What Is the 70-20-10 Rule for AI Investment? A Budget Framework for Mid-Sized Companies
The 70-20-10 rule for AI investment allocates 70% of AI budget to people and training, 20% to process and workflow redesign, and 10% to technology and tools. This framework reflects a consistent finding across AI research: PwC analysis shows technology delivers only about 20% of an AI initiative’s value, while the other 80% comes from how people use that technology and how workflows adapt around it. AI Smart Ventures has validated this ratio across close to 1,000 organizations, documenting 50% average time savings when investment priorities follow this pattern versus the 15-20% adoption rates typical of technology-first approaches.
Most organizations get this backwards. They spend 70% on software licenses and platforms, 20% on implementation, and 10% on training as an afterthought. Then they wonder why adoption stalls at early adopters and ROI never materializes.
The math is simple. A $100,000 AI investment split 70-20-10 on people-process-technology outperforms a $500,000 investment split the opposite way. Every time. The difference isn’t the tools. It’s whether anyone actually uses them.
Why Do Most Companies Invert the AI Investment Ratio?
Technology spending feels productive. You can point to a purchase order. You can demo new capabilities. You can show the board a roadmap with platform names on it.
People spending feels soft. Training doesn’t photograph well. Workflow redesign sounds like consulting jargon. Change management seems optional until adoption fails.
This bias toward visible technology investment creates predictable failures. BCG research shows 70% of AI pilots never reach production. MIT found 95% of enterprise AI projects show zero measurable ROI within six months. These statistics reflect technology investments made without corresponding investment in the people expected to use that technology.
Vendor incentives amplify the problem. Software companies sell software, not training. Their sales teams are measured on license revenue, not adoption rates. Implementation partners bill for technical work, not organizational change. The entire AI industry ecosystem pushes organizations toward technology spending because that’s where vendor revenue comes from.
Mid-sized companies face additional pressure. With limited budgets, spending $50,000 on training feels like money that could have bought better tools. The logic is intuitive but wrong. Better tools with undertrained users deliver less value than adequate tools with well-prepared teams.
For a deeper look at why these failures happen, see what are the biggest AI implementation mistakes and how to avoid them.
What Does 70% on People Actually Mean?
The 70% people allocation covers everything required to build organizational AI capability that persists after any engagement ends.
AI literacy training ensures everyone understands what AI can and cannot do, how to evaluate AI outputs, and when to trust or verify AI recommendations. This isn’t tool-specific training. It’s building judgment that applies across any AI system.
Role-specific skill development teaches people how to use AI in their actual jobs. A marketing manager needs different AI skills than an operations director. Generic “intro to AI” workshops waste time. Targeted training on specific workflows delivers immediate value.
Prompt engineering and output verification transforms mediocre AI users into effective ones. The difference between useful and useless AI output usually isn’t the tool. It’s how questions are framed and how outputs are evaluated.
Change management and adoption support addresses the human resistance that kills most AI initiatives. People fear job displacement, distrust unfamiliar technology, and resent disruption to established routines. Addressing these concerns requires investment, not just announcements.
Leadership development prepares executives to guide AI transformation rather than just approve budgets. Leaders who understand AI make better decisions about where to invest and when to push forward or pull back.
AI Smart Ventures has trained over 20,217 professionals in Applied AI across these dimensions. The consistent pattern: organizations that invest in comprehensive people development achieve adoption rates above 80%, while those that skip training plateau at 15-25%.
What Does 20% on Process Mean?
The 20% process allocation covers workflow redesign that captures AI value.
Workflow mapping and analysis documents how work currently flows before introducing AI. You cannot improve what you don’t understand. Many organizations discover during this phase that their biggest problems aren’t AI problems at all.
Process redesign changes how work gets done to take advantage of AI capabilities. Adding AI to existing processes captures minimal value. Redesigning processes around AI capabilities captures substantial value.
Integration planning connects AI tools to existing systems and data sources. This isn’t technology spending. It’s understanding what connections matter and how information should flow.
Quality and verification frameworks establish how AI outputs get reviewed and validated before use. Without these frameworks, organizations either over-rely on AI (accepting errors) or under-utilize AI (reviewing everything manually and negating the efficiency gains).
Performance measurement creates systems to track whether AI implementation is working. Measurement enables course correction. Organizations without measurement cannot distinguish successful initiatives from failing ones.
