SEO Best Practices That Still Matter in 2021

SEO best practices in 2021 are less about chasing algorithm updates and more about getting the fundamentals right at a higher standard than your competitors. Core Web Vitals became a confirmed ranking factor, content quality thresholds rose, and Google’s ability to assess topical authority improved meaningfully. The sites that gained ground were the ones that had already done the unglamorous work: clean architecture, genuine expertise on the page, and a link profile built on something real.

This article covers what actually moved the needle in 2021, where most sites were still leaving performance on the table, and how to think about the year’s changes without overcorrecting on things that never mattered as much as the SEO industry suggested.

What Actually Changed in SEO in 2021

Every year the SEO industry generates a wave of content about what changed, most of it written within 48 hours of a Google announcement and revised quietly when the predictions turned out to be wrong. I spent a long time in agency environments watching teams scramble to brief clients on algorithm updates before anyone had enough data to say anything meaningful. The scramble was mostly theatre. The clients who fared best were the ones we had already put on a sound footing, not the ones we rushed to “fix” after every core update.

2021 had three genuinely significant developments. First, the Page Experience update rolled out, incorporating Core Web Vitals as ranking signals. Second, Google confirmed that passage indexing was live, meaning it could rank specific passages within a page rather than just the page as a whole. Third, product reviews got a dedicated update targeting thin, affiliate-heavy review content that added no real value over manufacturer descriptions.

Everything else, the speculation about link velocity, the debates about exact-match anchor text ratios, the endless discourse about meta descriptions, was noise. Useful noise in some cases, but noise nonetheless.

If you want to understand how all of these pieces connect into a coherent approach, the Complete SEO Strategy hub covers the full picture, from technical foundations through to content and authority building.

Core Web Vitals: What They Measure and What They Don’t

Core Web Vitals measure three things: how fast the largest visible element loads (Largest Contentful Paint), how quickly a page responds to the first user interaction (First Input Delay), and how much the layout shifts during loading (Cumulative Layout Shift). Google confirmed these became ranking factors in June 2021 as part of the Page Experience update.

The practical effect was smaller than the industry anticipated. Google was explicit that a page with excellent Core Web Vitals scores would not outrank a page with significantly stronger relevance and authority simply because it loaded faster. What the update did was create a tiebreaker, and it raised the floor for sites in competitive markets where relevance and authority were roughly equal across the top results.

Where the scores genuinely mattered was in sectors where user experience had been neglected for years. I worked with a number of e-commerce businesses over the years where the product pages were technically functional but genuinely unpleasant to use on mobile. Cumulative Layout Shift was often the culprit: ads loading late, images without specified dimensions, banners injecting themselves into the page after the content had already painted. Fixing those issues improved both the scores and the conversion rate, which is a more direct commercial win than any ranking change.

The Semrush breakdown of bad SEO practices is worth reading alongside your Core Web Vitals audit. Many of the technical issues that drag down performance scores are the same ones that create other crawling and indexing problems.

Content Quality in 2021: A Higher Bar, Not a Different Game

The content quality conversation in SEO has been muddied by years of conflating word count with depth, keyword density with relevance, and publication frequency with authority. In 2021, Google’s systems got better at distinguishing between these things, and the distinction started to show up in rankings in ways it had not previously.

The product reviews update was the clearest signal. Pages that summarised existing information without adding any original perspective, testing, or expertise lost visibility. Pages written by people who had actually used the products, or by editors who had developed genuine category knowledge, held their ground. This is not a new principle. It is the same principle that has always governed good editorial work. What changed is that Google got better at detecting which side of the line a page was on.

When I was running agencies, one of the consistent mistakes I saw was treating content as a volume problem. Teams would brief writers on keyword lists and word counts and call it a content strategy. The output was technically correct but commercially inert. It did not rank well because it did not deserve to rank well. It added nothing to the conversation. The sites that invested in genuine subject matter expertise, either by hiring people who knew the industry or by creating structured processes for extracting knowledge from internal experts, consistently outperformed the volume-first approach over any 18-month horizon.

