How to measure SEO when Google hides the search terms
Google will not tell you which search brought each visitor. Here is what can still be measured honestly and which numbers mislead owners most.
The gap in your reporting
Your analytics shows organic traffic arriving and says nothing about what anybody searched for. Google stopped passing the search term to analytics more than a decade ago when it moved search to a secure connection. The keyword column has been empty ever since and no setting restores it.
Search Console fills part of the gap because it reports queries directly from Google. It fills it imperfectly. Rare queries are withheld to protect the identity of searchers. The interface returns a capped number of rows so a long tail is invisible. History is limited and the oldest months drop off. Clicks reported there will never match sessions in your analytics because the two systems count different events at different moments.
What measurement cannot do and what it costs
No tool recovers the search term for an individual visitor. Any product offering keyword level attribution for each session is modelling a guess from aggregate data and presenting it as fact. Rank tracking tools do not fill the gap either because a position on a screen in a data centre says nothing about what happened on your site.
The cost is setup time and discipline afterwards. Conversion events have to be defined and tested. Landing page groups have to be agreed and maintained. Somebody has to export Search Console data regularly if a long history matters. An analytics and measurement setup carries its scope and rate on the service page and professional fees attract GST at 18%. The ongoing cost is the monthly half hour somebody spends actually reading the result.
When this is worth paying for and when it is not
It is worth it when organic search is already a real channel and there is a conversion worth counting. An enquiry form a quotation request a checkout or a phone call all qualify as long as the event can be recorded.
It is not worth it when the site receives a handful of organic visits a day. At that volume the numbers are noise and a careful dashboard will make you confident about conclusions the data cannot support. Build the content first. It is not worth it when nothing on the site constitutes a conversion. Measuring traffic to a brochure site tells you the traffic changed and nothing else. It is also the wrong purchase if the aim is to settle a dispute with an agency. Better measurement will not prove whether past work was good because the baseline was never recorded.
What to measure instead of keywords
Measure landing pages. Search Console tells you which queries reached which page and your analytics tells you what happened once somebody arrived on that page. Joining the two by URL gives a reliable chain from demand to outcome without ever knowing the term for a single visitor.
Group queries by intent rather than reading them one by one. A group for people naming your brand. A group for the problem you solve. A group for comparison and pricing language. The brand group must be separated out and filtered on every variant and misspelling of your name because brand demand moves for reasons that have nothing to do with search work.
Record a baseline before anything changes and keep a dated log of what was published or fixed. Without that log a change in the numbers four months later is unattributable to anybody and the argument about causation cannot be won.
The number that matters and the number that flatters
What matters is non brand clicks and conversions by landing page group measured over months rather than weeks. That combination is slow and it responds to work.
Average position flatters and misleads more than any other figure. It averages every query including ones you never targeted so it worsens whenever a page starts appearing for new searches. That is a good event reported as a bad one. Total sessions flatters because brand traffic sits inside it. Impressions flatter because they rise when Google tests a page and fall when it stops. None of the three should be the headline on a report.
The common mistake
The usual error is judging search work on total sessions or average position because both are on the front screen of the tools. It is expensive in a particular way. It funds work that produced brand traffic and cuts work that was slowly building non brand demand. By the time the mistake surfaces the pages that were working have been abandoned.
A note on consent and the data you are allowed to hold
Measurement sits inside data protection rules. India's data protection framework and the consent notices that come with it mean a share of visitors can decline tracking and that share is not recoverable in any report. The mechanism is simple enough. Where consent is refused the analytics tag does not fire and the visit is absent from your totals rather than wrong in them. How the rules apply to your particular collection and storage is a question for a qualified professional who can review your setup.
This guide explains how something works. It is not legal, tax or financial advice, and rules change. Confirm anything regulatory with a qualified professional before you act on it.