, they don't bounce. If a reader gets depth, they don't bounce. If a researcher needs everything, they don't bounce.

Optimization 4: Build Direct Traffic (chromeInTotal)

The leak confirmed Chrome browser views feed into quality signals. Sites with real audiences ranking direct traffic — not just search traffic — get an authority lift.

How to build it:

Email newsletter (drives Chrome views every send)
Active Twitter/LinkedIn (drives clicks to your site)
Podcast appearances (referral traffic from listeners)
Reddit/Hacker News presence (spike traffic events)

Every direct visit is a positive signal that compounds over months.

Sources for this chapter:

Chapter 4: Topical Authority — The Indie Shipper's Cheat Code

The Case Study

Koray Tugberk Gubur (Holistic SEO) documented a case study in May 2022:

New site, no significant backlinks
0 → 128,000 organic monthly visits in 123 days
One topic cluster, comprehensively covered

Full methodology at holisticseo.digital/theoretical-seo/topical-authority.

How Topical Authority Beats Domain Authority

Conventional SEO advice says: build domain authority over years, then you can rank for anything.

Topical authority breaks this rule. The leaked siteAuthority module appears to be scoped by topic, not just site-wide. A site that completely covers a narrow topic gets treated as authoritative for that topic — even if its overall domain age is 6 months.

The mechanism:

Google identifies what topic your site is "about"
Within that topic, it measures coverage completeness
Sites with high topical coverage get ranking lifts within that topic
Sites that are scattered (covering many topics shallowly) get weaker signals everywhere

The Topical Map

Before writing anything, map the topic space.

Step 1: Identify your topic cluster

Pick a single, narrow topic. Not "marketing." Not "fitness." Something like:

"Conversion-focused web design for SaaS startups"
"Habit tracking app reviews and comparisons"
"Indie hacker SEO strategies"
"Nutrition for endurance athletes"

The narrower, the better. Topical authority compounds faster on narrow topics.

Step 2: Build the content map

For your chosen topic, list every:

Head term (1-2 word queries)
Mid-tail variation (3-5 word queries)
Long-tail question (full sentences people search)
Adjacent topic (related concepts)
Use-case variant (industry, audience, scenario splits)

For example, if your topic is "habit tracking for indie hackers":

HEAD TERMS:
- habit tracking
- habit tracker app
- habit tracker

MID-TAIL:
- best habit tracker app 2026
- habit tracker for entrepreneurs
- habit tracker with analytics

LONG-TAIL QUESTIONS:
- how to build a habit that sticks
- why habit trackers fail
- difference between streaks and consistency in habits

ADJACENT:
- atomic habits methodology
- behavioral psychology of habits
- habit stacking strategy

USE-CASE:
- habit tracking for fitness
- habit tracking for writers
- habit tracking for remote workers

That's 15+ articles right there. A complete topical map for indie habit tracking might be 50-80 articles. That sounds like a lot — it's actually 12-18 months of content if you publish 1 post/week.

Step 3: Identify the pillar page

The pillar page is the master article that:

Ranks for your head term (or tries to)
Links out to every supporting article
Gets linked back from every supporting article
Acts as the structural center of your topic cluster

Example: "The Complete Guide to Habit Tracking for Indie Hackers" links out to all 50 supporting posts. Each supporting post links back to the pillar.

Step 4: Internal linking the cluster

Within the cluster, every post should:

Link UP to the pillar
Link SIDEWAYS to 3-5 related posts
Link DOWN to deeper dives if applicable

Orphan content (posts not internally linked into the cluster) doesn't contribute to topical authority. Don't publish orphans.

Why This Works for Indie Shippers Specifically

You probably can't outrank Forbes or HubSpot on broad terms like "marketing tips." They have domain authority you can't match in 5 years.

But you can absolutely outrank them on narrow, specific topics. Forbes covers "habit tracking" in one shallow listicle. You cover it in 80 deep, interconnected posts. Google's topical authority signal favors you.

This is the cheat code. Pick a niche so narrow you'd be embarrassed to tell people about it. Cover it completely. Win the topic.

Sources for this chapter:

Holistic SEO — Topical Authority

Chapter 5: Entity SEO and N-E-E-A-T

The Knowledge Graph Track

Google's algorithm has two parallel tracks:

Link/PageRank track (the original)
Entity/Knowledge Graph track (added 2012, now dominant for trust queries)

Bill Slawski spent 15 years analyzing Google's entity patents at seobythesea.com. Jason Barnard (Kalicube) operationalized the playbook.

What Is an "Entity" to Google?

An entity is a thing Google has identified, classified, and stored in its Knowledge Graph. Entities have:

A unique identifier (machine-readable ID)
A category (Person, Organization, Product, Place)
Relationships to other entities
A confidence score (how sure Google is this entity exists/is what it claims)

Examples of entities:

"Daniel Castellani" (Person, founder of publishd.app)
"publishd.app" (Organization, web design service)
"Next.js" (Software framework)

When you become an entity in the Knowledge Graph, you get:

Knowledge panel in search results
Increased trust scoring across all your content
Eligibility for AI Overview citations
Brand SERP control (your branded queries show what you want them to show)

How to Become an Entity

Step 1: Establish an Entity Home

Your Entity Home is the canonical URL Google associates with you. For a person: typically your About page. For a business: typically your homepage.

Add structured data (JSON-LD) declaring the entity:

{
  "@context": "https://schema.org",
  "@type": "Person",
  "name": "Daniel Castellani",
  "url": "https://publishd.app/about",
  "image": "https://publishd.app/daniel.jpg",
  "jobTitle": "Founder",
  "worksFor": {
    "@type": "Organization",
    "name": "publishd",
    "url": "https://publishd.app"
  },
  "alumniOf": "Amazon",
  "sameAs": [
    "https://twitter.com/publishd",
    "https://linkedin.com/in/danielcastellani",
    "https://github.com/publishd"
  ]
}

The sameAs array is critical. It tells Google these are all the same entity across the web.

