Somebody Lied to You
About AI.
Boston Consulting Group studied AI transformation across hundreds of organizations and found something that should make every executive pause before signing another software contract. 70% of AI success has nothing to do with the technology. The 10-20-70 rule is here — and it is about to change how you think about everything you have spent so far.
Picture this. It is 9am on a Monday. You are in a conference room with your leadership team. A vendor just finished a 45-minute presentation — beautiful slides, impressive demo, a ROI calculator that shows your investment paying back in 14 months. The room is buzzing. You approve the budget. Contract signed by Wednesday.
Six months later, the tool is technically live. Adoption is at 31%. Three of your most senior people have figured out workarounds that completely bypass the system. Your IT team is drowning in integration tickets. And the 14-month ROI? Nobody has checked. Because nobody defined what it was supposed to look like before you bought the thing.
Sound familiar? I am not asking to embarrass anyone. I am asking because this exact scenario plays out in organizations every single week — and Boston Consulting Group’s research tells us exactly why.
— A composite story from dozens of real conversations. Every detail is real. Only the names are missing.BCG studied AI transformation across hundreds of organizations and found something that should make every executive pause before they sign their next software contract. The companies capturing the most value from AI spend their effort like this: 10% on algorithms, 20% on technology and data, and 70% on people and processes. That is the 10-20-70 rule. And here is what makes it interesting — most organizations are doing it almost perfectly backwards.
How Most Organizations Are Spending Their AI Budget vs. How the Winners Spend Theirs
Let me show you two organizations. Same industry. Similar size. Both announced AI initiatives in the same quarter. One is the case study you want to be. The other is the cautionary tale you probably recognize.
“The companies capturing the most value from AI follow the 10-20-70 rule. 10% of their AI effort goes to designing algorithms, 20% to building the underlying technologies, and 70% to supporting people and adapting business processes.”— Vinciane Beauchene, Managing Director & Partner, Boston Consulting Group
An Analogy That Will Ruin Every AI Sales Pitch You Sit Through
Buying AI Without Investing in People Is Like Buying a Professional Stand Mixer and Expecting Bakery Croissants.
The equipment matters — sure. A professional stand mixer is a genuinely impressive piece of technology. It is faster, more powerful, and more consistent than anything you would do by hand. But the mixer is maybe 20% of the equation. The other 80% is knowing what you are doing with it. The recipe. The technique. The muscle memory built from doing it wrong seventeen times before you get it right. Without that knowledge, the mixer just makes bad dough faster. And that is exactly what AI does to broken processes at scale. It automates them. At speed. With confidence. Before anyone notices they were broken in the first place.
Why the People Gap Is Bigger Than Anyone Is Comfortable Admitting
I want to sit with that 5% number for a second. Five percent. That means 95 out of every 100 AI pilots fail to deliver revenue growth. And before you assume those failing organizations just chose the wrong technology — they largely did not. Many of them licensed the same enterprise models as the 5% that succeeded. The difference was not in the algorithm. It was in everything that happened around it.
I spoke with a COO recently — brilliant, experienced, genuinely committed to getting AI right. She told me they had spent $2.4 million on their AI platform in the first year. When I asked how much they had invested in workforce readiness and process redesign, she paused. “We did two training days at launch,” she said. “And we put some resources on the intranet.”
Two training days and some intranet resources to support $2.4 million in technology investment. That is not a failure of leadership — she is a great leader. That is a failure of the mental model most organizations are still using to think about AI. The platform is the visible thing. The people work is the invisible thing. Boards want to see dashboards. Nobody’s asking for a slide on “are your people actually ready.”
— A real conversation from Q1 2026. Details changed for privacy.Breaking Down the Part of AI Success That Nobody Is Investing In
When BCG says 70% of AI success comes from people and processes, a lot of executives nod politely and then go back to evaluating platforms. So let me be specific about what that 70% actually means in practice — because it is more operational and less fuzzy than the phrase “people and processes” suggests.
Workflow Redesign
Not “add AI to the existing process.” Rethink the process entirely for AI. What steps exist only because a human had to do them manually? Which checkpoints exist because someone kept making mistakes? Which approvals add time but not value? AI-ready workflows are fundamentally different from human-designed ones.
Most organizations skip this entirelyAI Literacy at Every Level
Not a two-day training and a LinkedIn Learning module. Genuine, role-specific capability building that teaches each person how AI changes their specific job — what they can trust, what they should verify, where their judgment still matters, and what new skills they need to develop.