Passage indexing added a structural dimension to this. If Google can surface a specific section of your article independently of the rest, the quality of each section matters more than it did when Google was evaluating the page as a single unit. A 3,000-word article with two excellent sections and 1,500 words of filler is now a weaker asset than it looks. Every section needs to hold up on its own terms.

Technical SEO Priorities That Held Up in 2021

Technical SEO is the area where the gap between what gets written about and what actually matters is widest. The SEO industry has a tendency to generate complexity because complexity justifies retainers. The honest version of technical SEO in 2021 was a short list of things that genuinely affected crawling, indexing, and rendering, and a long list of things that were theoretically interesting but practically irrelevant for most sites.

The short list: crawl budget management for large sites, structured data implementation, mobile-first indexing compliance, canonical tag accuracy, and Core Web Vitals scores as discussed above. If those five areas were in good shape, the technical foundation was solid. Everything else was marginal.

Structured data deserves specific attention. Schema markup does not directly improve rankings, but it enables rich results, and rich results improve click-through rates in ways that can meaningfully affect traffic even at stable ranking positions. FAQ schema, How-To schema, and Review schema were all active in 2021 and worth implementing where the content genuinely warranted it. The emphasis on “genuinely warranted” is deliberate. Marking up content as a FAQ when it is not actually structured as questions and answers is the kind of shortcut that tends to get corrected in subsequent quality updates.

For teams that want to work through their technical baseline systematically, the Crazy Egg roundup of SEO tools covers the audit tools worth using, and the Buffer guide to free SEO tools is useful if budget is a constraint. The tools are not the strategy, but they make the diagnostic work faster.

E-A-T: What It Is and What It Is Not

E-A-T stands for Expertise, Authoritativeness, and Trustworthiness. It comes from Google’s Search Quality Evaluator Guidelines, the document that human quality raters use to assess search results. It is not a direct algorithmic ranking signal in the sense that there is no E-A-T score that Google computes and applies to your page. What it is, is a framework that shapes how quality raters evaluate pages, and those evaluations feed into how Google trains and refines its ranking systems over time.

The practical implication is that E-A-T matters most in YMYL categories, which stands for Your Money or Your Life. Health, finance, legal, and safety content is held to a higher standard because the consequences of bad information in those categories are real. A misleading article about a kitchen gadget is annoying. A misleading article about drug interactions or investment risk is dangerous. Google’s quality raters apply more scrutiny to YMYL pages, and that scrutiny has shaped the algorithm’s behaviour in those sectors.

For YMYL sites, the signals that support E-A-T are concrete: author credentials displayed on the page, editorial review processes documented in site policies, citations to primary sources, accurate and current information, and a site-level reputation that can be verified through external references. I judged the Effie Awards for several years, which gave me a useful perspective on what genuine credibility looks like versus what is performed. The difference is usually visible in the details. Credentials that are vague, testimonials that are unverifiable, and “expert” claims that cannot be traced to anything real are the patterns that erode trust, both with readers and with the systems that are trying to assess trustworthiness at scale.

The Ahrefs guide to SEO for medical practices is a useful worked example of how E-A-T considerations translate into practical decisions for a high-stakes YMYL sector.

Link Building in 2021: Quality Over Quantity, With No Exceptions

Link building in 2021 was still the area where the most money was being wasted on things that either did not work or actively created risk. The guest post link mill, the private blog network, the “editorial” link sold through a broker: these approaches have been degraded by successive algorithm updates and manual actions, and the sites still relying on them were either lucky or had not yet been caught.

What worked was the same thing that has always worked: earning links by publishing something genuinely worth linking to, and then making sure the right people knew it existed. Original research, comprehensive resources, tools, and data-driven analysis attracted links in 2021 for the same reason they attracted links in 2011. They gave other publishers something useful to reference.

