The Complete Guide To Best Practices Nobody Actually Follows
There is a playbook. You've seen it. Twenty-three PowerPoint slides about header tags and crawl budget and mobile-first indexing. Someone with "SEO Strategist" in their LinkedIn headline presented it at a webinar you attended while pretending to be on a call. They cited studies. They showed charts shaped like hockey sticks. They said the words "best practice" forty-seven times in forty minutes.
And then you looked at their site and it ranked for nothing except their own name.
This is the entire industry. A monument to advice nobody takes, least of all the people giving it. Best practices are the participation trophy of search optimization. Everyone holds one up. Nobody wins with it.
Welcome to the complete guide to the things you're supposed to do that the people actually ranking don't do.
The Best Practice Industrial Complex
Best practices exist because tool companies need upsells and consultants need retainers. You cannot monetize "make a better page." You can monetize a seventeen-point technical audit that finds problems a script invented.
The loop is perfect. Tools flag violations. Gurus explain violations. Courses teach you to fix violations. Agencies sell you violation fixes. The site still doesn't rank. The tool finds new violations. The carousel spins forever.
Best practices are not designed to make you rank. They're designed to make you believe ranking is a checklist you haven't completed yet. One more plugin. One more schema type. One more canonical tag and maybe this time Google will love you back.
Spoiler: Google doesn't read your robots.txt like a love letter.
What The Rankings Actually Show
Here is what sites ranking in the top three have in common: they answer the query better than the sites below them. That's it. That's the list.
They do not have matching meta descriptions at exactly 155 characters. Half of them are missing H1 tags. Their image alt text is "IMG_2847.jpg" from someone's phone. They have render-blocking JavaScript older than TikTok. Their Core Web Vitals look like a heart attack in real time.
And they rank. They rank because they did the one thing best practices forgot to mention: they made a page someone wanted to click on and stay on and link to.
Best practices treat search like a form you fill out for the DMV. You bring the right documents in the right order and Google gives you a ranking. This has never been how it works. This will never be how it works. But it's easier to sell a checklist than to admit the truth: most of what ranks is unmeasurable, unrepeatable, and utterly allergic to seventeen-step frameworks.
The Best Practices You're Definitely Ignoring
Let's go through the big ones. The commandments. The stuff they tell you will tank your site if you skip it.
Alt Text On Every Image
You should do this. For accessibility. For users who need it. Not because Google ranks you higher. Google has seen seventeen billion product images titled "image1.png" and ranked them just fine. Alt text is the right thing to do. It is not the ranking thing to do. There is a difference and the gurus keep pretending there isn't.
Unique Meta Descriptions For Every Page
Google rewrites seventy percent of them anyway. They rewrite them because yours are bad. They are bad because you wrote them for Google instead of for the person deciding whether to click. The best practice says be unique. The actual practice is: be interesting or get rewritten. Pick one.
Internal Links With Exact Match Anchor Text
This came from a 2011 SEO playbook that also recommended article directories and three-way link exchanges. It stayed alive because tools can audit it and agencies can fix it for $4,000. Meanwhile the sites ranking on page one are saying "click here" and "read this" like normal human beings who don't optimize conversations.
Keyword Density Between 1-3 Percent
This best practice is hold music from the Bush administration. Nobody with a site in the top ten has ever counted keyword density. Nobody currently ranking is counting it now. Keyword density is what happens when you let a formula replace judgment. Google is not grading your essay with a percentage ruler. Write like you're trying to explain something, not like you're trying to trick a robot into thinking you explained something.
Mobile Page Speed Under Three Seconds
Great goal. Not a ranking factor the way they sold it to you. Google's own sites don't hit three seconds. News sites loaded with ad sludge rank fine at eleven seconds. Page speed matters when it's so bad people leave. It does not matter the way tool companies matter when they're trying to sell you a CDN plan.
Why Smart People Keep Repeating Dumb Advice
Because it's safe. Nobody ever got fired for recommending best practices. You can lose a client by saying "your content is boring" but you'll keep them for six months by saying "your schema markup needs tuning."
Best practices are the SEO equivalent of thoughts and prayers. They let you look busy without risking anything. They give you something to put in a report. They turn the uncontrollable chaos of trying to rank into a project plan with milestones and deliverables.
And they don't work. But by the time the client figures that out you've already attended a conference and learned seventeen new best practices to sell them next quarter.
What The People Actually Ranking Are Doing Instead
They're ignoring you. They're ignoring the playbook. They're ignoring the best practice brigade and doing the stuff that best practices warn you not to do because it's "not scalable" or "too risky" or "not what Google wants."
They're publishing six-thousand-word guides that violate every readability score. They're using keyword stuffing that would make a 2009 affiliate marketer blush—except it works because the page is actually useful. They're buying links. They're building tools that exist only to get links. They're doing weird technical things with JavaScript that the audits flag as critical errors.
And they rank. They rank because they optimized for the user or the algorithm or the link or the moment, not for the checklist.
Best practices optimize for compliance. Rankings optimize for results. These are not the same game.
The Practices Google Actually Follows
Here's a fun one: go look at Google's own properties. Search Console. Google Ads. Gmail. The tools they build and rank with.
