Stop Watching Rankings. Watch Impressions and Enquiries

Most SEO reports arrive looking like a share price. Keyword up two, keyword down four, this one jumped to position three on Tuesday and fell to nine by Friday. The owner reads it, cannot tell whether that is good or bad, and files it. Three months later nobody can say whether the money did anything.

The problem is not the reporting software. The problem is that a daily keyword position is a measurement of something that no longer exists in a single fixed form. You are tracking a number that is partly about you and largely about everything else.

Why a daily position is almost meaningless

Think about what a rank tracker actually does. It sends an automated search to Google, pretending to be a person in a particular place, and records where your site appears. Every part of that sentence introduces noise.

Location. A search for "CA firm near me" from Laxmi Nagar and the same search from Noida Sector 62 return different results. Your tracker picks one point on the map. Your customers are scattered across dozens.
Device. Mobile and desktop results differ, and mobile results differ again depending on how much map pack sits above the normal listings.
Personalisation and history. A logged-in user who has visited your site before may see you higher than a stranger does.
The shape of the page. Position four used to mean four links down. Now position four might sit below an AI Overview, a map pack, a People Also Ask block and a shopping row — so it is somewhere near the bottom of the first screenful, or well below it.
Google's own testing. Google runs experiments continuously and confirms it publishes thousands of changes a year. Some of what you see on a Tuesday is simply a test you were included in.

So when a report says you moved from six to four, you genuinely do not know whether anything changed. And there is a deeper issue: nobody has ever paid you because you were in position four. They paid you because they saw you, clicked, read something convincing, and picked up the phone. Measure that chain instead.

The three numbers that actually matter

You need three layers, in this order. Each one answers a different question, and if the bottom layer is broken the top layer cannot be fixed by more SEO.

1. Impressions — are you being shown at all?

Google Search Console, which is free and takes ten minutes to verify, records every time a page of yours appeared in search results, what the query was, and whether anyone clicked. That is the real exposure number. It aggregates across every location, device and personalised variation, which is exactly what a rank tracker cannot do.

Impressions rising while clicks stay flat tells you something specific: Google is showing you for more searches, but your title and description are not persuading anyone. That is a copy problem, not a ranking problem, and it is cheap to fix.

Two honest caveats about Search Console. It holds sixteen months of data, so export quarterly if you want longer-term comparison. And it omits some rare queries to protect user privacy, so your query list will never add up perfectly to your totals. Neither of these stops it being the best free instrument you have.

2. Clicks and arrivals — are they coming?

Clicks in Search Console and sessions in your analytics will never match exactly. Do not spend an afternoon reconciling them. Pick one as your official number — Search Console clicks is the cleaner choice for organic search — and stay with it.

What matters more than the total is the split. Group your queries roughly into brand (people searching your business name) and non-brand (people searching for the service). Brand clicks rising mostly means your other marketing is working. Non-brand clicks rising is the thing SEO is supposed to produce. Agencies that report only the combined figure are often hiding a flat non-brand line behind a busy social campaign.

3. Enquiries — did anything happen?

This is the only number that pays salaries, and it is the one most small sites cannot measure. If your enquiry form does not record which page the person came from, and your phone number is a plain text line on the page, you are flying blind.

The minimum useful setup:

A hidden field on every form that captures the page URL and, if present, the referring source.
Every form submission creating a record somewhere you can count — a spreadsheet, a shared inbox with labels, or properly, a CRM where enquiries are logged and tagged.
One question in your intake script: "How did you find us?" Imperfect, people forget, but over fifty enquiries the pattern is real.
WhatsApp click-to-chat links with a source tag, so a tap from the website is distinguishable from a tap from Instagram.

On call tracking numbers: they work, but be careful. If you place a tracking number on your website that differs from the number on your Google Business Profile and your listings, you create an inconsistency that local search generally punishes. If you use call tracking, use dynamic number insertion that leaves the number unchanged for crawlers, or do not use it at all.

What I would not pay for

Plenty of things are sold in this category that measure nothing useful.

Daily rank-tracking dashboards. A login with sixty keywords updating every morning feels like value and generates panic. If you track positions at all — and there is a limited case for it, explained below — look monthly, on a small set, and never react to a single week.

"Number of keywords ranking" as a headline metric. Tools estimate this, and the estimates swing wildly when the tool updates its database. A business can gain four hundred "ranking keywords" and not a single additional enquiry, because the new ones are all position 73 for phrases nobody searches.

Screenshots of a number one position. Taken in one location, on one device, possibly while logged in. It proves nothing and it is trivial to produce.

Domain Authority, Domain Rating and similar scores. These are third-party inventions, not Google metrics. Useful as a rough comparison between two link sources. Useless as a monthly progress report, and worse than useless when an agency sells you links to raise it.

