Most SEO advice optimises for a search engine that no longer exists. Here is what still moves rankings, what has stopped working, and how to tell the difference.
Most SEO advice fails for a simple reason: it optimises for a search engine that no longer exists. Keyword density, exact-match anchor text and monthly article quotas were tactics for an algorithm that read pages as bags of words. That has not been how search works for a long time.
What follows is what still reliably moves rankings, what has stopped working, and how to tell the difference when someone tries to sell you the latter.
Search intent decides everything upstream
Before writing anything, look at what currently ranks for the phrase you want. Not to copy it, but to identify what kind of answer the results are made of.
If the first page is all product categories, a blog post will not rank there no matter how good it is. If it is all step-by-step guides, a sales page will not either. Google has already learned what people who type that phrase want, and the results are that judgement made visible. Publishing the wrong format is the most common reason a genuinely good page never ranks.
This also tells you where commercial pages belong. A phrase like "web development services" is transactional and deserves a service page. "How to speed up a website" is informational and deserves a guide. Trying to make one page serve both usually serves neither.
Cover the topic, not the keyword
Search engines now assess whether a page addresses the subject completely, using related concepts rather than repeated phrases. A page about two-factor authentication that never mentions authenticator apps, recovery codes, SIM-swap risk or passkeys is visibly incomplete, regardless of how many times the target phrase appears.
A practical method: write down every question a reasonable person would ask about the topic, then check that your page answers them. The questions people also ask, shown in the results themselves, are a useful free source. So is your own support inbox, which usually contains better questions than any keyword tool.
Length is a consequence of this, not a target. Pages get long because the topic requires it. Padding a thin page to hit a word count adds nothing a search engine values and makes the page worse to read.
Demonstrate experience you actually have
Google's quality guidance emphasises experience, expertise, authoritativeness and trust. Of those, experience is the one small businesses systematically underuse and the one that is hardest to fake.
Concretely, this means writing things only someone who has done the work would know. The failure mode you see repeatedly. The step where clients usually change their mind. The recommendation you would give against your own commercial interest. A page that says "based on the migrations we have run, the step people underestimate is DNS propagation during business hours" is doing something a generic article cannot.
The supporting signals matter too: a real author with a real biography, a business address, published prices or at least a clear explanation of how you price, and citations to primary sources when you state a fact. None of these are ranking tricks. They are how a reader — and an evaluator — decides whether to believe you.
Technical foundations that still matter
Technical SEO has narrowed to a few things that genuinely affect whether you can rank at all.
Crawlability and indexing
A page that cannot be crawled cannot rank. Check that important pages are not blocked in robots.txt, do not carry an accidental noindex tag, and are reachable through links rather than only through search or a sitemap. In Google Search Console, the pages report tells you what has been excluded and why — this is the first place to look when a page simply is not appearing. Fixing what it surfaces is usually a development task, and our web development team handles the ones that turn out to be structural.
Core Web Vitals
Page experience is part of the ranking system, with three published thresholds: Largest Contentful Paint at or below 2.5 seconds, Interaction to Next Paint at or below 200 milliseconds, and Cumulative Layout Shift at or below 0.1, all measured at the 75th percentile of real visits. Google documents these on web.dev.
Keep this in proportion. Speed rarely lifts a weak page above a strong one, but between two comparable pages it is a genuine advantage — and it affects conversion regardless of ranking. Our website speed guide covers how to find which of the three you are failing. Where the cause turns out to be server response time rather than front-end weight, that is work for our infrastructure team rather than a content fix.
Structured data
Schema markup does not directly improve rankings. It makes eligibility for rich results possible — review stars, FAQ panels, event details, product pricing. Those change how much space you occupy in the results and how many people click, which is worth having. Mark up what is genuinely on the page; marking up things that are not is a policy violation.
Internal linking
This is the most underused lever available and it costs nothing. Links between your own pages distribute authority and tell search engines which pages matter and what they are about. A new article with no internal links pointing at it is close to invisible. Whenever you publish, add links to it from existing relevant pages — not just from the blog index.
Local search, if you serve a place
If you have a physical location or a defined service area, your Google Business Profile frequently drives more enquiries than your website does. Complete it properly, keep the name, address and phone number identical everywhere they appear, choose categories precisely, and collect reviews consistently rather than in occasional pushes. We cover the detail in our guide to local SEO.
Links: still important, still misunderstood
Links from other sites remain one of the strongest signals, which is why the market for buying them persists. Bought links breach Google's spam policies and the risk sits with you, not the seller.
What works is slower and duller: publishing something worth referencing, contributing genuine expertise where your industry gathers, supplier and partner relationships, local press and trade publications, and being the business people cite when they explain something. A handful of links from relevant, trusted sites outperforms hundreds of low-quality ones.
Digital PR is the honest version of link building, and it works because it produces something a journalist actually wants rather than something an algorithm was supposed to fall for.
What has stopped working
| Tactic | Why it fails now |
|---|---|
| Keyword density targets | Search engines match meaning, not repetition. Hitting a percentage makes writing worse and changes nothing. |
| Publishing on a quota | Volume without usefulness is what the helpful content guidance targets. Thin posts can hold back the whole site. |
| Exact-match anchor text everywhere | An unnatural pattern that reads badly and looks manipulated. |
| Separate pages per town | Near-identical pages with the place name swapped are treated as doorway pages. One strong page beats forty thin ones. |
| Meta keywords tag | Unused by Google for many years. Harmless, but no reason to spend time on it. |
| Chasing a single ranking | Results are personalised and vary by location and device. Track enquiries and trends, not one position. |
Measuring it honestly
Rankings are a means, not the outcome. The measures worth reporting are organic sessions to pages that matter, the enquiries or sales attributable to them, impressions and click-through rate in Search Console, and how many distinct pages bring in traffic — a healthy site earns visits across many pages, not just the homepage.
Be patient in a specific way: expect months, not weeks, for new content in a competitive area, and judge on trend rather than week-to-week movement. Anyone promising first-page results within thirty days is either targeting phrases nobody searches for or planning to do something that will cost you later.
Frequently asked questions
How long does SEO take to work?
For a new site or a competitive term, typically several months before meaningful change and longer for the full effect. Technical fixes can show up faster because they remove a barrier rather than build authority. Anyone quoting a guaranteed timeframe is guessing.
Is AI-written content against Google's rules?
No. Google's spam policies target content produced at scale primarily to manipulate rankings, regardless of how it was made. The test is whether the page is genuinely useful and accurate. Content that is thin, generic or wrong is the problem — and that is true whether a person or a machine produced it.
How many blog posts should we publish?
Fewer and better beats more and thinner, particularly for smaller sites. One thorough, accurate page that answers a question completely will outperform four shallow ones, and it will not drag down the rest of the site.
Do we need to publish constantly to stay ranked?
No. Updating and improving existing pages that already rank is usually a better use of effort than adding new ones. Freshness matters for topics that genuinely change; an evergreen guide does not need rewriting quarterly.
Should we buy links?
No. It breaches Google's policies, the penalty risk sits with your site, and the links are usually worthless anyway because they come from sites nobody reads.
Where to put your effort
If you are starting from a standing position: fix anything preventing indexing, match your page types to what actually ranks for your target phrases, make your most commercially important pages genuinely thorough, and link between your own pages deliberately. That sequence solves more problems than any tactic further down the list.
Our SEO services begin with an audit that identifies which of these is actually holding you back, rather than applying the same checklist to every client. If the answer is that your pages are fine and you need links, we will say so. Get in touch for an honest assessment.