Step 2: Cross-Reference Across Trusted Sources

Google verifies entities through multiple corroborating sources. Get yourself listed on:

Wikidata (free, anyone can submit — no Wikipedia notability required)
Crunchbase
LinkedIn (with consistent name + title)
Google Business Profile
Industry directories (Clutch, G2 for SaaS, etc.)

Critical: the information must be consistent across all of them. Same name spelling, same title, same company. Inconsistencies make Google less confident the entity exists.

Step 3: Get Mentioned by Existing Entities

Co-occurrence is the entity track's version of backlinks. When a known entity mentions you in context, Google updates its model of what you are.

Example: If Mike King (a known entity) mentions "Daniel from publishd.app" in a podcast, Google's natural language processing connects these entities. This affects classification and trust — even without a backlink.

How to get mentioned:

Guest on podcasts (your name + role gets indexed)
Get quoted in articles (HARO, Featured.com, Qwoted)
Speak at conferences (transcripts get indexed)
Get cited in industry roundups

N-E-E-A-T: The Updated Trust Framework

Google's published framework is E-E-A-T (Experience, Expertise, Authoritativeness, Trustworthiness). Jason Barnard at Kalicube extended it to N-E-E-A-T by adding Notability.

Notability is whether you're recognized as a known thing in your field. The other components are about quality; notability is about whether you exist as far as the algorithm is concerned.

For a personal brand:

N: Are you mentioned in your industry?
E: Have you actually done what you claim?
E: Can you demonstrate expertise?
A: Do authoritative sources reference you?
T: Do reviews/testimonials confirm trust?

For a business:

N: Is the company recognized in its category?
E: Track record of completed work
E: Demonstrable expertise (case studies, results)
A: Authority signals (press, industry associations)
T: Reviews, certifications, transparency

The framework matters because Google's Search Quality Rater Guidelines (the rubric humans use to train the algorithm) explicitly score pages on these dimensions.

Sources for this chapter:

Chapter 6: Technical SEO That Actually Matters

Most technical SEO advice is bloated. There are 200+ technical factors; 90% don't matter for most sites. Here's the 10% that does.

The Reasonable Surfer Patent (Why Internal Linking Is Underrated)

U.S. Patent 8117209B1, filed 2004, granted 2012. Key shift from original PageRank: links don't pass equal equity. A link's weight is modulated by estimated click probability based on:

Position on the page (body > sidebar > footer)
Visual prominence (color, size, surrounding whitespace)
Anchor text relevance
Surrounding context

What this means concretely:

A link in the middle of an article passes 5-10x more equity than the same link in the footer. Stop putting your important links in nav/footer alone.
Navigation links pass high equity to a few pages and deplete fast. If your nav has 40 links, each gets ~1/40th of homepage authority. If it has 5 links, each gets ~1/5th. Fewer nav items = stronger nav links.
Orphan pages are invisible. A page with zero internal links from your site receives zero PageRank flow. Even if Google indexes it (which is unlikely), it functionally doesn't exist for ranking. Action: Audit your site monthly. Find orphan pages. Either link them in or delete them.
Deep pages (4+ clicks from homepage) get reduced equity flow. On large sites, pages that require 4+ clicks to reach are at risk of shallow crawling. Keep important pages within 3 clicks.
Color-distinctive links are weighted higher than blend-in text links. Make your links visually distinct. The patent literally says click-probability estimates use visual features.

Bill Slawski's full patent breakdown: seobythesea.com/2010/05/googles-reasonable-surfer

Core Web Vitals (The Actual Thresholds)

Google's published thresholds for "Good" Core Web Vitals:

Metric	Good	Needs Improvement	Poor
Largest Contentful Paint (LCP)	< 2.5s	2.5–4.0s	> 4.0s
Interaction to Next Paint (INP)	< 200ms	200–500ms	> 500ms
Cumulative Layout Shift (CLS)	< 0.1	0.1–0.25	> 0.25

INP replaced FID as a Core Web Vital in March 2024. Most older guides still reference FID. INP is stricter — it measures the worst interaction on the page, not the first.

Practical optimizations that move the needle:

Compress images. Use WebP or AVIF format. Most sites have 2-5MB of unoptimized images. Compressing to 200-500KB drops LCP by 1-2 seconds.
Defer non-critical JavaScript. Analytics, chat widgets, marketing pixels — all should load after the main content. Use defer or async attributes.
Preload your hero image. Add <link rel="preload" as="image" href="hero.webp"> in <head>. This often drops LCP by 500ms+.
Eliminate layout shift. Set explicit width/height on images. Reserve space for ads/embeds. CLS issues are almost always caused by content loading after the page renders and pushing things around.
Don't ship a CSR-only React/Vue site for SEO content. Client-side rendering means Google sees a blank page on first crawl. Use Next.js, Astro, Nuxt, SvelteKit, or any framework that supports SSR/SSG.

Schema Markup (The Hidden Lever)

Schema markup is structured data that tells Google explicitly what your content is. It's the difference between Google guessing and Google knowing.

The schema types that matter for indie shippers:

Article — for blog posts. Enables "Article" search features.
Organization — for your business entity. Foundation of entity recognition.
Person — for the founder. Builds personal brand entity.
SoftwareApplication — for apps. Enables app-specific rich results.
Product — for products with reviews/pricing.
FAQ — for Q&A sections. Can earn FAQ rich results in SERPs.
HowTo — for tutorials. Can earn step-by-step rich results.
BreadcrumbList — for navigation hierarchy.

