AI Isn’t About Age. It’s About Domain Expertise

Follow me on LinkedIn - AI, GA4, BigQuery

Most of the AI influencers you see today are in their 20s. For many, it’s their first career. They’ve got a 2–3 year head start, and they know the tools inside out.

But here’s the catch. They don’t have deep domain expertise.

They build automation for any industry or niche and lean heavily on their clients for context and domain expertise.

And that’s not a knock on them. They simply haven’t lived and worked long enough to develop deep domain expertise.

If you’re in your 30s, 40s, 50s or beyond, you might wonder if you’ve missed the AI gold rush.

The truth is, you are right on time.

Your real competition isn’t age, it’s mindset.

20-year-olds may have an early-mover advantage. But you have something far more valuable.

You have years/decades of pattern recognition, critical thinking, and context that can’t be duplicated overnight.

Even if AI automates 100% of your job (whether it’s SEO audits, campaign optimization, or analytics reporting) tomorrow, the mental toolkit you have built, knowing what to ask, why it matters, and how to evaluate the output, is priceless.

AI is just a power tool. You’re still the architect.

So,

Don’t think of switching careers into AI as starting over. Think of it as leverage. You are not discarding your expertise; you’re multiplying it.

If you know how to diagnose why a campaign isn’t working, why a landing page doesn’t convert, or why a dataset looks off, congratulations, you already have the foundation of advanced prompt engineering.

What to ask and how to ask come from domain expertise, which in turn stems from experience.

For example,

The majority of GA4 users don’t understand that they can not benefit from AI (chatbots, LLMs, MCP) for their data analysis requirements unless they already have a DEEP understanding of GA4 event schema, dimensions, metrics and scopes.

The lack of such knowledge will always reflect in their prompts (aka questions).

AI tools don’t magically “fill in the blanks” if the user’s question is vague, incomplete, or based on incorrect assumptions.

In GA4’s case, without knowing how events, parameters, dimensions, metrics, and scopes actually work, you can’t even frame the right question for AI to give you a useful, actionable answer.

It’s a bit like hiring a master chef but only telling them, “Cook me something nice”, without mentioning your ingredients, dietary needs, or cuisine preference. They’ll try, but the result may not be what you expect.

The following is an example of a bad prompt because of a lack of domain expertise:

“Show me sales trends.”

Why it fails:

>> Doesn’t specify the GA4 metric (purchase_revenue, ecommerce_purchases, etc.).

>> No dimension to break trends by (time, region, device, campaign).

>> No scope (are we looking at event-level purchases, sessions, or user cohorts?).

>> No date range or filters.

The following is an example of a good prompt because of domain expertise:

Using GA4 data, show me a month-over-month trend of ‘purchase_revenue’ for the last 12 months, segmented by deviceCategory, and include only sessions where country equals "United States".

Why it works:

>> Names a specific metric: purchase_revenue.

>> Specifies a time dimension: month-over-month for 12 months.

>> Adds a breakdown dimension: deviceCategory.

>> Uses a filter: country = United States.

>> Context: Assumes GA4 event schema and naming conventions.

The same logic applies to GA4 BigQuery as well.

If you don’t understand the GA4 BigQuery schema, you won’t be able to use AI to query the data correctly.

That’s why, even with my GA4 BigQuery Composer (a custom ChatGPT I developed for automating SQL generation for GA4 BigQuery), I always teach students the schema first so they can customise the output if needed.

You need domain expertise to do effective prompt engineering in any field.

I recently created the following video using a five-pages long text prompt.

The prompt didn’t just say “make a cool matrix movie style video.”

I have to map out characters, sound effects, visual effects, and frame-by-frame progression.

I have to specify technical parameters like resolution, frame rate, and style, the kind of details you only know if you’ve worked/studied video production or VFX.

That’s the point.

Prompt engineering isn’t just about writing instructions in plain English. It’s about knowing what matters in your field.

AI is the tool, but domain expertise is what makes the output extraordinary.

Today, most AI developers are generalists, building across multiple industries.

But as AI adoption deepens, demand is already shifting toward domain-specific solutions (e.g., AI for healthcare compliance, AI for real estate, legal contract review, financial risk modeling....).

In these contexts, technical ability alone isn’t enough.

Domain expertise will always matter.