The 10-20-70 Rule for AI: Where Your Budget Really Matters

You've allocated a six-figure budget to AI. You've purchased best-in-class software. You've deployed cutting-edge models. And yet, your ROI remains flat. Why?

According to Boston Consulting Group's research on AI transformation, 70% of AI projects fail to deliver expected returns—not because the technology is bad, but because companies invest in the wrong things. The culprit? A fundamental misalignment in budget allocation.

Enter the 10-20-70 rule for AI, a framework that has become the gold standard for executives serious about AI transformation. This principle challenges everything you think you know about AI spending and offers a clear roadmap for success.

What Is the 10-20-70 Rule for AI?

The 10-20-70 rule is a budget allocation framework developed by BCG that prescribes where organizations should invest their AI transformation dollars:

10% on algorithms and models – Software, SaaS tools, and model selection
20% on technology and data infrastructure – Cloud platforms, data harmonization, APIs, and system integration
70% on people, processes, and organizational change – Training, change management, workflow redesign, and role redefinition

The most counterintuitive insight? 70% of AI value is generated by business process transformation, not by the algorithms themselves. This means the real work of AI adoption isn't technical—it's organizational.

BCG's analysis of high-performing AI deployments confirms that companies adhering to this 10-20-70 split systematically outperform competitors who skew toward technology spending. Organizations that violate this ratio—such as those spending 80% on software and 20% on talent—consistently underdeliver on ROI.

The BCG 10-20-70 Principle: Why It Matters for CEOs

The BCG 10-20-70 principle emerged from studying hundreds of AI implementations across industries. The finding was clear: technology is the easy part. Transformation is hard.

Here's why this matters to executives:

The Talent-to-Software Ratio: BCG identified that high-performing programs maintain a 1.2x talent-to-software spending ratio. In other words, for every dollar spent on software, organizations should spend $1.20 on people and implementation. Programs falling below this threshold systematically underperform.
Process Redesign Drives Value: Bolting AI onto existing workflows doesn't work. The 70% allocation funds the critical work of redesigning processes, redefining roles, and preparing teams for an AI-augmented future.
Cultural Change is Non-Negotiable: Without proper investment in change management, employee training, and communication, even world-class AI tools gather dust.

Forbes has highlighted this principle in its executive guides, emphasizing that CEOs who understand the 10-20-70 rule gain a competitive advantage in their AI transformation journey.

How to Allocate Your AI Budget: A Practical Framework

The 10%: Select "Good Enough" Over Custom Builds

Too many mid-market companies waste their 10% algorithm budget on custom model development. The 10-20-70 rule advocates a different approach: deploy a strong off-the-shelf SaaS solution in Week 1.

The rationale is simple: unless you're building a proprietary competitive advantage, custom models are slower, more expensive, and less maintainable. Modern SaaS AI tools are mature enough for 80% of use cases. Focus your innovation energy elsewhere.

The 20%: Build the Data Foundation

Your 20% funds the infrastructure that makes AI actually work. This includes:

Cloud platform setup (AWS, Azure, GCP)
Data warehousing and harmonization
API integrations with existing systems
Data governance and security frameworks
Human-in-the-loop safeguards to prevent hallucinations and errors

This foundational work is unglamorous but essential. Without it, your algorithms will fail.

The 70%: Invest in Transformation

The largest allocation funds the work that actually drives ROI: people and process transformation. This includes:

Training programs: Equip employees with skills to work alongside AI
Change management: Build organizational buy-in and overcome resistance
Workflow redesign: Don't automate broken processes; fix them first
Role redefinition: Update job descriptions and KPIs to reflect the AI-augmented operating model
Strategic leadership: Hire or contract senior implementation experts and AI transformation officers

This 70% is where the real value is created—and it's also where most organizations underinvest.

What Is the 20-60-20 Rule for AI? (And Why It's Different)

You may have encountered the 20-60-20 rule in your research. Here's the important distinction: this is not the same as the 10-20-70 rule.

The 20-60-20 rule is a development lifecycle framework for software projects, specifically popularized by Robert Menetray in Drupal development. It allocates time as follows:

20% to planning
60% to development
20% to testing and deployment

This rule answers the question: "How should we spend our time during a software development project?" The 10-20-70 rule answers a different question: "How should we allocate our transformation budget?"

Do not confuse the two. Using the 20-60-20 development timeline as your budget allocation framework will lead to underinvestment in people and process—and project failure.

Decoding the "$900,000 AI Job": Understanding AI Transformation Roles

If you've heard about a "$900,000 AI job," you're witnessing the market value of the 70% allocation in action.