The distribution side is where most content teams underinvest. Publishing is not enough. The people most likely to link to your content are editors, journalists, and researchers who are actively covering your topic area. Identifying them, building relationships with them, and making your content easy to find and cite is the work. It is slower than buying links and it does not scale as cheaply, but it builds something that compounds rather than something that can be undone by a single algorithm update.

One useful tactical resource is the Moz piece on keyword labelling, which is relevant to how you organise your content strategy to support link acquisition. Understanding which content clusters have the strongest link potential helps you prioritise where to invest in depth and original research.

Keyword Strategy in 2021: Intent Over Volume

Keyword strategy in 2021 was less about finding high-volume terms and more about understanding what a searcher actually needed when they typed a query. Google’s ability to match intent had improved to the point where a page that answered a question well could outrank a page that had optimised more aggressively for the exact keyword phrase.

The practical shift this required was in how teams briefed content. Instead of starting with a keyword and building a page around it, the better approach was to start with the question behind the keyword and build a page that genuinely answered it. This sounds obvious. It was not how most content briefs were written in agencies I encountered, where keyword targets and word counts dominated the brief and the actual reader experience was an afterthought.

Volume still mattered for prioritisation. A keyword with no search volume is not worth targeting regardless of how well you could answer it. But volume was a starting point for prioritisation, not a proxy for content quality or strategic fit. A 200-search-per-month keyword that converts at 15% is more valuable than a 20,000-search-per-month keyword that converts at 0.1%, and the content required to rank for each is completely different.

For teams building out a keyword taxonomy, the Moz approach to keyword labels is a practical way to organise a large keyword set by intent type, which makes prioritisation and content briefing more systematic.

Accessibility as an SEO Factor

Accessibility is not traditionally discussed as an SEO best practice, but the overlap is more significant than most teams appreciate. The same signals that make a page accessible to users with disabilities, descriptive alt text, logical heading structure, readable font sizes, sufficient colour contrast, also make a page easier for search engines to interpret. The Moz analysis of accessibility and SEO makes the case clearly: the two disciplines are not competing priorities, they are complementary ones.

In 2021, with mobile-first indexing fully rolled out, the accessibility of the mobile experience became directly relevant to how Google assessed and ranked pages. Sites that had built their mobile experience as an afterthought, with small tap targets, unreadable text, and content that required horizontal scrolling, were penalised in ways that showed up as both ranking drops and conversion rate declines. The commercial and the technical were aligned, as they usually are when you look closely enough.

What Most Sites Got Wrong in 2021

The most common mistake I saw in 2021, and this has been consistent across every year I have been in the industry, was treating SEO as a series of isolated tactics rather than a coherent system. Teams would fix their Core Web Vitals scores without addressing their content quality. They would publish new content without auditing what was already on the site. They would build links to pages that had not been optimised for the queries they were targeting. Each tactic was technically correct. The system was incoherent.

The second most common mistake was measuring the wrong things. Organic traffic as a headline metric tells you very little without being broken down by landing page, by query type, by device, and by what happened after the visit. I have sat in more client meetings than I can count where an agency presented organic traffic growth as evidence of success while the commercial metrics were flat or declining. Traffic that does not convert is not an asset. It is a distraction.

The third mistake was confusing activity with progress. Publishing 50 articles in a quarter is activity. Publishing 10 articles that each rank in the top three for commercially relevant queries is progress. The industry has historically rewarded the former because it is easier to measure and easier to sell. The discipline is in holding to the latter even when the short-term output looks less impressive.

There is a broader framework for how all of these considerations fit together. If you are building or refining your approach, the Complete SEO Strategy hub is the right place to work through the full picture, from technical health through to content architecture and authority building.

For teams that want to develop their skills alongside their strategy, the Crazy Egg guide to SEO courses covers the training options worth considering. And for a useful external reference point on how Google itself approaches SEO standards, the Search Engine Land piece on Google grading its own SEO practices is a revealing read.

About the Author

Keith Lacy is a marketing strategist and former agency CEO with 20+ years of experience across agency leadership, performance marketing, and commercial strategy. He writes The Marketing Juice to cut through the noise and share what works.

Frequently Asked Questions