Articles
The tactical archive
New posts on AI search visibility, schema strategy, and indie-SaaS SEO. Some are free; some are part of the course archive.
An agent audited 42 Cloudflare zones. Here is what it changed, and what it refused to touch.
A redacted build log: one read-only audit across a 42-domain Cloudflare account, 68 approved setting changes, and a before/after re-audit. Includes the checks, the approval rules and what still needs a human.
LIPAI WANG ·
Auditing a 40-site Cloudflare portfolio with an agent: what we found and what we fixed first
A redacted build log: 42 domains on one Cloudflare account, 37 live sites, 36 RankPropel audits with a single organization key, an independent crawl, and four pull requests for the fixes that survived a second check.
LIPAI WANG ·
Cloudflare has no spending cap. Here are the safeguards we put around an agent API on Workers.
A build log of the cost and abuse controls on RankPropel's agent API: escalating budget alerts, usage alerts, a WAF rate limiting rule tested with a real burst, per-key limits in the Worker, early rejection and monthly caps. With measured CPU costs and links to Cloudflare's current pricing.
LIPAI WANG ·
Core Web Vitals and rankings: what Google says they do, what they don't, and how to read your own data
Google says Core Web Vitals are used by its ranking systems, that good scores don't guarantee top rankings, and that relevance comes first. Here are the current metrics and thresholds, the difference between field and lab data, and a work order for SaaS and indie sites.
LIPAI WANG ·
From site audit to pull request: six suspected issues, three real ones
A redacted build log of RankPropel's SEO pipeline on a multilingual professional-services website: how an evidence gate withdrew half the findings, and why the fixes went to the owner as a pull request instead of a deploy.
LIPAI WANG ·
Internal links: what Google documents, and what the click-weight theory does not prove
Internal linking advice now comes with specific claims: that link position and link clicks carry ranking weight. Here is what Google's documentation supports, what it does not, how to read 'Discovered – currently not indexed', and a test design for small sites.
LIPAI WANG ·
Crawled or Discovered – currently not indexed? What each Search Console status means, by Google's own definitions
A status-by-status guide to Search Console's Page indexing report: what Google says each reason means, which ones are working as intended, which need a fix, and which popular explanations go beyond the documentation. Includes what sitemaps, recrawl requests and the Indexing API can and cannot do.
LIPAI WANG ·
Mining Search Console queries for pages that already rank: what the data shows and what it hides
A method for finding the searches a page already appears for, deciding which ones deserve a section or a new page, and measuring the result. Each step is checked against Google's Search Console documentation, including the limits that make keyword lists incomplete.
LIPAI WANG ·
Show the source: how we cite Google's documentation with highlighted evidence
Why RankPropel puts highlighted, dated screenshots of Google's own documentation next to important claims, how we capture them, the licence that allows it, and what a screenshot cannot prove.
RankPropel ·
Subdomain or subfolder for your SaaS blog, docs or app: what Google documents and what it leaves open
Practitioners say subdomains split authority and subfolders inherit it. Google says to do what makes sense for your business. Here is what Google does document about hostnames (crawl budget, robots.txt, site names, Search Console, moves), and a decision table for blogs, docs, help centres and apps.
LIPAI WANG ·
Updating old content without faking freshness: dates, dateModified and sitemap lastmod, by Google's rules
When to change a page's visible date, what Google does with dateModified and sitemap lastmod, why date-only edits backfire, and which queries actually reward fresh content. A practical policy for SaaS blogs and docs, checked against Google's documentation.
LIPAI WANG ·
What RankPropel can and cannot see when your agent uses it
A plain-language guide to the RankPropel MCP service's data boundary: your agent sends public page HTML, RankPropel returns advice, and your code, CMS and hosting stay on your side.
LIPAI WANG ·
Google SEO and AI Search: what to check in October 2026
A dated review of Google's AI-content guidance, Search Console reports, eligibility controls and September spam update, with practical audit checks.
RankPropel ·
Build log: adding a $29 plan for indie developers without breaking billing
Why RankPropel added a cheaper plan instead of a discount code, the billing check that would have locked out discounted customers, and the database change it needed.
