Most businesses like yours are using AI shallowly: drafting emails, summarizing calls, brainstorming in a chat window. A much smaller group uses it structurally, with AI running whole workflows, powering client-facing tools, and making the business visible to AI engines. The two groups own the same subscriptions. They get completely different results.
MIT's 2025 GenAI Divide research put numbers on the gap: about 95% of AI pilots stall with no measurable return, while the roughly 5% that succeed embed AI deep into specific workflows instead of layering generic tools over old processes. The encouraging part is that the winning pattern, one pain point executed well, is more available to a small business than to any enterprise.
- Most businesses use AI shallowly, drafting and summarizing in a chat window, which is why most report no measurable return.
- MIT's GenAI Divide research found about 95% of AI pilots stall, while roughly 5% produce rapid, measurable results.
- The successful 5% embed AI inside specific workflows rather than adding generic tools on top of old processes.
- Consultants, coaches, and advisors use AI structurally for intake, proposals, content systems, and getting recommended by AI engines.
- One pain point executed well beats ten experiments, and a small business can rebuild one workflow faster than any enterprise.
Skip the AI Course. Get It Installed.
The AI Native Activation is one working session. You leave with AI installed on your machine, loaded with your business, and producing real work the same day.
See the AI Native ActivationRather talk it through? Book a Rapid Transformation Call.
What are most business owners actually doing with AI right now?
Most owners are using AI as a smart assistant for individual tasks, and stopping there. The pattern repeats across industries so consistently that researchers now treat it as the default state.
The common layer looks like:
- Drafting emails, posts, and proposals in ChatGPT or Claude, then editing by hand.
- Summarizing meetings, transcripts, and long documents.
- Brainstorming offers, names, and angles in a chat window.
- A few custom GPTs or saved prompts that package the above.
All of that is real value, and all of it is capped: the person still runs every process, so the business underneath is unchanged. Harvard Business Review, reporting on MIT Media Lab research, noted that 95% of organizations see no measurable return on their generative AI investment. The gap is depth of use, not tooling.
What separates the businesses getting real results from AI?
Structure separates them. MIT's GenAI Divide report found that the roughly 5% of AI efforts producing rapid, measurable results share one habit: they build AI into specific workflows, while the stalled 95% run generic tools alongside old processes.
| The stalled majority | The successful 5% |
|---|---|
| Generic chat tools for everything | AI shaped to one specific workflow |
| Person carries work between tools | Work flows through connected steps |
| Ten shallow experiments | One pain point, executed well |
| AI helps do the old process | AI changes how the process runs |
As the MIT researchers put it, generic tools excel for individuals but stall in business use because they do not learn from or adapt to workflows. The divide is not about budget or technical talent. It is about whether the business itself changes shape around the capability.
How are consultants and coaches specifically using AI in their businesses?
The service businesses using AI structurally point it at the repetitive middle of their operation, the work between winning a client and delivering the result. I build these systems for clients, so the examples below are from real installs.
Where it shows up
- Intake and qualification: AI synthesizes discovery calls and applications into a brief, scored the way the owner scores fit.
- Proposals and RFPs: one client's proposal workflow drafts from a library of past wins and improves with every submission.
- The method, captured: the owner's frameworks live in a knowledge base AI can draw on, so drafts come out in their voice with their thinking.
- Client-facing tools: diagnostics and calculators that deliver the expert's judgment before the first call.
- Being found: content structured so AI engines can read, cite, and recommend the business to buyers who ask.
Each one removes the owner from a loop while keeping their judgment in charge.
Should I copy what big companies are doing with AI?
No. Most of what big companies are doing with AI is failing, and the ways they fail are built into their size. The MIT research that found 95% of pilots stalling was studying exactly those enterprise efforts: committees, integration projects, and tools that never reach the actual work.
Your size is the advantage here:
- One decision-maker. You can rebuild a workflow this month; an enterprise needs three approvals to schedule the meeting.
- Workflows small enough to actually rebuild. Your proposal process is knowable end to end. Theirs spans four departments.
- Direct feedback. You see within weeks whether the rebuilt workflow gives time back.
The winners MIT describes pick one pain point, execute well, and stay narrow. A small expert business is the natural shape for that play. Watch what the quiet, structured small operators do, not what enterprise vendors announce.
Where should a business like mine start with AI?
Start with one workflow that repeats every week and already annoys you, because that is where structural AI pays back fastest and teaches you the most.
A practical sequence:
- Pick the workflow you dread on repeat, proposals, intake write-ups, follow-up emails, content production.
- Write down how you actually do it, every step, including the judgment calls. Your documented method is what makes AI output yours instead of generic.
- Rebuild that one workflow with AI running the steps and you approving the result.
- Let the reclaimed hours fund the next rebuild.
What does not work is sampling: a new tool every week, each one shallow, none of them changing how the work runs. Depth in one place beats breadth everywhere. If you want the first rebuild done with a guide instead of alone, that first working workflow is what the AI Native Activation delivers.
When I look inside businesses, and I have been doing that since 2015, across 70+ programs, I see the same picture the MIT researchers measured: everyone is using AI, and almost nobody's business is built on it. The owner has ChatGPT open all day and the operation underneath runs exactly as it did in 2019. Using AI and being built for AI are different conditions, and the second one is where the results live.
The divide research validated something I had been watching for two years: the winners are not the ones with the best tools or the biggest budgets. They are the ones whose foundation changed shape, whose expertise is captured where AI can use it, whose workflows connect instead of routing everything through the owner's hands. The tools are rented. The foundation is owned. That difference is the whole game.
Here is the part I find genuinely hopeful for small operators: every failure mode in that 95% is an enterprise failure mode. Committees, integration debt, tools that never touch real work. You have none of that gravity. A business your size can cross the divide in a season, one deeply rebuilt workflow at a time, and end up with something the big players openly struggle to build: a smarter business, not a bigger one.
Cindy Anne Molchany · Founder
Behind the majority, no, because drafting in a chat window is the majority. The MIT GenAI Divide research suggests roughly 95% of AI efforts stall at that shallow layer. You are behind the small group using AI structurally, and that gap is the one worth closing, because the structural users are pulling away in time, capacity, and visibility while the field stands still.
No. The divide between stalled and successful AI use is structural, not financial. Subscription-level tools are enough to rebuild a workflow end to end; what the successful group invests is clarity, writing down their method and redesigning one process around it. Enterprises spend millions and still land in the stalled 95%, which is strong evidence that money is not the differentiator.
Usually the repetitive middle of the operation: proposals, intake write-ups, follow-up sequences, and content production. Those workflows recur weekly, consume owner hours, and have clear outputs, so an AI rebuild shows results within weeks. Visibility work, making your business readable and recommendable to AI engines, compounds more slowly but reaches buyers you would never have met. Structural businesses run both tracks.
Weeks, not quarters, for internal workflows. One rebuilt process, proposals or intake, for example, starts returning hours as soon as it runs, and you feel the difference within the first few cycles. Visibility results move on a slower clock, because AI engines need time to crawl and cite new content. Rebuild an internal workflow first, and let its reclaimed hours fund the visibility work.
Your customers now start inside AI, which names a few businesses instead of listing links. Here is what changed, why it happens, and how to see what AI says about you.
If AI made you busier, you are likely using it in pieces on top of your old way of working. Here is why that adds work, and when AI actually gives time back.
Vibe coding means telling AI what you want in plain English and letting it write the code. Here is what the term covers, and why it puts custom business tools within your reach.