How to Replace $200k of Consulting with a $20 ChatGPT Subscription

Let me tell you a story.

Recently, I caught up with some old friends—consultants who had just landed a big project with a large corporate client.

They were charging $200k for an AI transformation program roadmap.

Their job? To help this company deploy AI tools and manage the “people transformation” that comes with it.

I know this story well.

Here’s how it goes:

• They’ll start by interviewing people internally, gathering knowledge that’s already floating around the organization, and formalizing it in endless slides.

• Next, they’ll present a series of hypotheses, forecasted numbers, and action plans, with all the classic elements: timelines, key success factors, stakeholder maps, governance structures.

• They’ll weave it all into a narrative, backed by past client successes and some fancy storytelling.

• Eight weeks later—boom. $200k paid.

And here’s the kicker: this process, while expensive, made sense…once, in the "old world".

But AI flipped the script. We're in a world where one simple tool can do most of this work.

One tool.

Take ChatGPT, for instance. $20 a month. Widely accessible. No hoops, no consultants required.

(There are better ways and tools, but let’s stick to it for the sake of simplicity).

Now, before I lose the consultants reading this, hear me out. I used to be one of you. I spent years in consulting, billing out projects like these. I know the value of expertise.

But the truth is, so much of what we charged for—formalizing knowledge, generating ideas, drafting plans—can be done by anyone who knows their way around an AI tool like ChatGPT.

That’s the reality of the “new world.” You don’t need a stack of tools. You don’t need a library of AI software. You just need focus—and the ability to use one tool effectively.

Let me show you what I mean.

Let's pick a problem

How to Replace a $1,000-a-Day Consultant…With a $20 ChatGPT Subscription

This isn’t just theory. This is a shift that’s already happening.

Let’s take a classic corporate problem:

Let’s say a law firm wants to create a roadmap for adopting AI across their departments.

In the "old world", this would be a huge project. You’d bring in consultants, spend weeks in workshops, formalize internal knowledge, and produce a polished, 100-page report.

(and some are still doing that, paying millions).

This week, I’m breaking down this example for you and showing you what I can do in 20 minutes.

I’ll keep it straightforward—because I know some of you are solo entrepreneurs or leading small teams. I chose a corporate-level problem to illustrate a point, so stick with me.

Most people think ChatGPT is perfect for quick tasks in small businesses but not quite ready for the heavy lifting of high-stakes projects. They believe that for the “big stuff,” you still need McKinsey or Deloitte on speed dial.

But here’s the thing: if ChatGPT can help a large company solve a complex problem, imagine what it could do for your business—with fewer layers, fewer people, and faster decision-making.

Use this example as proof:

$20 a month and a few focused hours can replace $20k in consulting (or agency) fees.

Let’s get into it.

Getting Started with Free ChatGPT

Alright, let me walk you through exactly what I’d do if I had to tackle this problem alone—in just 20 minutes.

I’ll document every step of the process. Not just the highlights, but everything: what worked, what didn’t, and how I adjusted when things fell flat. No cherry-picking.

To start, I’ll use the Golden Circle framework to structure the problem.

The Golden Circle breaks down into three core questions: Why, How, and What. Here’s the prompt I’ll use:

“Act like Simon Sinek reviewing [topic] using the Golden Circle.

Start by diving deep into the heart of [topic]. Answer: Why does it truly matter? What’s its fundamental purpose beyond just profit or function? If [topic] encapsulated a belief, what would that belief be?

Once the ‘Why’ is clear, transition to understanding how [topic] brings this purpose to life. Which unique actions, methods, or processes ensure its core is realized? How does it distinctively stand apart in this journey?

After grasping the ‘How’, proceed to describe ‘What’ [topic] really is. Detail the product, service, or idea, ensuring its alignment with the previously defined ‘Why’ and ‘How’.

Wrap up your exploration by reflecting on the Golden Circle’s resonance. Consider how leading with ‘Why’ can cultivate trust, loyalty, and genuine enthusiasm among those who connect with [topic].”

For this example, I’ll feed ChatGPT with the topic: “change management and AI adoption in my company.”

First, I’ll run this through GPT-4o (the free version) to see what comes out.

Alright, we’ve got some interesting output, but there’s a lot of fluff—generic statements, broad ideas, nothing concrete.

It’s like a bad consultant’s first pass: decent overview, but not actionable.

(believe me, some consultants can sell this s***!!)

So, let’s refine. I’ll add more context about the company to see if the same LLM can produce something sharper.

Here’s the next prompt:

“Act as a senior management consultant hired to accelerate our AI adoption.

Context: We’re a law firm with 4,000 employees across multiple cities. Our central office has 2,000 staff, with the remaining 2,000 spread across six locations. We’re deploying a single AI tool company-wide.