LIPAI WANG ·
Answer-first page shape: what is official, what is folklore
Word counts, keyword density, H1 rules, meta description length, FAQ blocks, TL;DR boxes and "chunking for LLMs", each graded against Google's and Bing's own documentation. Plus a page-shape template we use, labeled as our method, and a way to test it.
LIPAI WANG ·
Bottom-of-funnel pages for SaaS: use-case, alternative, comparison and integration pages that stay out of doorway territory
Which buyer-stage pages a SaaS site should build, what each one needs on it, where Google's doorway and scaled content policies draw the line, how to write a comparison page you can defend, and how to measure the set in Search Console and your own conversion data.
LIPAI WANG ·
DA, DR and Trust Flow are not Google metrics: what each one measures, when it helps and how to measure links instead
How Moz, Ahrefs and Majestic say they compute Domain Authority, Domain Rating, Trust Flow and Citation Flow, what Google has said about third-party authority scores, where the scores are useful for comparing link profiles, why they make poor KPIs, and a measurement plan built on Search Console and your own conversion data.
LIPAI WANG ·
Do clicks rank pages? What the DOJ trial and the 2024 leak do and do not show
An evidence review of click data in Google ranking: what Google says, what the US v. Google court found from sworn testimony about Navboost and Glue, what the 2024 API documentation leak only names, why none of it supports CTR manipulation, and what a site owner can legitimately do about clicks.
LIPAI WANG ·
How to vet an SEO or GEO agency: questions, evidence, red flags and contract hygiene for SaaS buyers
A buyer's checklist for SaaS and B2B teams hiring SEO or AI search help: what to ask, what evidence to demand (baselines, denominators, dated reports), the red flags Google and the FTC already describe, how to read a GEO pitch, and what belongs in the scope and contract.
LIPAI WANG ·
Local SEO myths, graded by evidence: what Google, Bing and Apple document, and what is folklore
Fifteen common local SEO claims, from keywords in the business name to geotagged photos, review velocity, city pages and AI assistants, each graded against Google Business Profile Help, Google's spam policies, Bing, Apple and the FTC reviews rule. Plus a way to test the rest without fooling yourself.
LIPAI WANG ·
Migrations without traffic loss: a site-move runbook built on Google's and Bing's documentation
How to plan a domain, URL or hosting migration: the pre-move inventory and baseline, a redirect map, staging checks, a launch-day checklist, what to monitor afterward, when to roll back, and what Google says normal fluctuation looks like.
LIPAI WANG ·
Recovering from a Google demotion: what is documented and what is folklore
How to tell a manual action, an algorithmic spam demotion, a core update shift and a technical loss apart, what Google says each one takes to recover, why moving to a new domain to escape a demotion can be treated as circumvention, and which popular recovery claims have no source behind them.
LIPAI WANG ·
Reddit's Google and AI visibility: what the data shows and what it doesn't
A dated, sourced look at how visible Reddit is in Google and in AI answers, why the published share figures disagree, what the Google and OpenAI data deals do and don't say, where Reddit's rules and the FTC draw the line, and how a SaaS team can take part honestly and measure it.
LIPAI WANG ·
The entity home page: an About-page checklist for SaaS
How to build one page that states who your company is, keep those facts the same everywhere you control, use Organization markup and sameAs the way Google documents them, and handle knowledge panels and Wikidata without overselling what any of it does, including for AI answers.
LIPAI WANG ·
From zero to indexed: a launch-day checklist for a content site (with the bug we almost shipped)
A redacted launch record by LIPAI WANG: provisioning, one canonical host, HTTPS, Search Console DNS verification, sitemaps, Bing import and IndexNow, plus the empty-index bug a live link count caught before any crawler did.
LIPAI WANG ·
What AI search engines cite when B2B buyers ask for a shortlist
A one-day panel by LIPAI WANG of buyer questions in an industrial B2B software category across Google, Google AI Mode, ChatGPT, Bing Copilot and Claude: who supplies the answers, how rivals get described, and what vendors can fix.
LIPAI WANG ·
Category audit: how eight B2B vendors' websites hold up for AI search
A public-data audit by LIPAI WANG of eight competing B2B software vendors: crawler access, rendering, structured data, buyer evidence and comparison pages, checked against Google's current guidance.
