What Can Your Website Tell Me Before I See Your Analytics?
When analysing a prospective client’s website, I often don’t have access to their Google Search Console, Analytics, CRM or sales figures.
Does that mean there’s no evidence to work with?
Far from it.
It means we need to be disciplined about what we know, what we can reasonably infer, and what remains a hypothesis.
And perhaps that’s not a bad discipline for an audit to have.
Start by Understanding the Business
Before worrying about rankings, keywords or conversion rates, there’s a more fundamental question:
What does this website tell us about the business behind it?
We can investigate:
- What does the business appear to sell or provide?
- Which products or services receive the greatest emphasis?
- What locations or markets does it serve?
- What differentiates its proposition?
- What actions does it want visitors to take?
None of this necessarily tells us what the business actually considers most commercially important. That’s something we’ll eventually want the people running the business to confirm.
But it gives us an evidence-based starting point.
Understand the Person, Topic and Page
A website doesn’t communicate with an abstract “user”.
Its language, propositions, imagery, calls to action and journeys inevitably speak more strongly to some people than others.
So we can ask:
Who does this website appear to be designed to help?
What problem might have brought that person here? What are they trying to accomplish? What concerns or constraints might prevent them from progressing?
Then we can look at what the site actually talks about.
Not simply:
Which keywords appear in the title tag?
But:
What topics does this website consistently communicate?
A crawl can show how topics are distributed through pages, headings, body content, internal links, supporting content and site architecture.
Then comes another important question:
What job does each important page appear to perform?
A service page, FAQ, guide and About page may all contribute to the same eventual business outcome while doing very different jobs.
One might introduce a service. Another establishes trust. Another answers an objection. Another captures early-stage discovery. Another is intended to generate an enquiry.
Understanding those relationships gives us something considerably richer than a collection of SEO checks:
Business → Audience → Topic → Page Purpose
Now we can start interpreting evidence in context.
Then Look Outside the Website
A website doesn’t exist in isolation.
Once we’ve formed hypotheses about its audiences and important topics, we can investigate the search landscape around them.
What searches appear around those needs?
What related questions are people asking?
What language do searchers use?
Which businesses and pages repeatedly appear?
What kinds of pages does Google surface?
And, does the website we’re auditing appear within that landscape?
Tools such as AnswerThePublic can help expose questions, comparisons and other search variations around a topic. Search-result analysis can then provide observable evidence about what appears for selected searches.
Suddenly, an audit without Search Console isn’t simply:
“You have 37 missing meta descriptions.”
Or:
“You’re not on page one.”
We’re developing a much more useful chain:
Business → Audience → Topic → Site Evidence → Search Landscape → Hypothesis
Your Website Has Topics.
Does Google See the Same Ones?
Suppose a website strongly communicates a particular topic.
Its architecture reinforces it. Several pages support it. The business appears to position it as commercially important.
Now look at the search landscape.
Does the website appear for searches associated with that topic?
If not, what does?
Which competitors repeatedly appear?
What topics and types of pages are being surfaced instead?
What related questions are people asking?
That doesn’t automatically tell us why the website isn’t visible.
But it gives us something worth investigating.
And that’s an important distinction.
Know What You Don’t Know — But It Was Always There
An audit doesn’t create the evidence. It helps us notice what’s already there.
Without Search Console data, I cannot legitimately conclude:
“This page isn’t getting impressions.”
I don’t have that evidence.
I might instead say:
“Across the searches we observed, this site had little visibility.”
Likewise, without Analytics I shouldn’t confidently declare:
“Visitors aren’t engaging with this page.”
I might observe:
“The page presents several potential UX or conversion obstacles that should be investigated against behavioural evidence.”
And without CRM, enquiry or sales data, I certainly shouldn’t announce that SEO is responsible for poor sales.
An evidence-led audit should distinguish:
Evidence → Inference → Hypothesis
An automated audit becoming more sophisticated doesn’t give it permission to become more certain than its evidence allows.
Then Let the Business Challenge the Audit
This approach isn’t only useful for prospective clients.
Perhaps the website led us to infer that Topic A was strategically important.
Then the business tells us:
“Actually, Topic B produces most of our valuable work.”
Excellent.
We’ve learned something.
That’s the point. Like an experiment, an audit hypothesis doesn’t have to be right to be useful. If the evidence disproves it, we’ve still learned something that improves the next decision.
Perhaps our external search analysis suggested weak visibility. Search Console later reveals substantial impressions but poor click-through.
Again, we’ve learned something.
Perhaps the site’s language appeared directed towards one audience, while the business tells us another generates most of its revenue.
That’s evidence challenging an assumption.
With access to first-party evidence, our understanding becomes progressively richer:
Business Understanding → Site Evidence → Search Landscape → Search Console → Analytics → Business Outcomes
Each layer doesn’t simply give us more data.
It gives us another opportunity to test whether our previous interpretation was actually correct.
An Audit Shouldn’t Pretend to Know More Than It Does
There’s plenty we can learn about a business before seeing its analytics.
We can investigate what the website appears to sell, whom it appears designed to help, what topics it communicates, what jobs its pages perform, how those pages work together and how those topics appear within the wider search landscape.
That’s enough to identify opportunities and form useful hypotheses.
But hypotheses are not facts.
The objective shouldn’t be to make an automated audit sound impressively certain.
It should be to gather enough evidence to ask better questions.
Then, when the business contributes its knowledge and first-party data becomes available, those questions, and our answers, can become better still.
Because perhaps the first job of an evidence-led audit isn’t to tell a business everything that’s wrong with its website.
Perhaps it’s to understand the business well enough to know what’s actually worth investigating.
Part of the Wider Discussion
This builds on two ideas I’ve explored recently:
The Discipline of Not Optimising – why evidence sometimes tells us to leave something alone &
Your Website Doesn’t Have “Users.” It Has People – why understanding the person behind the data changes how we interpret search and performance.
Further Reading
Google Search Console’s Search Analytics documentation explains the first-party search data available through Search Console, including queries, pages, clicks, impressions, CTR and position.
AnswerThePublic’s Search Listening material provides a useful perspective on using search questions and modifiers to uncover people’s needs, concerns, comparisons and stages of their journey.
🔗 Follow AWE on LinkedIn: https://www.linkedin.com/company/awe-conversion
🔹 Follow the discussion on Facebook for daily AWE strategies: https://www.facebook.com/awe.conversion