Guides

Every answer here is backed by numbers from one product taken from zero to 100K+ users on search alone, with nothing spent on advertising.

Clicks and positions

• Average CTR by Google position, measured on real data: What click-through rate to expect at each position, measured on 1.2 million impressions from one site, and why the published averages will mislead you.
• What is a good click-through rate in SEO?: Why a single "good CTR" number is meaningless, what actually determines yours, and how to tell a snippet problem from a ranking problem.
• Why is my CTR dropping in Search Console?: Falling click-through rate is usually arithmetic, not decline. How to tell the difference before you rewrite titles that were never the problem.
• Lots of impressions but no clicks: what it actually means: Impressions without clicks is a specific, diagnosable condition with two very different causes and two very different fixes.
• Getting from position 8 to position 3 is a different game: The work that takes a page from invisible to position 8 is not the work that takes it from 8 to 3, and confusing the two costs quarters.

Getting found at all

• Website not showing up on Google: work through this in order: The checks that actually find the cause, cheapest first, before you rewrite anything or buy anything.
• How long does Google take to index a new page?: Indexing, ranking and settling are three separate clocks. Confusing them is why people give up in week three.
• Bing is not indexing your site, and it matters more than you think: A second search engine can vanish without warning and recover on request. What that looks like, and what it costs to ignore.
• Pages are indexed but get no traffic: Indexed and invisible are the same thing from a user’s side. How to tell which of your pages are dead and which are merely unpersuasive.
• Traffic dropped suddenly: how to diagnose it without panicking: A gradual decline and a cliff have different causes and different remedies. Telling them apart is the first and cheapest step.

Choosing what to target

• How to find keywords without paying for a tool: A method that costs an afternoon and nothing else, using the sources that show you what people actually type.
• Should you target high-volume keywords?: The biggest query in a category is usually the worst investment available, and here is the measured reason why.
• Are long-tail keywords worth it? Yes, with one caveat: Long-tail queries rank more easily and convert differently from how everyone claims. The measured version.
• Keyword research for a brand new website: What to do when you have no data, no authority and no budget, and every guide assumes you have all three.
• How to tell if a keyword is too competitive for you: Difficulty scores are a proxy. Reading the actual results page is the measurement, and it takes two minutes.

Starting from nothing

• How long until a new website ranks?: A month-by-month account from one site that went from zero to six figures of users, including the months that looked like failure.
• SEO for a brand new website, in the order that matters: What to do first when you have no authority, no traffic and no budget, and why most checklists are in the wrong order.
• The first 90 days of SEO, and what the numbers will look like: A realistic account of the first quarter, so you can tell normal from broken while it is happening.
• No traffic after three months: what to check before giving up: Three months of nothing has several causes and only one of them means the plan is wrong.

Authority and links

• Do you need backlinks to rank? Honestly, it depends where: On-page work has a measurable ceiling. Knowing where it sits tells you exactly when links start to matter.
• How to get backlinks when you hate outreach: Cold outreach converts terribly. What actually earned links here, measured, including the campaign that produced almost nothing.
• Domain Rating explained, and what it does not tell you: A third-party score modelling your link graph. Useful as a rough sort, dangerous as a goal.
• Is Domain Rating worth chasing?: You can raise it in a weekend. Here is exactly what that bought, measured, and what it did not.
• Are directory submissions still worth it in 2026?: Yes, once, for a specific reason that is not the one usually given.
• Why nobody links to your site: Usually because there is nothing on it that helps the person who would be doing the linking.

Structured data

• Is schema markup worth it?: It is cheap, it is not magic, and one specific mistake with it can cost you a manual penalty.
• How to add schema markup, and how to check it stayed added: Generating the block is the easy half. Knowing it is still on all 200 pages six months later is the half nobody does.
• Schema is valid but no rich result is showing: Valid, eligible and displayed are three different states, and most people conflate the first and the last.
• BreadcrumbList schema: what it is for and how it breaks: The one structured-data type whose entire value is in its contents, and the three ways it silently goes wrong.
• Does FAQ schema still work?: Its rich result was largely withdrawn. It is still worth adding, for a different reason than the one you were sold.

Pages and content

• Should you start a blog for SEO? The measured answer: One site, both approaches, same period, same author. The gap is larger than anyone admits.
• Why your blog gets no traffic: Usually because it is competing for queries where a page that does the job wins, and no amount of writing changes that.
• How many pages does a website need for SEO?: More than you think, and the reason is about entry points rather than volume.
• What is thin content, and how thin is too thin?: Word count is not the test. Whether the page answers the query it targets is the test.
• Programmatic SEO: what it is, and the distribution nobody prints: Generating pages at scale works for a reason that is not the one usually given. Here is the full distribution from one site.
• Programmatic SEO mistakes that get whole sites handled as a block: Four failure modes, each survivable at sixty pages and fatal at ten thousand.

