You can log into Google Analytics and learn that 1,240 people visited your website last month. You still can’t answer the question your finance director would ask: did any of them become work?
That gap is what this post is about.
What are website analytics, and what they are not
Website analytics is the record of what happens on your website: who arrives, where they came from, which pages they look at, and what they do next. That last part is where most setups quietly fall over.
Think of it like a showroom. Standard analytics is a door counter. It tells you 1,240 people came in. It doesn’t tell you who spoke to a salesperson, what they asked about, or who left a phone number on the way out. The door counter is useful. It is not the same as a sales ledger, and it was never designed to be one.
So when the question is “is the website working”, you need both: the traffic record, and a separate record of enquiries tied back to where each one came from.
Why more traffic doesn’t always mean more enquiries
Why no single dashboard can answer the question
Each tool has documented blind spots. Not bugs. Deliberate design.
Google Search Console hides the actual search terms behind a large share of your clicks. Google’s own rule is that queries not searched by more than a few dozen people over a two to three month window are stripped out of the query table, though those clicks stay in your headline total. That is why your keyword list never adds up to the number at the top. Ahrefs has measured how big the gap is three times: 46.08% of clicks hidden in 2022, 45.02% in April 2024, and 46.77% in April 2025 across 22 billion clicks.
Google Analytics has its own limits. It applies system-defined thresholds that withhold rows with low user counts, and you can’t adjust them. Once a report has more than about 500 unique values a day it starts bundling the smaller ones into a row simply labelled “(other)”, and filters don’t look inside that row.
Then there is consent. If a visitor declines cookies, standard analytics doesn’t count them. Published consent rates vary enormously depending on how the banner is built and how the figure is calculated, so treat any single headline number with suspicion. The safe assumption is that some of your traffic is invisible, and you don’t know which part.
None of this makes the tools bad. It makes any one of them, on its own, an incomplete answer.
Why one tool is never the whole picture
Number of searches the same website was found for in the same period, according to two different sources.
993
16
The six sources we pull, and what each is for
- Google Analytics 4 – behaviour on the site, consented visitors only.
- Google Search Console – what Google showed you for and what got clicked. Google’s own record, so the most reliable search data there is.
- Umami – lightweight, privacy-first site analytics that fills some of the gap left by consent.
- Ubersuggest and DataForSEO – independent rank and competitor checks, run side by side so we can spot where the two disagree.
- Our own lead tracking plugin – server-side, so it records the enquiry itself: which page it came from, which source sent the visitor, and the path they took to get there.
What is website conversion rate tracking and how does it work.
The last one is the piece most businesses are missing, and it is the one that turns a traffic report into a business report.
Same website, two traffic sources, ten times the difference
Share of visits that ended in an enquiry, by where the visitor came from.
13%
1.2%
Visitors arriving from AI tools were roughly ten times more likely to enquire than visitors from Google search. Google still sent far more people, so both matter. But without enquiry-level tracking, the AI traffic looked like a rounding error.
The same data showed what a serious buyer does before they get in touch.
Read More:
What a visit that ends in an enquiry actually looks like
Median time on site and pages viewed, comparing visits that led to an enquiry against engaged visits that did not.
For a British ceramics engineering firm, the question was different: did the rebuild work?
Before and after a rebuild, measured properly
Matched 31 day windows either side of a new website going live, taken from Google Search Console.
The three questions your monthly report should answer
- Who arrived, and from where? Not just a total. Split by source, with AI referrals counted separately.
- What did they do? Which pages carry the enquiries. In the signage example, one quote page converted 27% of the people who reached it, which makes getting more people to that page the single clearest priority.
- Who actually enquired? A number you can check against your inbox and your diary.
If your report answers all three, you can make decisions. If it only answers the first, you have a door counter.
What website analytics are best to track.
What to ask whoever runs your website
- Can you show me enquiries, not just visits?
- Which page did each enquiry come from?
- How many of my visitors are you not counting, and why?
- Where do your figures and Google’s figures disagree, and which do you trust?
Four honest answers to those questions are worth more than any dashboard.
Frequently Asked Questions
What are website analytics in simple terms?
They are the record of what happens on your website: how many people visit, where they came from, which pages they view and what they do. Most standard setups measure visits well and enquiries poorly, which is why a strong traffic month and a quiet enquiry month can happen at the same time.
Is Google Analytics enough on its own?
For behaviour on the site it is a reasonable starting point. On its own it will not tell you what people searched to find you, it will not count visitors who decline cookies, and it will not reliably tie an enquiry back to the page and source that produced it.
Why do Google Analytics and Search Console show different numbers?
They measure different things. Search Console records what happened in Google results, before anyone reaches your site. Analytics records what happened on your site, and only for visitors it is allowed to count. A gap between them is normal and expected, not a sign that something is broken.
Why don’t my Search Console keywords add up to the total clicks?
Google removes search terms that too few people used, to protect those people’s privacy. The clicks still count in your total, but the search term itself is never shown. Across large studies this accounts for close to half of all clicks, so treat the keyword list as a sample rather than a full record.
Do I need cookie consent to run website analytics?
In the UK, non-essential cookies including standard analytics require consent before they are set. That is one reason we run a privacy-first analytics tool alongside Google Analytics, and why enquiry tracking is done server-side rather than depending on a visitor accepting a banner.
How do I know which enquiries came from my website?
You need tracking that records the enquiry event itself, along with the page it came from and the source that sent the visitor. Once that is in place you can reconcile the report against your inbox, which is the only check that really settles the question.
