Key takeaways
- Technical health, intent-led pages and real mentions still do the work
- Volume content, link packages and ranking reports are where budget leaks
- Six months is a realistic horizon for compounding results
Every year produces a new theory about how SEO died. Every year the same four things keep working, for the same reason: they are proxies for a page that deserves to rank.
Still works: technical health
Fast pages, clean indexing, one canonical version of every URL, and a sitemap that matches reality. None of this is clever, which is why agencies skip it in favour of content calendars that produce more pages for an unindexable site.
If you can answer yes to four questions, you are ahead of most competitors: does every page have one canonical URL, does the sitemap return only 200s, is the mobile layout genuinely usable, and is the largest content element loading in under 2.5 seconds.
Still works: one page, one question
Pages written to answer a specific question, thoroughly, outperform broad guides that answer everything shallowly. The pattern that wins is unglamorous: state the answer in the first hundred words, then expand with the detail that proves you know what you are talking about.
This is also the structure AI answer engines reward, which is convenient — the same work serves both.
Still works: buying intent over volume
A keyword with 20,000 monthly searches and the wrong intent sends you traffic that will never buy. A keyword with 200 searches and an evaluative intent is worth more than the other one, and costs less to win. Map every target to a stage of the buying journey before writing anything.
Still works: mentions from sites with their own audience
Not guest posts on content farms. Mentions from industry publications, local media, associations, partners and suppliers — places that had a reason to exist before somebody sold a link package.
Stop paying for: bulk content
Fifty thin pages targeting variants of one keyword will lose to one page that earns the click and keeps it. If your content plan is measured in articles per month with no quality gate, the plan is the problem.
Stop paying for: link packages
If a vendor cannot tell you the domain name of a link before you pay, the link is not an asset, it is a liability with a delivery date.
Stop paying for: technical audits that change nothing
A 120-page audit is not a deliverable. A prioritised list of twelve fixes, with the three highest-impact ones shipped, is. Ask what gets implemented, not what gets reported.
Stop paying for: ranking reports
Rankings move daily and mean less than they used to. Search Console clicks, impressions and conversions are the numbers that correspond to money. If the report cannot show those, it is decoration.
Stop paying for: monthly "strategy" that is a call
A call is not a deliverable. The deliverable is a written decision: this month we do X, we stop Y, and here is what we expect to see in ninety days.
Want the audit before the retainer?
One-time SEO audit — $750A realistic timeline
Technical fixes: two to four weeks. Content and authority: three to six months of consistency before the curve bends. In competitive categories, longer. Anybody who tells you otherwise is either selling something or has not done it recently.