Implementation: JSON-LD format, in <script type="application/ld+json"> tags. Can be in <head> or anywhere on the page.

Test your schema: Use Google's Rich Results Test before publishing.

The Technical SEO Non-Negotiables

A minimum viable technical SEO setup:

☐ HTTPS enabled (free SSL via Let's Encrypt or hosting provider)
☐ XML sitemap auto-generated and submitted to GSC
☐ robots.txt allows JS/CSS crawling (common WordPress mistake)
☐ One canonical URL per page (self-canonical by default)
☐ Schema markup on homepage minimum (Organization + WebSite)
☐ Mobile-responsive (test on actual phone, not just dev tools)
☐ Core Web Vitals "Good" status in GSC
☐ No broken internal links (run Screaming Frog crawl quarterly)
☐ Internal linking audit (no orphan pages)
☐ Hreflang tags if multi-language
☐ 301 redirects (not 302) for permanent moves
☐ 404 page returns proper 404 status (not 200)

That's it. If you nail these 12 items, you're more technically sound than 90% of websites.

Chapter 7: Content Strategy — Passage Indexing + Freshness

Passage Indexing (Section-Level Ranking)

Google announced Passage Indexing in October 2020 (live 2021). The change: individual passages within a page can rank independently for queries the overall page doesn't target.

Mechanism: dense vector retrieval. Each passage gets its own embedding, compared against query embeddings separately from the page-level signal.

What this unlocks: A single comprehensive 5,000-word post can rank for 50+ queries if each H2/H3 section is structured as a self-contained, direct-answer unit.

The Passage Structure That Captures Rankings

## [Question phrased as the user would type it]

[40-60 word direct answer — complete standalone sentence, no "as mentioned above"]

[Supporting explanation, 200-400 words]

[Optional: example, screenshot, data point]

Why this works:

The H2 question matches the search query (relevance signal)
The 40-60 word direct answer is what AI Overviews and Featured Snippets pull
The standalone phrasing means the passage works out of context (no "as mentioned earlier")
The supporting explanation provides depth for engaged readers

Example — Bad:

## Page Speed
Page speed is important for SEO. We talked about this earlier in the technical chapter, but here are some more details. Various factors affect page speed including images, scripts...

Example — Good:

## How fast should a website load for SEO?

A website should load in under 2.5 seconds (LCP threshold for Google's "Good" Core Web Vitals rating). Sites loading in 1-2 seconds rank higher and convert 40% better than sites loading in 4+ seconds.

The 2.5-second target comes from Google's published Core Web Vitals documentation. Specifically...

Freshness — Query Property, Not Site Property

Google's "Query Deserves Freshness" (QDF) signal, introduced 2007 by Amit Singhal: freshness weighting is applied per query, not uniformly. Topics with spiking search volume get temporary recency boosts. Most topics don't.

What doesn't work:

Updating only the published date without changing content (Google maintains its own first-crawl timestamps)
"Refreshed for 2026!" with no actual refresh (manual action risk)
Republishing old content as new

What actually triggers freshness signals:

Adding new statistics with cited sources (re-crawl detects substantive change)
Adding a new section answering a question the original didn't cover
Referencing recent events that didn't exist when original was published

The freshness strategy that works:

Write evergreen content, then make it easy to add a "2026 Update" section at the bottom annually:

[... original 5,000-word evergreen content stays exactly as written ...]

## 2026 Updates

### What changed since publishing:
- [New stat with citation]
- [New section: Why X is now Y]
- [New tool/tactic that emerged]

### Still works in 2026:
- [Confirmation that original points hold]

The rest of the page stays stable. Google sees genuine additive change. Your evergreen rankings don't risk losing position from over-editing.

Content Length: When Long Wins

The data on content length is misread constantly. The correlation isn't "longer content ranks better." The correlation is "comprehensive content ranks better, and comprehensiveness usually requires length."

Practical guide:

Query type	Target length
Definition / quick answer	800-1,500 words
How-to guide	2,000-3,500 words
Comprehensive guide	4,000-8,000 words
Comparison / vs.	1,500-3,000 words
Listicle	2,500-5,000 words
Case study	1,000-2,500 words

Don't pad to hit a word count. Cover the topic completely; the length follows.

Chapter 8: AI Overviews and LLM Citation Strategy

The State of AI Overviews (As of 2026)

Current data:

AI Overviews show on ~25.8% of US searches (Stackmatix, January 2026)
Informational queries trigger AI Overviews ~39.4% of the time
E-commerce queries trigger them ~4% of the time
Zero-click searches went from 56% → 69% between May 2024 and May 2025 (Semrush)
Organic CTR dropped 61% on queries with AI Overviews — from 1.76% to 0.61% (Dataslayer/Ahrefs)
Getting cited inside an AI Overview: +35% organic CTR, +91% paid CTR (Semrush)

The implication: Ranking #1 organically is no longer enough on informational queries. You need to be cited inside the AI Overview.

How AI Overviews Choose Citations

Semrush tracked AI Overview citation overlap with organic top-10 results from May to September 2024. Overlap grew from 32% to 54%.

The primary prerequisite: Rank in the organic top 10 for the query. AI Overviews don't pull from page 5 — they pull from results Google already considers authoritative.

Secondary factors that earn citations:

Direct, citable factual statements with specific numbers
Definitional structure ("X is...", "X works by...")
Original data (surveys, experiments, primary research)
Author bios with credentials (E-E-A-T signal)
Recent publication or update date

The Citation Optimization Framework

For every page, include:

Element 1: A 40-60 word factual lead

"Topical authority is a Google ranking signal that scores 
sites based on coverage completeness within a single topic 
cluster. A new site can rank competitively for a topic if 
it covers all subtopics comprehensively, even with low 
domain authority. Documented case: 0 to 128,000 monthly 
visits in 123 days, no backlinks."