Most organizations do a fraction of thisGovernance That Enables Instead of Blocks
Clear rules for what AI can do autonomously, what requires human review, and what requires senior authorization. Built before deployment — not written reactively after the first incident. Governance that people actually understand and follow, rather than route around with shadow tools.
Over 50% of employees use unauthorized AI toolsChange Management That Is Actually Managed
Not a launch email and a FAQ document. Sustained communication about why the change is happening, what it means for each role, how concerns are being addressed, and what the timeline looks like from here. Change management that treats people as adults who can handle real information.
Consistently the most underinvested elementFeedback Loops That Actually Loop
Structured channels for employees to report what is working, what is creating friction, and what the AI is getting wrong — that are actually read, responded to, and acted upon. The organizations with the highest adoption rates treat employee feedback as AI improvement data, not employee complaints.
Most are one-directional: launch and listen to nothingMeasurement Against Business Outcomes
Not adoption rates. Not system uptime. Not “number of queries processed.” Revenue. Cost. Customer satisfaction. Quality. The specific, measurable business outcomes that justified the AI investment — captured against a baseline, measured monthly, reported to leadership who actually care about the result.
Only 23% of organizations do this consistentlyWhere Most Organizations Actually Are vs. Where They Need to Be
Five Moves That Bring Your Investment Closer to 10-20-70
You do not have to overhaul your entire AI strategy in one quarter. But here are five things you can do right now that will start closing the gap:
Run the Honest Budget Exercise
Pull your actual AI spending from the last 12 months and bucket it honestly: algorithms/models, technology/data, people/processes. See where your percentages land. Most organizations that do this exercise are genuinely surprised by what they find — and that surprise is the beginning of a better conversation.
Ask Five Frontline People What They Are Actually Using
Not what they are supposed to be using — what they are actually using. Listen without judgment. The gap between those two answers is your adoption gap in miniature, and it will tell you more about your AI readiness than any adoption dashboard.
Pick One Workflow and Redesign It — Not Just Automate It
Choose the highest-value AI use case in your organization. Now ask the question BCG recommends: not “how can AI fit into this workflow?” but “how would we design this workflow differently if we were building it for AI from scratch?” The answer is almost always faster, simpler, and more valuable than the automated version of the old process.
Define a Business Outcome Metric — Before the Next Deployment
Before your organization approves the next AI platform investment, require a written statement of the specific business outcome it will deliver — in revenue, cost, quality, or time — with a baseline captured before launch and a monthly measurement commitment from a named owner. This single discipline will change how you buy AI.
Create a Shadow AI Amnesty
Over 50% of employees are using unauthorized AI tools because the official ones are too cumbersome or irrelevant. Declare a brief amnesty period where people can share what they are actually using without fear of consequence. What you discover will tell you exactly where your official AI program is failing — and what to fix first.
The Good News Nobody Is Telling You
Remember the organization from the opening story? Six months after applying the 10-20-70 lens — without buying a single new AI tool — they saw a 34% improvement in adoption of the tools they already had. Not because they added technology. Because they finally invested in the 70% they had been skipping.
The same tools. The same models. The same licenses they had already paid for. The only thing that changed was that they stopped spending all their energy on the algorithm and started investing in the people who were supposed to be using it.
That is the good news the AI market does not want you to hear: you may already have everything you need to get a dramatically better result. You just need to stop buying more of the 10% and start investing in the 70%.
— A real outcome. Same organization, six months later.“The best AI investment your organization can make right now might not be a new platform. It might be finally investing in the people who are supposed to be using the platforms you already have.”— Kelli Gilmore, COO, MindFinders
The MindFinders Difference
We Are the 70%. That Is Literally What We Do.
MindFinders was built on the conviction that AI transformation is a people strategy first and a technology strategy second. The operational readiness frameworks, the workflow redesign, the change management, the workforce upskilling, the governance architecture, the performance measurement — the things BCG’s research identifies as 70% of what makes AI succeed — that is our practice. That is where we live.
- We conduct the honest budget exercise with your leadership team — and show you exactly where the investment imbalance is
- We design and deliver role-specific AI literacy programs that build genuine capability, not checkbox completion
- We redesign workflows for AI — not automate the old ones
- We build governance frameworks that enable speed rather than generate paperwork
- We structure change management programs that treat your people as adults who deserve real information
- We create the feedback loops and measurement infrastructure that make improvement continuous rather than accidental
What Percentage of Your AI Budget Is Going to the 70%?
Let’s run the honest exercise together — see where your investment actually sits, identify the gaps that are costing you adoption and ROI, and build the people and process infrastructure that makes your AI investment finally deliver what the vendor’s calculator promised.
Schedule Your Free Consultation