Key considerations:

1. Is a one-size-fits-all AI approach optimal for our diverse workforce?

2. How can we ensure consistent AI implementation across different locations?

3. What specific AI tools would benefit a law firm most?

4. How do we measure the ROI of our AI adoption?

5. What training programs should we implement to maximize AI utilization?

Refine your initial analysis. Provide a concise, actionable report addressing these points. Remove any fluff. Every sentence should provide actionable value.”

Humm...Still getting a lot of generalities.

It’s more focused now, but it’s still dancing around specifics.

Alright, let’s try running the same prompt through o1-preview, which has better reasoning capabilities.

Going Deeper with O1

Alright, let’s start with the Golden Circle Prompt again—this time using “o1-preview” (the paid version) instead of GPT-4 free.

Much better. We’re getting more depth, more actionable insights. Here’s what stands out:

• It recognizes that establishing a strong “why” for employees is crucial—not just pushing tech for tech’s sake.

• It highlights that this isn’t just a technology problem; it’s about culture, too. Giving people flexibility in tool choice can drive engagement.

• It gets that adoption shouldn’t be a rigid, top-down process.

• And it emphasizes inclusivity in the approach.

Among many other valuable insights (try it).

Honestly? This is a solid foundation for just a few seconds of AI-generated brainstorming.

In the “old world,” you could easily pay a few thousand dollars for someone to package this up with polished slides and bullet points.

But let’s take it deeper. Now that we have a rough framework, let’s get specific. I’ll run the same second prompt through "o1-preview" to see what happens.

“Act as a senior management consultant hired to accelerate our AI adoption…”

Not bad at all. This draft actually has some solid thinking:

• It questions the “one-size-fits-all” approach—something a lot of companies ignore when rolling out a single tool to thousands of employees.

• It raises the issue of measuring “adoption,” which is deceptively complex.

• It touches on ROI and KPIs—critical if we want to track real impact.

Now, we’ve got key insights and ideas. But ideas are just the start.

Execution is everything.

Let's ask for an execution plan. We want to start tomorrow morning.

We Want Execution Over Ideas: Action Plan

Let’s ask ChatGPT to generate a concrete action plan—something we can start doing tomorrow.

Here’s the prompt:

“Formalize an action plan for the next 3 months with SMART KPIs and achievable objectives. Generate key success factors. Determine the number of people we need to involve, their responsibilities, and create a clear 12-week project timeline in a table format. Be specific to our law firm and avoid generalities or non-actionable advice.”

Honestly, we’re starting to see something that could pass as a “first draft” from McKinsey. You could take this, run it through a tool like Gamma(.)app, and have a polished presentation ready for stakeholders.

Is it perfect? No—it needs refinement and adjustments. But by listing the flaws and challenging the output, you’re already well ahead. Here’s what we’ve got in just a few minutes:

• SMART objectives with KPIs that are actually achievable.

• Key success factors that give you a clear focus.

• Operational plan that’s ready to share and get feedback on.

• Initial staffing requirements—who you need, what their roles are.

• Defined responsibilities within the team.

• Timeline broken down week by week for the next 3 months.

In a corporate setting, I’ve seen teams of 2-3 people take days to produce something comparable. Here, we’re getting a draft in under 10 minutes, plus another few minutes to generate presentation slides with Gamma(.)app.

Now, let’s take it deeper.

Next prompt: “Calculate the potential ROI in dollars. Explain your hypotheses and formulas. Generate a chart illustrating: the evolution of costs versus gains, the share of gains per category, and the share of estimated costs per category. Include any other relevant projections that might help leadership make informed decisions.”

Okay, the output looks good, but let’s push back on some of the assumptions.

Follow-up prompt: “The projected gains seem overly optimistic. Your assumptions include a high adoption rate and rapid AI proficiency among users. Please factor in a realistic learning curve and provide more conservative estimates for potential gains.”

The updated assumptions and projections are solid—conservative, grounded, and consistent. Of course, no AI tool gives you a flawless strategy on the first pass. You’ll still need to challenge the assumptions, upload any existing data or calculations you already have, and refine it further.

But in just a few minutes, we’re getting realistic estimations, ROI projections, and a clear value measurement framework.

So now we have a plan and estimations. Next, let’s bring it to life with visuals.

We can generate charts, graphs, and slides to present this in a way that’s easy to digest.

We Want Graphics and Visuals

We’ve got solid information, but to make it digestible, we need charts and visuals. Let’s get ChatGPT to generate some.

Here’s the prompt:

“Generate a chart illustrating: the evolution of costs versus gains, the share of gains per category, and the share of estimated costs per category. Include any other relevant projections that might help leadership make informed decisions. Generate and execute Python code to create the charts. Use the data generated above.”