This format is what AI Overviews lift directly.

Element 2: Specific stats with attribution

"AI Overviews now show on 25.8% of US searches as of 
January 2026 (Stackmatix data)."

LLMs cite primary sources. Be the primary source.

Element 3: Definition-style headers

"What is [topic]?"
"How does [topic] work?"
"Why does [topic] matter?"
"When should I use [topic]?"

Element 4: Author bio with credentials

<author>
  <h3>About the author</h3>
  <p>Daniel Castellani has shipped 100+ websites at 
  publishd.app. Former Amazon engineer. Specializes in 
  conversion-focused web design for indie founders.</p>
</author>

Author entities transfer expertise to content.

Element 5: Original data You don't need a research team. Original data can be:

A survey of your customers
An analysis of your own client data
A documented experiment ("I tested X for 30 days...")
A comparison of competitors you actually ran

LLMs heavily favor primary sources over re-reporters.

Tracking LLM Mentions

Tools that track when AI systems cite your site:

Profound — LLM visibility tracker
Otterly.ai — AI search monitoring
AthenaHQ — brand mentions across LLMs

ChatGPT, Perplexity, and Gemini citations correlate strongly with traditional rankings. Same authoritative sources get pulled across all LLMs.

Kevin Indig's Mid-Length Query Insight

From Kevin Indig's Growth Memo: mid-length queries (6-9 words) are the fastest-growing segment of search.

Why: AI assistants and voice search shifted user behavior toward conversational queries. People search "how do I rank a new site without backlinks in 2026" instead of "SEO tips."

Practical application: Optimize for specificity, not head terms. A page targeting "how to rank a new site without backlinks in 2026" will outperform one targeting "SEO tips" for the same effort.

Chapter 9: Programmatic SEO — The Indie Hacker Unfair Advantage

What Is Programmatic SEO?

Programmatic SEO is generating large numbers of pages from a structured dataset, where each page targets a specific long-tail query.

Famous examples:

Zapier — "[App A] + [App B] integration" pages (10,000+ pages, ranks for thousands of queries)
Wise — "Convert [Currency A] to [Currency B]" pages (hundreds of pages, dominates currency conversion queries)
Nomad List — "Best places to live in [Country]" pages (city-by-city pages)
G2 — "Best [software category]" pages (programmatic + reviews)

The pattern: one template + a dataset = N pages targeting N variations of a query.

Why It Works for Indie Shippers

You can't beat Forbes on broad terms. But you can ship 500 highly specific pages overnight that target queries Forbes will never bother with.

The math: If each page gets 10 visitors/month, 500 pages = 5,000 monthly organic visitors. From content you wrote once and templated.

The Programmatic SEO Playbook (Indie Hackers Version)

Step 1: Identify a templatable query pattern

Look for queries with consistent structure:

"[App] alternatives"
"How to [task] in [language/tool]"
"[Currency A] to [Currency B] conversion"
"Best [tool category] for [use case]"
"[X] vs [Y] comparison"

Step 2: Build the structured dataset

Pick a specific pattern. For each variation, you need data points the page will use:

Names and IDs
Descriptive content
Stats or comparison data
Images or screenshots
Unique value-add per page (CRITICAL — see Step 4)

Example: For "[App] alternatives" pages, your dataset has:

App name
App description
Top 5 alternatives (with their data)
Pros/cons of each alternative
A unique data point (price comparison, feature matrix, etc.)

Step 3: Build the page template

The template needs:

Dynamic title tag using the variable
Dynamic H1
Dynamic meta description
Schema markup populated dynamically
Canonical tag pointing to itself
Unique content sections (not just data dump)

Example title: "{app_name} Alternatives: 5 Better Options in 2026"

Step 4: Add unique value per page

This is where most programmatic SEO fails. Google's August 2025 spam update specifically targeted "scaled AI content abuse" — pages that are technically unique but provide no unique value.

Each page needs at least one of:

An original screenshot
A unique data point
A comparison table built specifically for that variation
A user testimonial relevant to that specific case
A custom infographic

If your pages are 95% template + 5% data, you'll hit the spam filter. If they're 60% template + 40% unique-per-page value, you'll rank.

Step 5: Ship 50-500 pages, monitor, prune

Indie hacker consensus: ship a meaningful batch (50+ pages) and monitor:

Which clusters get crawled
Which get indexed
Which start ranking
Which get zero traffic after 90 days

After 90 days, kill the dead clusters. Keep what works. Expand the winners.

When Programmatic SEO Fails

No unique value per page → spam filter
Targeting queries with zero search volume → indexed but no traffic
Bad CSR-only setup → Google can't crawl the dynamic pages
No internal linking between programmatic pages → orphan content
Bad data quality → users bounce, NavBoost penalizes you

Source

Full playbook: Indie Hackers — Complete Guide to Programmatic SEO

Chapter 10: Free Tools as Link Magnets

Why Free Tools Outrank Blog Posts

Free, single-purpose tools earn backlinks 5-10x more than blog posts in indie hacker case studies. Three reasons:

They get cited as "use this to do X" in tutorials, Reddit posts, forum answers
They rank for high-intent queries ("calculator", "checker", "tester", "generator")
AI Overviews recommend them ("here's a tool that does X")

Tools That Have Built Companies

Ahrefs — Free SERP checker, backlink checker (lead-magnet for paid Ahrefs)
Smallpdf — Each tool (compress PDF, merge PDF) ranks separately
Hemingway Editor — Free writing analyzer drives 100K+ monthly visitors
Word Counter — A counter ranks #1 for "word count" with millions of monthly searches

The Indie Tool Playbook

Step 1: Find a tool query with traffic and weak competition

Use Google Search Console + Keyword Surfer:

"[task] calculator"
"[task] generator"
"[task] checker"
"[task] converter"
"[task] tester"

You're looking for queries with 1K-10K monthly searches where the top results are basic, ad-heavy tools.

