For a long time, I looked at my website the way most marketers do.
Can Google crawl it?
Are the important pages indexed?
Are the titles clear?
Is the schema in place?
Are the internal links working?
Are the right keywords appearing in the right places?
All of that still matters.
But over the last few months, while auditing websites for AI visibility, I started noticing something uncomfortable.
I was asking other brands questions that I had not fully asked about my own website.
Not:
Can Google find this page?
But:
If someone asks ChatGPT about this business, does the website give it enough information to understand what the company actually does?
And then the harder question:
Would an AI system have enough evidence to recommend it?
That changed the way I looked at my own website.
So I started rebuilding parts of Digital Platform 271.
Not because the site was broken.
Because I realised that being searchable, being understandable, and being recommendable are three different things.
The first problem was clarity
People are good at filling in gaps.
We can read a website and understand that two services are related.
We can interpret vague phrases like “growth solutions” or “digital transformation.”
We can infer who a company is probably trying to serve.
But the more interpretation a website requires, the more ambiguity it creates.
So one of the first things I changed was the service structure.
Instead of presenting several overlapping services, I simplified the core offer around three clear stages:
- AI Visibility Audit
- GEO Foundation
- AI Growth System
Each one solves a different problem.
That sounds like a simple copywriting exercise.
It wasn’t.
It forced me to answer a much more important question:
What exactly do I want this business to be known for?
That is a branding question.
It is also increasingly an AI visibility question.
Then I started building pages around actual questions
I also stopped thinking only in terms of traditional service pages.
Instead, I started looking at the questions someone might actually ask an AI assistant.
For example:
- What is an AI visibility audit?
- How do I check whether my brand appears in ChatGPT?
- What is Generative Engine Optimization?
- How is GEO different from SEO?
- Why can AI understand a brand but still not recommend it?
Those questions started influencing the structure of the website.
One practical example was creating a dedicated AI Visibility Checker page.
The goal was not just to create another page around a keyword.
The goal was to create one clear destination for one clear intent.
If someone is looking for a way to check AI visibility, the page should make it immediately obvious what the tool does and what happens next.
That kind of clarity helps humans.
It also reduces ambiguity for machines.
I stopped hiding the methodology
This was one of the bigger changes in my thinking.
Like many businesses, I originally focused heavily on the outcome.
What the audit does.
What the client receives.
What the score means.
But the more I worked on AI visibility, the more I realised that the process behind a conclusion matters just as much as the conclusion itself.
So I started publishing more of the methodology behind the work.
How the audit is structured.
What signals are evaluated.
How research is conducted.
How AI visibility is measured.
Where the limitations are.
This matters because AI visibility is becoming crowded with very big promises.
“Rank #1 in ChatGPT.”
“Guaranteed AI visibility.”
“Get recommended everywhere.”
I do not think that is a useful way to approach this field.
AI recommendation systems are not simple ranking tables.
If the work is serious, the methodology should be visible too.
The About page became more important than I expected
Most businesses treat the About page as a biography.
I started looking at it differently.
More like an identity page.
Who is behind the company?
What exactly does that person work on?
What topics are they associated with?
Where else can that identity be verified?
What research, profiles, articles or projects connect back to them?
That changed the way I structured my own profile page and the relationship between my personal identity and Digital Platform 271.
I also strengthened the structured data connecting the founder, organisation, services and published content.
None of this replaces good content.
But it removes unnecessary uncertainty.
And uncertainty is something I am increasingly trying to eliminate from every part of the site.
Then came the machine-readable layer
Some of the changes were much less visible.
Schema.
Canonical URLs.
Sitemaps.
Robots directives.
FAQ structure.
Service information.
Organisation and Person data.
Research article markup.
Product and service markup where relevant.
I also added files such as llms.txt.
But this is important:
I do not believe adding an llms.txt file suddenly makes a website AI-optimised.
The same applies to schema.
These are supporting layers.
They can make information easier to interpret or discover.
But they cannot compensate for weak positioning, unclear services or a lack of evidence.
That distinction matters because AI optimisation can very quickly turn into another technical checklist.
I think that would be a mistake.
The biggest lesson came after the technical work
After making these changes, I started testing the website differently.
Not only with SEO tools.
I started asking AI systems the kinds of questions potential clients might actually ask.
Questions like:
- Who provides AI visibility audits in India?
- Which consultants work on GEO?
- How can a D2C brand improve its visibility in ChatGPT?
- Who offers AI visibility analysis?
That is where the difference between being understood and being recommended becomes very clear.
An AI system may understand that your company exists.
It may describe your services correctly.
It may even surface your content.
But that does not mean it will recommend you.
Recommendation is a much higher bar.
Your website can explain who you are. But it cannot be the only one saying you are good.
This may be the most important lesson from everything I have been working on recently.
Your website is first-party evidence.
Naturally, it will say positive things about your business.
The more difficult question is:
What exists outside your website?
Are independent websites mentioning the company?
Are there reviews?
Are there credible profiles?
Interviews?
Research references?
Editorial mentions?
Directories?
Comparisons?
Independent evidence?
I have started paying much more attention to this layer.
Because when an AI system has to choose between several businesses, self-description is only part of the picture.
The web around the brand matters too.
This also changed how I think about AI visibility scores
I still believe scores are useful.
They help identify missing signals.
They create a baseline.
They make audits easier to interpret.
But I no longer think a website score should ever be treated as the final answer.
A technically strong website can still have very little real recommendation presence.
At the same time, another brand can have a weaker website but still appear frequently in recommendations because it has strong category recognition, reviews, third-party mentions and external authority.
That is why I now separate two questions:
Is the website ready to be understood?
and
Is the brand actually being chosen?
Those are not the same test.
So what am I measuring now?
My own website work is no longer just about rankings.
I am watching a wider set of signals.
Does Digital Platform 271 appear when relevant questions are asked?
Does it get mentioned when the brand name is not included in the prompt?
Does it appear in recommendation-style queries?
Which sources does the AI rely on when talking about the company?
Does the answer remain consistent when the same question is tested again?
Are third-party mentions increasing?
Is the brand becoming more strongly associated with AI visibility and GEO?
Some of these things will take time.
That is exactly why I think they are worth tracking.
The way I look at websites has changed
I still care about SEO.
I still care about indexing.
I still care about technical quality.
AI visibility does not replace any of that.
But I no longer stop there.
When I look at a website now, I ask four additional questions:
- Can AI clearly identify what this business does?
- Can it understand who the business is for?
- Can it find enough evidence to trust the claims being made?
- And does it have a reason to recommend this brand instead of another one?
Those questions have changed the way I am building my own website.
They have also changed the way I audit other brands.
And I suspect this is only the beginning.
Because the next stage of search is not simply about being found.
It is about being understood, compared and chosen.
That requires a different way of thinking about your website.