When AI Costs Your Business

"Just See What ChatGPT Says"

This is a sentence I’m hearing more and more from clients and I’ll admit, I’m getting a little tired of hearing it.

Not because I’m against AI. Quite the opposite. At Nova X, we use AI every day across research, content, administration, ideas, processes and problem-solving.

The reality is, by the time we present something to a client, there’s a very good chance we’ve already asked ChatGPT/AI.

We’ve prompted it skilfully. Refined the crap out of the response. Challenged what it gave us. Added context. Removed the generic fluff. Then layered in something AI simply doesn’t have.

Real-world experience and knowledge of the business we’re working with.

So when the finished work is then handed back with a request to “put it through ChatGPT and see what it thinks”, something interesting happens.

AI stops being a productivity tool.

It becomes another person in the approval process.

When AI becomes the second client

This is something I’m increasingly hearing discussed among VAs, outsourced professionals and consultants.

A client engages someone because of their experience, judgement and ability to understand their business.

That person does the work, potentially using AI as one of several tools in the process.

The work gets reviewed by the client > then ChatGPT gets asked for its opinion > The consultant is asked to review ChatGPT’s opinion > revisions are made then someone asks ChatGPT again.

At some point, we have to ask:  Are we actually saving any time here?

Because what was supposed to create efficiency has just introduced another layer of review, discussion and billable time.

And yeah, sometimes it’s a little offensive.

If you’re paying someone for their experience, judgement and knowledge of your business, continually asking them to run their work back through AI is essentially saying:

“Cool. But can the robot double-check you?”

The bigger problem isn’t ChatGPT

This goes much further than consultants being mildly annoyed.

One of the biggest risks I see for businesses right now is unmanaged AI adoption.

AI has entered workplaces incredibly quickly. Employees, contractors, managers and business owners can all be using different AI tools, different accounts and completely different approaches.

• One staff member might use it to draft an email.
• Another might upload a document and ask it to analyse it.
• Someone in marketing might use it to write social content.
• A manager might use it to help make a business decision.
• Meanwhile, nobody has actually established the rules.

What information can be entered?

• What shouldn't be entered?

Which tools are approved?

What needs human review?

How should the business's brand voice be maintained?

When is AI appropriate, and when should someone rely on their own professional judgement?

“We use ChatGPT” is not an AI strategy.

Giving everyone access to AI without establishing how it should be used can create an entirely new category of inefficiency and risk.

Your brand can quickly become AI's brand

There’s another problem we see particularly in marketing and content.

Without strong brand guardrails, AI tends to pull businesses towards the middle.

• The language becomes familiar.
• The sentence structures become familiar.
• The captions start sounding like everyone else’s captions.

Suddenly every business is “delighted to announce”, “passionate about helping” and ready to “elevate” something.

You might be producing more content, but that doesn't necessarily mean you're producing better content.

AI needs context.

It needs examples of how your business communicates. It needs to understand terminology you use and terminology you avoid. It needs to know your audience, services, positioning and boundaries.

Most importantly, someone still needs to know when its answer isn't right.

That judgement doesn't come from a prompt.

It comes from experience.

Poor AI use can cost you money

The promise of AI is largely centred around efficiency.

Do more. Work faster. Reduce repetitive tasks and used properly, it can absolutely deliver that.

But businesses also need to start measuring the time being spent managing AI itself.

If a team member saves 20 minutes drafting something but then spends 45 minutes prompting, checking, rewriting and comparing responses, have you gained productivity?

If a consultant produces a recommendation and is then paid to analyse an alternative recommendation generated by ChatGPT, have you saved money?

If five employees are independently building prompts and teaching AI the same information about your business, is that efficient?

If AI-generated content has to be heavily rewritten because nobody established brand guidelines in the first place, what did the shortcut actually achieve?

There is a cost to poor AI use.

It's just often hidden inside everyone's working day.

Businesses need AI guardrails now

You don't necessarily need a 40-page AI manual.

But businesses do need to start establishing some basic rules around how AI fits into their operations.

• approved AI platforms and accounts

• what business, client and personal information can be entered

• privacy and confidentiality requirements

• brand voice and messaging guidelines

• approved prompts or workflows for repetitive tasks

• where human review is mandatory

• who is responsible for checking AI-generated information

• when employees should use AI and when they should use their own expertise

• how AI use fits into existing systems rather than creating another layer of work

For businesses working with external consultants and contractors, there is another useful principle:

Trust the people you are paying for their expertise.

By all means, ask questions. Challenge recommendations. Ask why someone has taken a particular approach.

But AI shouldn't automatically become the final judge of whether professional work is correct.

We’re not anti-AI. Far from it.

AI has become one of the most useful tools available to small businesses.

We use it.  We encourage appropriate use of it. We will continue building it into the way we work.

But the businesses that benefit most from AI won't necessarily be the ones using it the most.

I suspect they’ll be the businesses that become very good at knowing where it belongs.

So the next time you're about to tell your employee, VA or consultant:

“Just see what ChatGPT says.”

There’s a reasonable chance the answer will be:

We already asked. Now here’s what we think.

Stay well. 

Storm Dawson
Business Owner | Nova X Solutions