Reading your own numbers

• How to read Search Console without fooling yourself: Four habits that turn the default dashboard from a source of false alarms into a diagnostic.
• Search Console vs analytics: they answer different questions: One tells you how people arrive, the other what they do. Using either alone leaves you half blind.
• What is a good conversion rate?: It depends entirely on what you count as a conversion, which is why the published benchmarks are useless.
• How to tell if your SEO is actually working: Four signals that lead the traffic number, so you know months before the chart does.
• What average position actually means, and why yours got worse: A rising average position number is frequently a sign of growth, which makes it the most misread metric in the dashboard.

AI and answer engines

• Will AI kill SEO?: It changes what a click is worth and it does not change what makes a page worth citing.
• How to get named in AI answers: There is no submission form. There is a mechanism, and it is slower and more boring than the advice suggests.
• What AI Overviews do to your traffic, and how to see it: The effect shows up as falling CTR at unchanged positions, which looks exactly like a snippet problem.
• What is llms.txt, and should you add one?: A proposed convention with low adoption, near-zero cost, and one honest reason to bother.

Growth without a budget

• How to get 100,000 users without spending on ads: One product, 224 days, zero advertising spend, and the full account of what actually did the work.
• SEO vs paid ads when you have no money: They are not alternatives, they are different instruments with different failure modes and different timelines.
• How to launch a product when nobody knows you: A launch is upstream of traffic, not downstream of it, which is why waiting until you are ready is the mistake.
• Free tools as a marketing strategy, measured: Why a working tool outperforms content about the same subject, on the same domain, in the same period.
• What SEO actually costs when you do it yourself: An honest accounting of one site: the hours, the tools, the money, and the mistakes that were expensive.

Technical foundations

• Crawl budget, and whether yours is actually the problem: What crawl budget is, how to measure yours from a log file, and why most sites worrying about it have a linking problem instead.
• The robots.txt mistakes that quietly cost you indexing: What robots.txt does and does not do, the five mistakes that actually deindex pages, and how to check yours in a minute.
• What a sitemap is actually for, and what it will not fix: Sitemaps aid discovery, not ranking. What to include, what to leave out, and why an accurate small sitemap beats a complete large one.
• Canonical tags: what they do, and the four ways they backfire: A canonical is a hint about duplicates, not a command. The mistakes that deindex pages, and how to check what yours actually say.
• Redirect chains: the cost, and when it is worth fixing: What a redirect chain costs in crawl, speed and signal, how to find yours, and the honest size of the problem.
• JavaScript and search: what gets indexed, and what gets deferred: Google renders JavaScript on a second pass. What that means for indexing speed, for other crawlers, and how to see what lands in your HTML.

Migrations and moves

• Site migration without losing your traffic: The order of operations for a migration, the checks that matter before and after, and the realistic recovery curve.
• Changing domain without losing your rankings: What actually transfers when you move domains, the steps in order, and how long the new domain takes to inherit the old one.
• Moving to HTTPS without splitting your site in two: HTTPS is a small ranking factor and a large duplication risk. The four things to get right so the two versions consolidate.
• URL structure: what matters, and what is worth changing: Which parts of a URL affect anything, why changing structure is rarely worth the migration cost, and what to get right the first time.

SaaS and product

• SEO for SaaS, from a product that did it with no budget: Why the standard SaaS content playbook underperforms, what replaces it, and the distribution to expect from a product-led approach.
• Product-led SEO: when the page is the product: What product-led SEO means in practice, which products it suits, and the distribution you should expect before you commit to it.
• Comparison pages: high intent, and easy to get wrong: Why "X vs Y" queries convert, what makes a comparison page rank rather than read as marketing, and when not to build one.
• "Alternative to X" pages, and why they beat comparisons: Alternative queries capture people already leaving a product. What the page needs to say, and the mistake that makes it useless.
• Your pricing page is a search result, whether you planned it or not: Pricing queries are the highest-intent traffic you will get. Why hiding prices costs rankings and trust, and what the page needs.
• Should your SaaS have a blog? Only under one condition: Measured on one dataset, a 75-post blog produced 2.32% of traffic. When a blog is worth it, and what to build instead.