The 10-20-70 rule implies that the 70% devoted to people often includes high-impact strategic roles such as:

AI Transformation Officers: Senior leaders who drive organizational change
Implementation Strategy Partners: Consultants who redesign workflows and manage change
AI Center of Excellence Directors: Executives who build internal AI capabilities
Enterprise AI Architects: Specialists who align AI strategy with business processes

In large enterprises, these roles command premium compensation—sometimes reaching $900,000+ in total compensation—because they directly control whether your organization captures 70% of the value at stake. A single great transformation leader can be worth millions in realized ROI.

This is why the 70% allocation makes financial sense: you're not just training employees; you're hiring strategic talent to redesign your business for the AI era.

Common Mistakes: How Organizations Violate the 10-20-70 Rule

Mistake #1: Treating AI as a Technology Problem

The Reality: Companies often allocate 70% to technology, 20% to data infrastructure, and only 10% to people and process. This is an inversion of the recommended split—and it fails.

When you over-invest in software and under-invest in people, you get powerful tools that nobody knows how to use effectively. The software gathers dust while employees continue old workflows.

Mistake #2: Building Custom Models Instead of Using SaaS

The Reality: Your 10% algorithm budget should fund SaaS solutions and model selection, not custom builds. Custom models are slower, more expensive, and require ongoing maintenance.

Unless you're a tech company or building a proprietary competitive advantage, off-the-shelf models will get you 80% of the way there. Focus your energy on the other 80% (process transformation).

Mistake #3: Confusing the 20-60-20 Development Rule with Budget Allocation

The Reality: The 20-60-20 rule is a development timeline, not a budget framework. Don't use it to allocate transformation dollars—you'll starve the people and process side of your initiative.

Mistake #4: Ignoring the Talent-to-Software Ratio

The Reality: BCG identified that high-performing programs maintain a 1.2x talent-to-software ratio. If your spending ratio is inverted (more software than talent), you're almost certainly leaving value on the table.

Conduct a quick audit: total your last three quarters of AI spending. What's your talent-to-software ratio? If it's below 1.2x, you need to rebalance.

Actionable Strategy: Implementing the 10-20-70 Rule

Step 1: Start with the 70%

Reverse-engineer your budget planning. Begin by allocating 70% to people and process, then work backward to the 10% and 20%.

Ask yourself: What processes do we need to redesign? What training will employees need? What roles will change? How will we manage cultural resistance? Budget accordingly.

Step 2: Choose Your First "Play"

BCG identifies three AI value plays. Start with one:

Deploy: Immediate productivity gains using off-the-shelf tools (quick wins)
Reshape: Transforming critical business functions (medium-term value)
Invent: Creating entirely new revenue streams (long-term innovation)

Most organizations start with "Deploy" to build organizational confidence and capture quick wins.

Step 3: Implement a 90-Day Cadence

Stop planning five-year transformations. Instead, adopt a 90-day cycle:

Months 1–3: Achieve one measurable quick win
Months 4–6: Scale what worked
Months 7–12: Tackle bigger, more complex problems

This approach maintains momentum, builds organizational confidence, and allows you to adjust strategy based on real results.

Step 4: Redefine Roles and KPIs Before Launch

Before deploying any AI tool, update job descriptions and performance metrics to reflect the new AI-augmented operating model. Train teams on the new workflow, not just the software.

If you don't redefine roles and KPIs, employees will continue working in old patterns—and you'll wonder why adoption is low.

Step 5: Calculate ROI Upfront

Before launching, define measurable ROI. Answer this question precisely: "When do we break even?" If you can't answer it, don't launch.

The 70% allocation makes financial sense only if you can connect it to business outcomes. Track metrics religiously.

The Leader's Guide to Transforming with AI: Key Takeaways

AI transformation is primarily an organizational challenge, not a technology one. Allocate 70% of your budget to people and process, not software.
Use the 1.2x talent-to-software ratio as your north star. If you're spending more on software than talent, you're off track.
Start with SaaS solutions, not custom builds. Your 10% should fund model selection and strong off-the-shelf tools.
Don't confuse the 20-60-20 development rule with the 10-20-70 budget rule. They answer different questions.
Design processes first, then deploy technology. Automation of broken workflows creates broken automation.
Redefine roles and KPIs before implementation. Organizational alignment precedes technology adoption.
Operate in 90-day cycles. Build momentum with quick wins, then scale.

Conclusion: The Budget Rebalancing Challenge

If your organization has been following the traditional tech-heavy approach to AI investment, the 10-20-70 rule will feel counterintuitive. Your instinct might be to invest more in algorithms and infrastructure. Resist that instinct.

The data is clear: organizations that adhere to the 10-20-70 split outperform competitors who don't. The difference isn't that they have better technology—it's that they've invested in the organizational capabilities to extract maximum value from that technology.

Start today. Audit your last three quarters of AI spending. Calculate your current allocation. If you're not close to 10-20-70, begin rebalancing. Your future ROI depends on it.