
AI isn’t underperforming in accounting firms.
It’s being under-directed.
That’s different.
Most firms open ChatGPT and type something like:
“Write me a tax blog about S-corps.”
AI produces something decent.
Technically correct.
Nicely formatted.
Totally usable.
Also completely interchangeable with what 1,000 other firms just generated.
And then the conclusion becomes:
“AI is fine, but it’s not amazing.”
Of course it’s not amazing.
You asked it for average.
The Real Problem
Most professionals are using AI like Google.
They expect:
- A clean input.
- A clean answer.
- One step.
- Done.
But AI doesn’t work like search.
It works like probability.
And probability sharpens when constraints increase.
Related: Your Website Is Your First Interview. Are You Getting the Job?
Let Me Show You What I Mean
Let’s take that same S-corp example.
Version one:
“Write a blog about tax planning for S-corps.”
Now let’s layer context:
“Write a tax planning article for profitable S-corp owners in {{city}} earning $500k+, focused on reducing reasonable compensation exposure while preparing for a future sale within 5–7 years.”
Same AI.
Different universe.
The first prompt creates content.
The second creates positioning.
The first is compliance marketing.
The second is advisory marketing.
That difference is thinking.
Not technology.
Where Firms Get Stuck
Here’s what I see happening.
A firm tries AI for three things:
- Writing blogs
- Drafting emails
- Summarizing documents
They don’t refine.
They don’t layer.
They don’t build systems around it.
So they conclude:
“It’s helpful, but it’s not game-changing.”
That’s like walking into a gym, lifting a 10-pound weight once, and deciding fitness doesn’t work.
The leverage isn’t in the first rep.
It’s in the structured repetition.
Related: Capital Advisory Is the Next Evolution of the Growth Minded Accountant
This Is Exactly How Advisory Should Work
Let’s shift to a client scenario.
A business owner says:
“I think I’m paying too much in taxes.”
That’s vague.
That’s:
“Silver hair. Looks familiar.”
A compliance mindset responds by reviewing the return.
An advisory mindset starts narrowing.
What’s the entity structure?
What’s the income mix?
Is there reasonable compensation pressure?
Are distributions optimized?
Is the spouse earning income?
What state exposure exists?
Are there exit plans?
Is QSBS in play?
Is real estate carved out?
Every question eliminates half the possible strategies.
That’s constraint layering.
That’s structured narrowing.
That’s how real advisory value is created.
And it’s exactly how AI gets sharper.

AI Doesn’t Replace Judgment
It Rewards Structured Thinkers
Here’s another example.
Prompt:
“Rewrite my homepage.”
Output?
Generic accounting firm language.
Now try:
“Rewrite my homepage for a boutique advisory firm serving dentists in {{city}}, focusing on practice acquisition, cash flow optimization, and exit planning within 10 years.”
Now the output has teeth.
AI didn’t get smarter.
The thinking did.
Most Firms Are Prompting.
Very Few Are Refining.
The difference matters.
Prompting is asking.
Refining is narrowing.
The firms that will win with AI will:
- Build vertical-specific prompt stacks.
- Layer geography and income tiers.
- Connect marketing prompts to advisory workflows.
- Turn one-off requests into repeatable systems.
That’s not automation.
That’s leverage.
Why This Matters Right Now
The market is getting noisier.
AI makes it easier to produce:
- Blogs
- Emails
- Landing pages
- Proposals
Which means generic output will explode.
The only firms that stand out will be the ones who:
- Think in constraints.
- Operate in systems.
- Narrow precisely.
- Build narrative around specificity.
Generic prompts create generic firms.
Structured prompts create differentiated positioning.
The Bigger Shift
We’re not moving into an era where:
“The firm with the best software wins.”
We’re moving into an era where:
“The firm that thinks most clearly wins.”
AI exposes weak thinking.
It amplifies structured thinking.
And that’s why some firms will see marginal gains from AI…
While others will see exponential leverage.
Same tools.
Different mindset.
If You’re Experimenting with AI Right Now
Don’t ask:
“Is AI good?”
Ask:
“Am I narrowing enough?”
Because the advantage isn’t in using AI.
It’s in guiding it.
And that’s a skill.
One that compounds.
Move into the new age of online growth with CountingWorks PRO here.