Reading the results page

• How to read a results page before you commit a quarter to it: A ten-minute manual SERP analysis that predicts winnability better than any difficulty score, and what each signal means.
• Featured snippets: how they are chosen, and whether you want one: What earns a snippet, the formats that get selected, and the honest tradeoff between visibility and the click you lose.
• People Also Ask, and what it tells you for free: PAA is a free map of the questions behind a query. How to use it for research, and the modest case for optimising into it.
• Competitor analysis that changes what you do next: Most competitive analysis produces a report nobody acts on. The four questions worth answering, and how to answer them cheaply.

Speed and experience

• Core Web Vitals: small ranking factor, large conversion factor: What the three metrics measure, how much they move rankings, and why the conversion effect is the one that pays.
• Why your Lighthouse score disagrees with your users: Lab tests and field data measure different things. Which to trust, why they diverge, and what a good score does not tell you.
• INP, the metric that measures whether your page feels broken: What Interaction to Next Paint measures, what causes a bad score, and the fix that is usually cheaper than making the work faster.
• Dead clicks: the cheapest conversion fix most sites never look at: A dead click is a person clicking something that does nothing. What causes them, how to find yours, and why the fix takes an afternoon.
• Does page speed affect rankings? Yes, and less than you hope: The honest size of the ranking effect, why speed still deserves the work, and how to tell if speed is genuinely your problem.

Domains and international

• Does your domain extension matter? The .cv case, measured: Country-code domains carry a geographic association that generic ones do not. What that costs, and what one site measured.
• hreflang, and whether you actually need it: What hreflang does, the errors that make it useless, and the much smaller number of sites that genuinely need it.
• International SEO when you are small: mostly, do less: The order to expand in, why translation is rarely the first move, and how to find which markets already want you.

Testing and proof

• You cannot A/B test SEO the way you test a landing page: Why the standard split test does not apply, what a valid SEO test looks like, and how to avoid claiming credit for noise.
• How long to wait before judging an SEO change: The sequence in which metrics move, why clicks are the slowest signal, and the window that actually settles a question.
• Statistical significance in SEO, without the false confidence: What a p-value tells you about a traffic change, what it does not, and why effect size matters more than the test.
• Seasonality: the reason your comparison is probably wrong: Weekly, monthly and annual cycles distort almost every before-and-after. How to spot yours and compare against the right period.
• How to run an SEO experiment you can actually believe: A practical design for testing a change across a page set, including what to write down before you start.

Logs and crawling

• Server logs: the only record of what crawlers actually did: What logs tell you that Search Console cannot, how to read them, and the privacy reason nobody offers this as a web tool.
• How to tell real Googlebot from something pretending: User agents are trivially forged. The two reliable verification methods, and what fake crawler traffic costs you.
• Soft 404s: pages that say 200 and mean nothing: What a soft 404 is, the three ways sites create them accidentally, and why they are worse than a real 404.
• Measuring crawl budget from your own logs: A step by step read of a log file that answers whether crawl is your constraint, with the numbers to compute.

Traffic and revenue

• What a search click is actually worth to you: Three ways to value organic traffic, which one to use when, and the mistake that inflates every SEO business case.
• Calculating SEO ROI without fooling yourself: The formula, the costs people leave out, and why the payback period matters more than the ratio.
• When SEO starts paying, on one measured timeline: The month by month shape of a real site from zero, where the inflection was, and what the early months look like.
• Traffic that does not convert, and the four reasons why: Ranking for the wrong intent, an interface that fails, a mismatch between page and promise, or a product nobody wants.
• From free tool users to customers, and why the bridge is hard: A free tool can produce enormous usage and almost no revenue. What the gap is made of, and the honest ways across it.

How to actually work

• An SEO audit that ends in decisions, not a 200-item list: The checks that change what you do, in the order that finds problems fastest, and the ones safe to skip.
• Prioritising SEO work when everything looks urgent: A ranking method based on what a change is worth in clicks, and the categories of work that reliably waste quarters.
• What to do first, in order, on a site with nothing: The first five moves for a site with no traffic, why they are in this order, and what to ignore until later.
• An SEO report a founder can actually act on: The four numbers worth reporting, the ones that mislead, and how to present a channel whose results lag by a quarter.
• When to hire for SEO, and what to ask before you do: The point at which outside help pays, the questions that separate operators from reporters, and what to never outsource.
• How to see your page the way a crawler does: View source is not the inspector. How to check what actually lands in your HTML, and what to look for once you have it.
• Internal linking: the lever you fully control: How internal links shape discovery, importance and crawl frequency, and the audit that finds your orphaned pages.
• Should you delete old content? Usually update, sometimes remove: How to decide page by page between keeping, updating, merging and deleting, using demand data rather than instinct.