Traffic as the goal. A blog post that brings two thousand readers who will never buy from you is a cost, not an asset. We have written elsewhere about matching the page to the kind of search behind it — a page that ranks for the wrong intent inflates your chart and empties your diary.

How to read the numbers without fooling yourself

The discipline matters as much as the metrics.

Compare 28 days to the same 28 days last year, not to last month. Almost every Indian business is seasonal in some way — exam cycles, wedding season, the long monsoon lull, the pre-Diwali rush, the dead fortnight around it. Month-on-month comparison mostly measures the calendar. Year-on-year measures you.

Allow a reporting lag. Search Console data typically settles two to three days behind. Do not check it this morning for yesterday.

Set a floor before you believe a change. If a page gets forty impressions a month, a move from two clicks to four is not a sixty-six per cent gain, it is noise. Only treat movement as real on pages with enough volume for the movement to mean something.

Keep a change log. One line per change, dated: rewrote the service page, fixed the slow images, added the location pages, switched hosting. When something moves in month four, you want to know what you did in month two. This single habit explains more than any dashboard.

A simple monthly scoreboard

Layer	Number	Source	What a fall means
Visibility	Non-brand impressions	Search Console	Indexing, technical or competitive problem
Appeal	Non-brand clicks and click-through rate	Search Console	Titles, descriptions, or a new SERP feature above you
Persuasion	Enquiries from organic	Your CRM or inbox	Page content, speed, trust signals, or the form itself
Local	Calls and direction requests	Google Business Profile performance	Profile, reviews, or map pack competition

Four rows. If visibility is up and enquiries are flat, the SEO is working and the website is not — which usually points at page speed, clarity and the enquiry path rather than at more content. If visibility is flat and nothing else moved, the SEO is not working and you should ask why before you renew.

Where rankings still have a place

Not never. Positions are a useful diagnostic for a small, hand-picked set of terms that genuinely drive your business — the five or ten phrases where being seen is the whole game. Checked monthly, they help you understand why impressions moved. They are a reason, not a result.

And there is a newer category nobody can measure well yet. When someone asks ChatGPT, Gemini or Perplexity for a recommendation, there is no position, often no click, and usually no entry in your analytics at all. Server logs will tell you whether AI crawlers are fetching your pages. Beyond that, honest answer: the measurement tooling for visibility inside AI answers is immature, the figures sold as "AI share of voice" rest on small samples of prompts, and anyone presenting you a confident monthly AI ranking is presenting you an estimate dressed as data. Treat it as a direction to work in, not a scoreboard to report on.

What to do this week

Three steps, in order, none of which need an agency.

Verify Google Search Console if you have not, and export the last sixteen months of query data to a spreadsheet so you own the history.
Add a hidden page-URL field to every enquiry form and start tagging enquiries by source, even if the tagging is a label in Gmail.
Build the four-row table above for last month and the same month last year. Look at it for five minutes. You will learn more than from a year of rank reports.

If the numbers are there but you cannot tell what they are saying, or your forms are not recording anything useful in the first place, tell us what you are trying to measure and we will tell you honestly whether the data exists to measure it. Sometimes it does not, and that is worth knowing before you commission anything.

Questions we get asked

How long before SEO work shows up in the numbers?

Nobody can promise a timeline, and anyone who does is guessing. What you can watch for is sequence: impressions usually move before clicks, and clicks before enquiries. If impressions for non-brand searches have not moved at all after several months of work, that is the signal to ask hard questions, regardless of what the ranking report says.

Why do Search Console clicks and Google Analytics sessions not match?

They count different things. Search Console counts a click on a search result; analytics counts a page actually loading with its tracking script running. Ad blockers, people leaving before the page loads, and consent banners all create a gap. Pick one as your official number and stop reconciling them.

Is average position in Search Console reliable?

It is more honest than a rank tracker because it averages across real searches from real locations, but it still needs care. It reports the average topmost position your site appeared at for that query, so a page that sometimes ranks second and sometimes thirtieth shows a misleading middle. Use it to spot direction, not to report a precise figure.

Should I pay for a rank-tracking tool at all?

For most small businesses, no — Search Console gives you better data free. A tracker earns its cost if you are in a tightly contested local category and need to see map pack positions from several points across a city. Even then, check it monthly on a short keyword list, not daily on sixty terms.

How do I measure whether AI tools like ChatGPT are sending me business?

Partly you cannot, and it is better to say so. Some AI tools pass a referrer you can see in analytics, many do not, and a person who reads your name in an AI answer may simply search for you afterwards, which then looks like brand traffic. Rising brand searches with no other obvious cause is currently the most useful indirect hint available.