Step 2: Build something better in a weekend

Don't overthink. The first version should be:

Single-purpose (one job, done well)
Fast (loads in < 1 second)
No signup required
No ads
Mobile-responsive
Beautiful enough to feel premium

Step 3: SEO the tool page

Your tool page should have:

The tool itself above the fold
Clear H1 matching the search query
800-1,500 words explaining what the tool does and how to use it
FAQ section answering common questions
Schema markup (SoftwareApplication or WebApplication)
Internal links to related tools/content

Step 4: Promote to relevant communities

Product Hunt launch
Hacker News (Show HN)
Reddit (relevant subreddit only — don't spam)
Twitter/X with a video demo
Indie Hackers showcase

Each launch generates initial backlinks. Good tools then accumulate organic links over time.

What Tools Should Publishd Build?

For your audience (founders shipping apps):

Website conversion audit tool (input URL, get a clarity score)
SEO audit checker (run automated checks, output report)
Landing page copy analyzer (paste copy, get feedback)
Free site speed checker (PageSpeed Insights wrapper with cleaner UX)
Domain age checker (paste URL, get age + early signals)

Each of these is a separate page that ranks separately, captures a separate query, and has a separate backlink profile.

Chapter 11: "Alternatives to X" Plays for SaaS

The Conversion Data

Comparison-style queries convert at 8.43% on average (Powered by Search) — among the highest-converting page types in SaaS.

Why: when someone searches "Mailchimp alternative," they've already decided they want to leave Mailchimp. They're at the bottom of the funnel.

The Three Comparison Patterns

Pattern 1: "[Competitor] alternative"

Query: "Mailchimp alternative"
Intent: Active churn from competitor
Conversion: Highest
Difficulty: Moderate (lots of incumbents)

Pattern 2: "[Competitor A] vs [Competitor B]"

Query: "Mailchimp vs ConvertKit"
Intent: Active comparison
Conversion: High
Difficulty: High (G2/Capterra dominate)

Pattern 3: "Best [category] for [use case]"

Query: "Best email marketing for ecommerce"
Intent: Research stage
Conversion: Medium
Difficulty: Variable (depends on use case specificity)

The Piggyback Strategy

From Powered by Search: a route-optimization SaaS called Circuit ranked #1 for "Postmates vs OnFleet" — a query about two competitors that don't include Circuit.

How: They built a three-way comparison page titled "Postmates vs OnFleet vs Circuit" — inserting themselves into a high-intent comparison query.

This works because:

The query has commercial intent (someone shopping)
Existing pages are duels (A vs B), not three-ways
Three-way pages are more comprehensive (better passage indexing)
Google's relevance algorithms reward the most useful answer

Action for indie shippers: Find two competitors who are commonly compared. Build a three-way comparison putting yourself between them. Be objective and honest — don't fake the comparison.

Avoiding the "G2 Dump" Mistake

A common error: shipping 50 comparison pages against tools nobody searches for.

Don't ship comparison pages for:

Tools with < 100 monthly searches
Tools that don't actually compete with you
Tools that aren't on the path your customers consider

Instead: Ship 3-5 high-quality comparison pages against actual market leaders. Each should be:

2,000-3,500 words
Honest (acknowledge competitor strengths)
Specific (concrete feature comparisons, not vague claims)
Updated annually
Schema-marked with Comparison data

The Comparison Page Template

# [Your Product] vs [Competitor]: Honest Comparison [Year]

[40-60 word lead summarizing the key difference]

## Quick Verdict
- **Choose [Your Product] if:** [3 specific scenarios]
- **Choose [Competitor] if:** [3 specific scenarios]

## Feature-by-Feature Comparison
[Table comparing 10-15 features honestly]

## Pricing Breakdown
[Detailed pricing comparison with examples]

## Use Case Walkthroughs
[3-5 specific scenarios with which tool wins each]

## Real User Reviews
[Quotes from G2/Capterra/Trustpilot — both products]

## Migration Guide (if applicable)
[How to switch from competitor to you, if relevant]

## FAQ
[Common questions about both products]

## Final Recommendation
[Your honest take, with caveats]

The honesty matters. Pages that obviously favor your product without acknowledging competitor strengths get bounced (badClicks). NavBoost will demote them.

Chapter 12: Link Building Without a Budget

The Skyscraper Technique 2.0

Brian Dean's original Skyscraper Technique generated a 110% traffic lift in 14 days for one of his case studies (backlinko.com/skyscraper-technique).

Original method:

Find content that's getting links
Build something significantly better
Email people who linked to the original

The 2.0 update: Don't just analyze what people linked to. Analyze why they linked. Then your outreach can speak to motivation, not just topic.

Outreach conversion: 2-3x higher when you reference the specific reason someone linked vs. generic "I built a better version" pitches.

HARO Replacement Tools

HARO (Help a Reporter Out) shut down in 2024. Replacements:

These are quote-driven backlinks. Journalists post requests for sources. You respond with expertise. If quoted, you get a backlink from a publication that's typically DR 60+.

The pitch format that works:

Hi [Reporter],

Re: Your request for [topic].

I've shipped 100+ websites at publishd.app. Specifically 
on [the specific question you asked]:

[Direct answer in 100-200 words. Specific, quotable, 
data-driven. No fluff.]

Happy to expand if useful. Bio for byline:

Daniel Castellani, founder of publishd.app, ships 
conversion-focused websites for indie founders.
[Link to your site]

Thanks,
Daniel

Hit rate: 5-15% if you respond fast (within 2 hours of request) and stay specific.

