It Revealed Something Bigger About Advisory Firms.

I was at Erewhon in Calabasas when I saw someone who looked familiar.

Late 50s. Maybe early 60s.
Silver-gray hair.
Lean, fit, chiseled jaw.
Law enforcement energy.
Someone came up and shook his hand.

He wasn’t A-list famous.
But he clearly wasn’t anonymous.

He was also with a very tall woman — over six feet — who could have been his wife.

That detail stuck.

I couldn’t place him.

So I ran an experiment.

I opened ChatGPT.

And I gave it almost nothing.

Related: How AI Helps Tax and Accounting Firms Get Found Before the First Call

The First Prompt (Intentionally Vague)

“Late 50s or 60s. Silver hair. Around six feet tall. Chiseled face. Maybe CBS. Maybe detective or western shows.”

Predictably, it suggested:

• Mark Harmon
• Tom Selleck
• Kevin Costner

Reasonable.

All wrong.

But this wasn’t about being right on the first try.

It was about what happened next.

The Power of AI Isn’t in the First Prompt

It’s in the Refinement.

Instead of repeating the question, I added constraints.

• Not A-list.
• More of a utility actor.
• Lean but solid.
• No mustache.
• Familiar, but hard to place.
• Law enforcement vibe.
• Could wear a cowboy hat.
• Someone recognized him.
• Very tall partner with him.

Each detail narrowed the probability set.

AI adjusted.

Eliminated obvious names.
Shifted to long-running procedural actors.
Weighted toward recurring network roles.

And then it landed on:

Neal McDonough.

Correct.

Not because AI guessed.

Because structured prompting works.

Related: The Real Role of AI in Referral-Driven Tax and Accounting Firms

What Actually Happened Under the Hood

This wasn’t magic.

It was:

• Pattern recognition
• Constraint layering
• Probability filtering
• Context stacking
• Iterative refinement

In other words:

Advisory thinking.

This Is Exactly How Advisory Services Should Work

When a client says:

“I think I’m paying too much in taxes.”

That’s the equivalent of:

“Silver hair. Looks familiar.”

It’s vague.

Your job isn’t to answer immediately.

Your job is to refine.

• Entity structure?
• Income mix?
• State exposure?
• Long-term exit plans?
• Ownership splits?
• Liquidity events?
• Risk tolerance?

Each added layer reduces noise.
Each constraint narrows strategy.

Eventually, the right advisory answer emerges.

Not from guessing.

From guided refinement.

Most Firms Use AI Like Google

That’s the mistake.

They type:

“Write a tax blog.”

And accept whatever comes back.

But high-performing firms are starting to understand:

AI gets exponentially better with structure.

Compare:

“Write a blog about tax planning.”

Versus:

“Write a tax planning article for dual-income households in {{city}} earning $400k+ with RSUs and equity compensation, focusing on liquidity planning before IPO.”

The second prompt produces a strategic asset.

Because context compounds.

Why This Matters for Accounting Firms Using AI

AI is not replacing accountants.

It’s amplifying structured thinkers.

The firms that will win are the ones who:

• Ask better questions
• Layer constraints
• Use guided AI workflows
• Refine outputs intentionally
• Build repeatable advisory systems

This is why AI workflow automation inside modern accounting platforms matters.

It’s not about generic output.

It’s about guided refinement.

Inside CountingWorks PRO, MAX doesn’t generate random content.

It narrows outputs based on:

• Niche
• Geography
• Service tier
• Client profile
• Engagement stage
• Advisory focus

The better the inputs, the sharper the output.

Just like identifying an actor.
Just like diagnosing a tax strategy.

The Bigger Lesson

AI is probabilistic.

Precision requires iteration.

Context compounds.

Structure wins.

The future of advisory firms isn’t:

“Who can prepare returns the fastest?”

It’s:

“Who can think in systems, guide AI, and refine strategy with precision?”

That’s where margin lives.

That’s where differentiation lives.

That’s where growth lives.

A Subtle But Important Shift

If you’re experimenting with AI in your firm right now, ask yourself:

Are you prompting?

Or are you refining?

Because there’s a difference.

And that difference determines whether AI produces noise…
Or leverage.

Curious What Structured AI Looks Like in Practice?

If you want to see how guided AI workflows can layer context across:

• Marketing
• Client onboarding
• Advisory positioning
• Proposal automation
• Niche targeting

That’s exactly what we’re building inside CountingWorks PRO.

Because cookie cutter is dead.

Narrative wins.

And precision scales.