They violate half the best practices they told you to follow. Render-blocking resources everywhere. No schema markup. Mobile speeds that would make a Lighthouse audit file a restraining order. Their own documentation pages have broken links and redirect chains and meta descriptions that cut off mid-sentence.
Do they rank? Yes. Because they're Google. But also because best practices were never the actual requirement. They were the cover story. The press release. The thing they told small sites to worry about while the big sites did whatever worked.
If Google doesn't follow Google's best practices, why are you?
When Best Practices Actually Matter
Sometimes they do. Not often. But sometimes.
Crawlability matters if Google literally cannot see your pages. Mobile usability matters if your site is unreadable on a phone. HTTPS matters because browsers will scare users away from HTTP.
These are not competitive edges. These are baseline functional requirements. Fixing them won't make you rank. Ignoring them might keep you from ranking at all. There's a difference between "this will help" and "this won't actively destroy you."
Most best practices are the second thing wearing a costume of the first thing.
What To Do Instead Of Following Best Practices
Stop auditing. Start looking. Go find the pages that rank for what you want to rank for. Not what the tools say they rank for. What they actually rank for when you type it into Google like a normal person.
Look at those pages. What do they do? Not what do they have in their meta tags. What do they do that makes someone click and stay and link?
Now go do that. Not the same topic. The same energy. The same utility. The same reason to exist.
Best practices will tell you this is not a strategy. Best practices are wrong. This is the only strategy. Everything else is cosplay.
The Real Best Practice Nobody Sells
Make something worth ranking. That's it. That's the whole guide.
Not worth ranking according to a rubric. Worth ranking because it's better than the ten things above it. Better written. Better researched. Better designed. More useful. More specific. More honest. More something that matters to the person searching.
You cannot audit your way to better. You cannot checklist your way to useful. Tools will tell you that you can because tools exist to be sold. Gurus will tell you that you can because gurus exist to sell courses about tools.
But the sites that rank—really rank, consistently rank, rank without tricks or schemes or nine-step frameworks—are the ones that stopped optimizing for Google and started optimizing for the thing Google is trying to approximate: actual value.
Best practices are what you do when you don't know what actually works. The moment you know what works, the practices stop mattering.
That moment is now.
Frequently Asked Questions
- Why does everyone talk about best practices but nobody can prove they actually work?
- Because best practices are easy to sell and impossible to disprove. When a site doesn't rank after following them, the consultant blames something else—competitive difficulty, domain age, content quality, anything except the checklist they sold you. Best practices exist in a perfect loop: they're always recommended, rarely questioned, and never held accountable for results. They're the industry's safety blanket. No one ever got fired for recommending schema markup, even when it changed nothing.
- Are SEO best practices just made up by tool companies trying to sell subscriptions?
- Not entirely, but the incentives are clear. Tool companies need reasons for you to pay monthly. Audits need problems to find. If the only advice was "make a better page," you wouldn't need a $300/month platform to tell you your H2s are out of order. Many best practices started as legitimate observations, then got weaponized into compliance requirements by companies that profit from finding violations. The practices aren't always fake. The urgency around them usually is.
- What happens if I ignore all the best practices and just focus on what actually ranks?
- You'll probably rank faster than if you spent six months fixing canonical tags. The sites in position one through three are not there because they followed a checklist. They're there because they did something better than the competition—answered the query more completely, earned more links, built more trust, created more value. Best practices are table stakes at most. Ignoring them to focus on actual competitive advantage is not reckless. It's strategy.
- Do Google engineers actually follow the best practices they tell us to follow?
- No. Google's own properties violate recommendations they publish in their documentation. Search Console has render-blocking resources. Their support pages have broken links. Their mobile speeds fail their own Core Web Vitals thresholds. This isn't hypocrisy—it's proof that best practices are not the actual ranking system. They're the public-facing advice for small sites while the real game runs on different rules. If Google doesn't need to follow them to rank, neither do you.
- Why do SEO gurus recommend best practices they don't use on their own sites?
- Because their sites don't rank from best practices either. They rank from brand searches, backlinks from speaking gigs, and mentions in industry publications. The guru's site exists to sell courses, not to prove the tactics work. Recommending best practices is safe, repeatable, and packageable into a $2,000 curriculum. Telling you what they actually did—bought links, gamed Reddit, exploited a loophole—doesn't scale into a webinar funnel.
- Is following best practices why my site still doesn't rank?
- Possibly. Best practices are often a distraction from the real problem: your content isn't better than what already ranks. You can have perfect technical SEO and lose to a site with broken images and no meta description if that site answers the query more completely. Best practices give you something to fix while avoiding the harder truth—you need to build something more useful, more linkable, or more relevant. Checklists feel productive. They're rarely sufficient.
- What best practices did sites that actually rank completely ignore?
- Most of them. Top-ranking pages routinely skip unique meta descriptions, have zero schema markup, use generic anchor text, exceed recommended word counts by thousands, load slow, have no H1 tag or three H1 tags, and violate every readability formula. They rank because they're comprehensive, trusted, linked-to, or just better at satisfying the search intent. Best practices assume ranking is compliance. Reality shows ranking is competition. The two have almost nothing to do with each other.