Podcast Guesting (Underpriced)

Podcast hosts are constantly looking for guests. Most podcasts have a podcast page that links to guests. Many have show notes that link to your site.

One decent podcast guest spot = a DR 50+ backlink + distribution to that audience.

How to land guest spots:

Find 20 podcasts in your niche (use Podchaser)
Listen to 2-3 episodes of each (so you can pitch with context)
Pitch with a specific episode angle, not "I'd love to come on your show"
Mention what value you'll bring (data, story, contrarian take)

Pitch template:

Subject: [Specific topic] for [Podcast name]?

Hi [Host],

I've listened to your episodes on [topic 1] and [topic 2]. 
The point about [specific insight from their episode] was 
exactly right.

I have a related angle I think your audience would dig: 
[Your specific topic with a hook].

For example, [one concrete data point or story from your 
proposed topic].

I shipped 100+ websites at publishd.app and have specific 
data on [your angle].

If interested, here's my one-sheet: [link]

— Daniel

Hit rate: 10-30% if you're specific and have a real hook.

Free Tools as Link Magnets

(See Chapter 10.) Free tools earn 5-10x more backlinks than blog posts.

Industry Directories and Citation Sites

For local businesses: get listed on Google Business Profile, Yelp, Bing Places, Apple Maps.

For SaaS: G2, Capterra, Product Hunt, AlternativeTo, Slant.

For services: Clutch, GoodFirms, DesignRush, UpCity.

Citation consistency matters: same name, address, phone (or "NAP") across all listings. Tools like Yext or BrightLocal automate this — or do it manually for free.

What Doesn't Work

Buying backlinks — Google detects this and penalizes you
Link exchanges — "I'll link to you if you link to me" detected as patterns
PBNs (Private Blog Networks) — directly violates spam policies
Comment spam — links from comments are typically nofollow anyway
Low-quality directory submissions — 50 listings on random directories add zero value
"Guest posts" on content farms — Google deprecated low-quality guest post networks

What Works (Summary)

Free tools that earn organic citations
Original research that gets quoted
Podcast guesting with real content
HARO/Qwoted/Featured responses
Skyscraper outreach with motivation analysis
Industry directory listings (real ones)
Building relationships with 5-10 site owners in your niche

The pattern: real value → real links. Anything else is either fragile or risky.

Chapter 13: The 90-Day Ranking Plan

The Honest Timeline

Rankings don't happen overnight. Realistic expectations:

Phase	Timeline	What's Happening
Foundation	Months 1-3	Site indexed, content published, no rankings yet
Early Traction	Months 3-6	Long-tail keywords start ranking, 50-200 visitors/month
Momentum	Months 6-12	Mid-tail keywords climbing, 200-1,000 visitors/month
Authority	Months 12-24	Competitive keywords ranking, 1,000-10,000 visitors/month
Compounding	Months 24+	Topical authority pays off, multi-keyword rankings

The 90-day plan below gets you through Foundation and into Early Traction.

Days 1-7: Audit and Setup

Day 1: Run the audit

Sign up for Google Search Console (free)
Sign up for Google Analytics 4 (free)
Run PageSpeed Insights on your homepage
Run Mobile-Friendly Test
Run Rich Results Test
Document current state: traffic, rankings, technical issues

Day 2-3: Fix critical technical issues

HTTPS enabled? (If not, fix immediately)
Mobile responsive? (Test on actual phone)
Core Web Vitals score? (If poor, this is your priority)
Submit XML sitemap to Search Console
Verify robots.txt allows JS/CSS crawling

Day 4-5: Set up entity foundation

Add Organization schema to homepage
Add Person schema to about page
Claim Google Business Profile (if applicable)
Update LinkedIn, Twitter, GitHub with consistent info
Submit to Wikidata (free)

Day 6-7: Plan the topical map

Pick your topic cluster (narrow!)
Brainstorm 50-80 article ideas
Identify head term, mid-tail, long-tail, adjacent, use-case
Pick the pillar page topic
Document the topical map in a spreadsheet

Days 8-30: Build the Foundation

Week 2: Write the pillar page

4,000-8,000 words
Comprehensive coverage of the topic
Strong passage-indexing structure (question H2s, direct answers)
Schema markup (Article)
10+ internal links to future cluster posts (placeholder URLs okay)

Weeks 3-4: Write 4 cluster posts

2,000-3,500 words each
Each targets a specific subtopic from the topical map
Each links UP to the pillar
Each links SIDEWAYS to 3-5 other cluster posts
Schema markup on each

By end of Day 30:

Pillar page published
4 cluster posts published
All internally linked
All submitted to Google Search Console for indexing
Title tags optimized for CTR
Meta descriptions are compelling ad copy

Days 31-60: Expand and Promote

Week 5-6: Continue cluster expansion

Write 4 more cluster posts (8 total now)
Update pillar page with links to new posts
Add author bios to all posts (entity signal)

Week 7: Outreach begins

Identify 20 potential link partners (sites in your niche)
Send 5 HARO/Qwoted responses (specific, fast, helpful)
Pitch 3 podcast hosts
Submit to 5 industry directories
Share content on Twitter/LinkedIn (drives chromeInTotal signal)

Week 8: Monitor and adjust

Check Search Console for indexing status
Look at "striking distance" keywords (positions 11-20)
Refresh title tags on any post showing impressions but low CTR
Add more internal links if any post is orphaned

Days 61-90: Iterate and Scale

Week 9-10: Continue cluster expansion

4 more cluster posts (12 total)
Expand pillar with new sections
Add a "2026 Update" section to the oldest post (test freshness signal)

Week 11: Build a free tool

Identify a tool query in your niche
Build a single-purpose tool over a weekend
Ship it with full SEO optimization (schema, content, internal links)
Promote on Product Hunt, Hacker News, Reddit

