How long does SEO take?

Three to six months before you see movement worth discussing. Six to twelve before it is producing enquiries at a level that changes anything commercially. Longer on a new domain, longer again in a competitive sector, and shorter than both if the site has an existing history and the problems are technical rather than competitive.

That range is unsatisfying because it is wide, and it is wide because the variables genuinely dominate. The same work on two different sites produces results months apart, and anybody quoting a confident single figure before looking at your domain is describing a sales process rather than a timeline.

These are the factors that genuinely shift the number, in rough order of how much difference they make. The first two swamp everything else.

This is the schedule to hold an agency to. If none of the first-band items has happened by month three, the problem is delivery rather than patience.

Rankings are a lagging indicator and a volatile one. These move earlier and are far better evidence that the work is landing, which matters because they let you make a continue-or-stop decision months before positions would have told you anything.

• Impressions in Search Console for the topics you are targeting. This moves first, often within weeks, and it means Google is now considering you for those queries at all.
• The number of distinct queries you appear for. Breadth grows before depth — you show up for more things before you show up higher for any of them.
• Average position on long-tail phrases, which improves well before head terms do.
• Pages indexed against pages published. If published pages are not being indexed, nothing else matters yet.
• Crawl frequency in the logs or in Search Console. A site being crawled more often is a site being taken more seriously.
• Branded search volume, which rises when other channels are working and is worth separating from the rest.
• Click-through rate on existing positions, which can be improved in days by rewriting titles rather than by ranking higher.

Patience is genuinely required in this discipline and it is also the most common excuse for poor delivery. The distinction is evidence: an engagement that is working produces visible movement in the leading indicators long before it produces revenue, so an absence of those is a real signal rather than impatience.

By month three, impressions should be moving for the topics being targeted. By month six there should be position improvement on something, even if it is not yet the phrases that matter commercially. If neither has happened, the honest questions are whether the targets were winnable in the first place, whether anything has actually shipped, and whether the site has a technical problem nobody found.

Most delays are not algorithmic. They are organisational, and they are worth naming because they are within your control in a way the competition is not.

• Approval cycles. Content sitting in review for three weeks turns a twelve-month programme into an eighteen-month one, and it is the single most common cause of slippage.
• Developer availability. Technical fixes queued behind a product roadmap can wait months, and they are usually the fastest wins available.
• Changing direction. Retargeting the strategy at month four resets most of the clock. Choose targets you can commit to for a year.
• Publishing too little. Content velocity is the main lever you control, and halving it roughly doubles the timeline.
• Chasing head terms first. The order that works is long-tail to short-tail, and starting at the top wastes the months that would have built the authority to compete there.