LIPAI WANG ·
Case study: preparing a long-form guide for Google's AI features before launch
A redacted case study by LIPAI WANG: applying Google's October 2026 guidance on AI content, images and human review to a long-form guide before it went live.
LIPAI WANG ·
Technical SEO case study: discovery files and homepage redirects
A redacted case study by LIPAI WANG: diagnosing and repairing discovery-file responses and a broken homepage redirect.
LIPAI WANG ·
Service pricing case study: making scope and next steps clear
A redacted implementation case by LIPAI WANG: explicit currency, clear deliverables, payment milestones and an inquiry path.
LIPAI WANG ·
The RankPropel method: turn search improvements into reusable evidence
A practical cycle for improving SaaS discovery, measuring outcomes, and turning tested lessons into better products.
LIPAI WANG ·
Directory sites in 2026: the business model, who actually profits, and a 5-question vetting framework
Indie devs are being pitched template-and-course bundles for AI directory sites by people whose revenue comes from selling the template, not from operating one. Here's the actual business model, where the money is at each layer, and a 5-question diagnostic for separating real opportunities from shovels-in-a-gold-rush plays.
LIPAI WANG ·
The protection racket model: how indie devs pick markets they can't win in 2026
Selling both the disease and the cure — AI detector AND AI humanizer, plagiarism scanner AND paraphraser, spam filter AND email warmup — is the protection racket model. In the 2026 AI-search era it's gone from clever indie-SaaS hack to structural liability: AI answers tend to treat you with caution or leave you out, trust signals are hard to build, and a single investigative thread is a multi-year overhang. Here's the five-test framework for picking markets GEO will actually endorse, plus how to choose the right side once you spot the pattern.
LIPAI WANG ·
The domain pre-purchase checklist for indie devs
Before you register a domain, run this free six-point audit on its history. Spammy backlinks, gambling redirects, AI-content-farm signals, and PBN footprints all transfer to your new site. Copy-paste checklist, 30-minute run, free tools only.
LIPAI WANG ·
The Scaled Content Abuse trap: how batch AI content sinks indie sites in 2026
Two Google spam policies, scaled content abuse and the site reputation policy, plus core ranking's view of helpfulness, are what batch AI content workflows run into. Here's the failure pattern, what reported cases look like, and the 7-rule checklist that keeps your transcript-to-article workflow honest. Plus a free single-page self-review tool.
LIPAI WANG ·
Case study: 239 clicks from 61,500 impressions — anatomy of a page-2 CTR collapse
A real 28-day Search Console panel — 0.4% CTR at average position 11.5. Setup, data, diagnosis, and the action plan. Plus the generalizable lesson: most 'ranking problems' indie devs bring me are click-through problems in disguise.
LIPAI WANG ·
Cloudflare's 'Manage robots.txt' will silently block GPTBot, ClaudeBot, and Perplexity if you let it
I audited 25 of my own Cloudflare-hosted indie sites in May 2026 and found 6 of them invisibly blocking every major AI crawler. Here's what's happening, how to detect it in 30 seconds, and how to fix it.
LIPAI WANG ·
IndexNow on Cloudflare: the 5-minute free win indie devs skip
IndexNow is a single POST that tells Bing, Yandex, Naver, Seznam, and a growing list of engines that your URLs changed. Cloudflare has a one-click integration. Almost nobody enables it. Here's the setup and what it's worth.
LIPAI WANG ·
Welcome to RankPropel
What this site is, what's free, what's paid, and why we're publishing this in 2026.
LIPAI WANG ·
When /robots.txt returns your homepage — the Cloudflare Pages SPA-fallback trap
If you deployed a static site to Cloudflare Pages without an explicit robots.txt or sitemap.xml file, you're serving your homepage HTML at those URLs. Bots see junk. Here's the trap and the one-minute fix.
LIPAI WANG ·
Why Perplexity and ChatGPT rarely cite the same sites
ChatGPT and Perplexity often cite different domains for the same query. Low overlap does not mean the engines are unrelated: both retrieve from the web, then select differently. Here is what is documented, what is only observed, and how to cover both without duplicating work.
LIPAI WANG ·