Week 12: Audit and double down

Pull Search Console data: which posts are gaining impressions?
Pull GA4 data: which posts have lowest bounce rate?
Identify your 3 best-performing posts
Expand each: add depth, add internal links, add schema
Plan next 90 days based on what's working

What to Measure

After 90 days:

Metric	Target	What it means
Indexed pages	12+	Foundation built
Search Console impressions	1,000+/month	Google sees your content
Search Console clicks	50+/month	Some queries ranking
Average position	25-50	Long-tail traction
Striking distance keywords (pos 11-20)	5+	Ready for next push

If you hit these targets, months 4-6 will see meaningful traffic. If you fall short, the issue is usually:

Topic too broad (lower topical authority)
Content too thin (under 1,500 words on competitive queries)
Technical issues blocking crawl
No internal linking structure

What Makes 90 Days Fail

The single biggest reason 90-day plans fail: inconsistency.

You need 12 cluster posts published in 90 days. That's one per week. Most people start strong (4 posts in week 1) and burn out by week 4.

The fix: schedule it like a job. 1 post per week, on the same day, for 12 weeks. No more, no less. Consistent output beats heroic sprints.

Chapter 14: When to Stop DIY and Hire Help

When DIY SEO Works

DIY is the right call if:

Your niche has low competition (long-tail keywords with < 1,000 monthly searches)
You have 5-10 hours/week for content
You enjoy the work or want to learn
You have technical skills (or your site doesn't have technical issues)
Your business doesn't depend on ranking in the next 90 days

When DIY Won't Work

DIY hits a ceiling when:

Issue 1: You're competing in a high-volume niche

Top results have DR 70+ domains
Top results have 8,000+ word content with original research
Top results have been live for 5+ years

You can still niche down (smaller topic cluster), but you'll never rank for the head term without significant authority.

Issue 2: Technical issues require real expertise

Site speed below 50 PageSpeed score
JavaScript-heavy single-page app
Migration disasters (lost rankings after redesign)
Schema markup needed but you don't write code

Issue 3: You don't have time

Need 5-10 hours/week for content
Need 2-4 hours/week for technical maintenance
Need 1-2 hours/week for monitoring and adjustment

If your business demands your full attention, SEO becomes the thing that never gets done.

Issue 4: Site architecture is fundamentally broken

Built on Wix or Squarespace (real SEO ceiling)
No CMS, hard to publish content
Site structure makes internal linking impossible
Mobile experience is broken

In this case, you need a rebuild before any SEO work helps.

What to Hire (And What to Pay)

Need	Hire	Cost Range
Strategic audit	SEO consultant (one-time)	$1,000-$5,000
Ongoing strategy + tracking	SEO consultant (monthly)	$2,000-$5,000/month
Content production	Content writer (per post or monthly)	$200-$1,500/post
Technical fixes	Technical SEO specialist	$3,000-$10,000/month
New site build	Web designer/developer	$999-$50,000
Comprehensive SEO + design	Specialized agency	$5,000-$50,000+/month

When Publishd Makes Sense

I built Publishd specifically for the gap I noticed running 100+ projects:

You hire Publishd if:

Your existing site has serious SEO/conversion problems
You need a custom-built site that ranks (not a template)
You want fast turnaround (4-6 weeks, not 6 months)
You want transparent pricing (not "starts at $5,000, ends at $50,000")
You want to own the code (no vendor lock-in)
You want direct founder access (no account managers, no relays)

Pricing:

Starter: $499 — Single landing page, conversion-focused, SEO-ready
Pro: $999 — Multi-page site with blog/CMS, full schema setup, technical SEO baked in
Custom: $1,999+ — Larger scope (e-commerce, programmatic SEO setup, multi-section sites)

What's included on every tier:

Custom design (not template)
Schema markup setup (Organization, Person, Article)
Technical SEO foundations (Core Web Vitals "Good")
Mobile-responsive
6-hour response time during build
Full code ownership at handoff
Documentation

You're not hiring me to "do SEO." You're hiring me to ship a site that ranks. Different thing.

What I Don't Do

Honest disclosure of where I'm not the right fit:

Pure content production at scale — I build sites; I don't run a 50-post-per-month content factory.
Enterprise SEO — I don't have the team for $50K/month engagements.
Black-hat or grey-hat SEO — Won't buy links, won't churn AI content, won't take shortcuts that risk your domain.
Paid ads — Not my specialty. Hire someone who lives in Google Ads daily.

If you need any of the above, I'll tell you and recommend someone better.

Appendix: The 24-Point Audit Checklist

Print this. Run through it monthly.

Technical SEO (10 points)

☐ HTTPS enabled (🔒 in URL)
☐ Mobile-responsive (test on actual phone)
☐ Core Web Vitals "Good" status in Search Console
☐ XML sitemap submitted to GSC
☐ robots.txt not blocking JS/CSS
☐ Schema markup on homepage (Organization minimum)
☐ Schema markup on key page types (Article, Product, etc.)
☐ Canonical tags set (self-canonical by default)
☐ No broken internal links (run Screaming Frog quarterly)
☐ No orphan pages (every page has internal links pointing to it)

Content SEO (8 points)

☐ Topical map documented (50+ article ideas)
☐ Pillar page exists (4,000+ words, comprehensive)
☐ At least 8 cluster posts published
☐ All cluster posts link to pillar; pillar links to all clusters
☐ Title tags optimized for CTR (not just keywords)
☐ Meta descriptions are compelling ad copy (160 char)
☐ Passage-indexing structure (question H2s, 40-60 word direct answers)
☐ Author bios on every post (entity signal)

Authority + Behavioral (6 points)

☐ Google Business Profile claimed (if applicable)
☐ Listed on Wikidata
☐ Consistent NAP across 5+ directory listings
☐ At least 5 backlinks from real sites (not paid/spam)
☐ Active brand search (people searching your name)
☐ Direct traffic source (newsletter, social, etc.)