For best results, I strongly recommend using “o1-preview” to generate the code first.

If you don’t understand Python code, don’t worry. You don’t need to.

I’m keeping this at “one-shot” level here, just to give you a quick sense of what’s possible. But for finer results, you could ask ChatGPT to tweak the charts—specify different styles, adjust labels, or focus on particular data points.

Now let’s switch to GPT-4o (the free version) to actually run the code (since o1-preview doesn’t execute code).

And there you have it—beautiful, clear visuals. You’ve got charts showing costs vs. gains, cost breakdowns, ROI projections, and more.

Drop these right into your presentation deck (or use Gamma(.)app to generate your deck in a few minutes)

I pulled this off in about 20 minutes. With focused effort, you could generate a world-class and sharper action plan, complete with visuals and other elements, in a few hours or days.

But Consulting is About “Expert Knowledge,” Isn’t It?

Now, everything we’ve done so far is just using the language model as-is.

Some people might push back: “McKinsey or BCG charge $1k a day for a junior consultant because they’re giving you their firm’s knowledge and expertise.”

Fair enough.

But if you want that knowledge and expertise? That’s easy. Here’s how you can enhance everything we just built—integrating insights that go beyond what even a Consulting Partner might bring.

Step 1: Pull in Expertise from Trusted Sources

Big consulting firms love sharing their knowledge. Head over to McKinsey, BCG, or Bain’s websites and look for their latest research on your topic.

For example:

These reports are goldmines. They cover trends, challenges, and specific benchmarks that these firms use to guide clients. And if you’re working in a specialized field, like law, these insights are invaluable.

Step 2: Add Internal Knowledge and Past Analysis

If your organization has internal documents, previous analysis, or reports from past projects, bring those in too. This context can help the AI generate more tailored and relevant outputs.

Step 3: Tap into Real-Time AI Research

For more up-to-date expert insights, you can turn to tools like Perplexity AI or ChatGPT’s web browsing feature.

Try this:

Or use ChatGPT with browsing enabled to pull in sharp, recent information.

For example, let’s say you find a benchmark report on ROI in a similar firm. Now you’re not just guessing at potential gains—you have real numbers to compare against. And you can access the original source to validate everything.

Step 4: Organize Your Knowledge Files

Once you’ve gathered all these resources, put them into one or two files. Structure them by topic—this makes it easier for the AI to pull in the right context.

Drag and drop these files into ChatGPT, and tell it to use and reference these materials as it builds out your project.

This approach gives you the best of both worlds: AI-driven efficiency and consulting-level expertise.

Get creative with your own knowledge and experience, too. With the right setup, you’re combining the strengths of AI with the insights of top consulting firms—at a fraction of the time and cost.

And What About Interviews and Gathering Data?

Do you want to pay a consultant $1,000 a day to gather information from people in your company?

I don’t think so.

Some might argue that consultants are valuable because they spend hours in meetings, interviewing employees to “refine the organization’s understanding” through a bottom-up approach. But here’s the thing—AI can handle that, too. And it’s way more efficient.

If you want to go the expert route, you can design a dynamic interview system: a conversational AI that adapts its questions based on responses. (That’s advanced stuff, so let’s keep it simple for now.)

For a quick start, just ask AI to generate static questionnaires.

Here’s how:

Switch to "o1-preview" and prompt it with this:

“To enrich our insights, let’s incorporate a bottom-up approach. Design questionnaires for different categories of employees to gather their perspectives on this project. Suggest relevant categories and create a tailored questionnaire for each.”

Now you have tailored questionnaires for each role—no endless meetings required.

Create the forms in MS Forms (or any survey tool of your choice) and send them out.

Once the responses come in, just export the results to an Excel file.

Then, drag and drop the Excel file into ChatGPT.

That’s it! You’re ready to analyze, iterate, and move forward—faster than any traditional consulting process. Why wouldn't you do that?

If you’re a consultant reading this, here’s your takeaway: structure your knowledge and sell it as a product. The real value you can provide isn’t in gathering data—it’s in automating the steps I’ve described here.

Work differently. It’s time.

Welcome to the New World

People ask me all the time, “What’s the best AI tool for my business?”

The answer? It’s not about the tool. It’s about how you use it.

Pick one. Learn it inside-out. Master it.

Because right now, we’re in a world where $20 a month can give you a level of leverage we never dreamed of.

And the best part? It’s only just begun.

Stop waiting for the perfect stack of  "AI tools." You need one tool, a clear goal, and the willingness to get your hands dirty.

Welcome to the new world.

If you're a business reading me and you have this kind of challenge, message me; I've got something for you.

That’s all for this one.

—Charafeddine