Tools Referenced (All Free or Freemium)

Google's own tools:

Google Search Console — free, mandatory
Google Analytics 4 — free
PageSpeed Insights — free
Rich Results Test — free
Mobile-Friendly Test — free

Keyword research:

Google Keyword Planner — free with Google Ads account
Keyword Surfer — free Chrome extension
AnswerThePublic — free tier
Bing Webmaster Tools — free, has keyword data

LLM citation tracking:

Profound — paid
Otterly.ai — paid
AthenaHQ — paid

Backlink/competitor analysis:

Ahrefs Free Tools — free tier
Ubersuggest — free tier

Outreach:

Qwoted — free for sources
Featured.com — free for sources
Help A B2B Writer — free
Podchaser — free podcast database

Sources (Every Citation in This Guide)

Google leak + algorithm research:

Behavioral signals:

Search Engine Journal — CTR Ranking Factors

Topical authority:

Holistic SEO — Topical Authority

Entity SEO:

AI Overviews data:

Google's own documentation:

Practitioner sources:

Indie / SaaS-specific:

Lead magnet + Trustpilot data:

Closing Note

If you've made it this far, you have more applied SEO knowledge than 95% of business owners. You also have more than most "SEO experts" who haven't read the leak documentation.

The honest truth: ranking is a long game. The principles in this guide are mechanical — they work because they reflect how Google's algorithm actually operates. But mechanical doesn't mean instant. You'll see meaningful results in 3-6 months and compounding results over 1-2 years.

If you do the work, you'll rank. Most people won't do the work. That's why ranking is still possible.

If you want me to build the site that runs all of this for you, either start a kickoff or jump straight to the website design page. Flat fee, you own the code, no lock-in.

If you want to do it yourself, this guide is everything I'd tell a friend over coffee.

Either way — go ship something.

— Daniel Castellani
publishd.app

Last updated: 2026. This guide is updated annually with new algorithm research, case studies, and tactics. If you found it useful, leave a Trustpilot review — it helps other founders find this guide.

[ companion ]

Want the skim version and the next step?

Use the printable checklist if you just want the execution list, or book a kickoff if you want me to run this against your site and build the thing properly.

Want it done?

If you'd rather skip the 90-day DIY phase, I can build the site, fix the structure, and ship the SEO foundations with it.

Start a kickoff Review on Trustpilot

Why This Guide Exists

Table of Contents

Chapter 1: The 2024 Google Leak — What Changed Everything

What Happened

Finding #1: NavBoost Exists (And Uses 13 Months of Click Data)

Finding #2: siteAuthority Is Real

Finding #3: hostAge Confirms the Sandbox

Finding #4: chromeInTotal Tracks Browser Data

What Most Guides Get Wrong (And You Now Know)

Chapter 2: How Google Actually Ranks (The Real Architecture)

The Stack (Simplified)

The Three Pillars That Actually Matter

Why New Sites Struggle

Chapter 3: Behavioral Signals — Confirmed Under Oath

The Court Testimony

Rand Fishkin's CTR Experiment

What This Means for Your SEO

Chapter 4: Topical Authority — The Indie Shipper's Cheat Code

The Case Study

How Topical Authority Beats Domain Authority

The Topical Map

Why This Works for Indie Shippers Specifically

Chapter 5: Entity SEO and N-E-E-A-T

The Knowledge Graph Track

What Is an "Entity" to Google?

How to Become an Entity

N-E-E-A-T: The Updated Trust Framework

Chapter 6: Technical SEO That Actually Matters

The Reasonable Surfer Patent (Why Internal Linking Is Underrated)

Core Web Vitals (The Actual Thresholds)

Schema Markup (The Hidden Lever)

The Technical SEO Non-Negotiables

Chapter 7: Content Strategy — Passage Indexing + Freshness

Passage Indexing (Section-Level Ranking)

The Passage Structure That Captures Rankings

People Also Ask (PAA) Optimization

Freshness — Query Property, Not Site Property

Content Length: When Long Wins

Chapter 8: AI Overviews and LLM Citation Strategy

The State of AI Overviews (As of 2026)

How AI Overviews Choose Citations

The Citation Optimization Framework

Tracking LLM Mentions

Kevin Indig's Mid-Length Query Insight

Chapter 9: Programmatic SEO — The Indie Hacker Unfair Advantage

What Is Programmatic SEO?

Why It Works for Indie Shippers

The Programmatic SEO Playbook (Indie Hackers Version)

When Programmatic SEO Fails

Source

Chapter 10: Free Tools as Link Magnets

Why Free Tools Outrank Blog Posts

Tools That Have Built Companies

The Indie Tool Playbook

What Tools Should Publishd Build?

Chapter 11: "Alternatives to X" Plays for SaaS

The Conversion Data

The Three Comparison Patterns

The Piggyback Strategy

Avoiding the "G2 Dump" Mistake

The Comparison Page Template

Chapter 12: Link Building Without a Budget

The Skyscraper Technique 2.0

HARO Replacement Tools

Podcast Guesting (Underpriced)

Free Tools as Link Magnets

Industry Directories and Citation Sites

What Doesn't Work

What Works (Summary)

Chapter 13: The 90-Day Ranking Plan

The Honest Timeline

Days 1-7: Audit and Setup

Days 8-30: Build the Foundation

Days 31-60: Expand and Promote

Days 61-90: Iterate and Scale

What to Measure

What Makes 90 Days Fail

Chapter 14: When to Stop DIY and Hire Help

When DIY SEO Works

When DIY Won't Work