What Does 10% on Technology Mean?
The 10% technology allocation may seem low, but it reflects a strategic reality: most mid-sized companies already own significant AI capability they aren’t using.
Microsoft Copilot is included in many Microsoft 365 subscriptions. Google Gemini is built into Google Workspace. CRM platforms like HubSpot and Salesforce include AI features most users never activate. Before purchasing new AI platforms, audit what you already have.
The 10% covers gaps that existing tools genuinely cannot fill, subscription costs for tools that provide clear value, and infrastructure improvements required for AI operation. It does not mean skimping on necessary technology. It means buying technology after establishing clear use cases rather than buying technology and searching for use cases.
For more on this approach, see what is an AI revamp and why you don’t need another AI tool in 2026.
How Do You Apply 70-20-10 to Your Actual Budget?
Translating the framework to real numbers requires honest assessment of current spending patterns.
| Budget Size | People (70%) | Process (20%) | Technology (10%) |
| $25,000 | $17,500 | $5,000 | $2,500 |
| $50,000 | $35,000 | $10,000 | $5,000 |
| $100,000 | $70,000 | $20,000 | $10,000 |
| $250,000 | $175,000 | $50,000 | $25,000 |
These allocations assume you’re starting fresh. If you’ve already purchased AI tools, your technology spending is committed. Reallocate remaining budget toward people and process rather than buying more technology.
The framework is a principle, not a rigid formula. Some organizations may need 60-25-15 or 75-15-10 based on their starting point. The directional guidance matters more than exact percentages: prioritize people over process, process over technology.
When Should You Adjust the 70-20-10 Ratio?
Certain situations warrant adjusting the standard ratio.
Increase people allocation to 80% when previous AI initiatives failed due to low adoption, when workforce AI literacy is minimal, or when organizational resistance to change is high. More training addresses the actual constraint.
Increase process allocation to 30% when existing workflows are undocumented or chaotic, when the AI use case requires significant process change, or when quality control systems don’t exist. Process work becomes the bottleneck.
Increase technology allocation to 20% when existing tools genuinely cannot address the use case, when data infrastructure requires upgrade, or when security or compliance requirements mandate specific platforms. Sometimes you do need to buy something.
What you should never do: increase technology allocation because it feels more concrete or because vendors are pressuring you. Those are the wrong reasons to deviate from a framework that reflects how AI value actually gets created.
What Mistakes Do Companies Make With AI Budgets?
Beyond inverting the ratio entirely, several specific mistakes undermine AI investment.
- Counting training as one-time cost. AI capabilities evolve constantly. Training requires ongoing investment, not a single workshop. Budget for continuous learning, not just initial rollout.
- Hiding people costs in project budgets. When training and change management get buried in “implementation,” they’re the first line items cut when budgets tighten. Make people investment visible and protected.
- Measuring technology spending as AI progress. Procurement is not progress. Adoption is progress. Organizations that celebrate tool purchases before measuring adoption confuse activity with results.
- Underestimating workflow redesign complexity. Process change takes longer than technology deployment. Budget time and patience, not just money.
- Skipping measurement investment. Without measurement, you cannot prove ROI, learn from results, or justify continued investment. Measurement feels optional until the CFO asks for evidence.
For comprehensive guidance on demonstrating AI value, see how do you measure AI ROI: a framework for business leaders.
Frequently Asked Questions
What is the 70-20-10 rule for AI investment?
The 70-20-10 rule allocates AI budgets with 70% going to people and training, 20% to process and workflow redesign, and 10% to technology and tools. This framework reflects research showing technology delivers only about 20% of AI value while people and process changes deliver 80%. Organizations that follow this ratio consistently outperform those that prioritize technology spending, achieving higher adoption rates and faster time-to-value.
Why should people get 70% of AI budget?
People receive the largest allocation because adoption determines whether AI investments succeed or fail. BCG research shows 70% of AI pilots never reach production, typically due to adoption failures rather than technology failures. Training builds AI literacy, role-specific skills, prompt engineering capability, and change acceptance. Without these investments, expensive technology sits unused. Organizations report 80% adoption rates with adequate training versus 15-25% without it.
Is 10% enough for AI technology?
Yes, for most mid-sized companies. The 10% reflects a strategic reality: most organizations already own significant AI capability through Microsoft 365, Google Workspace, and existing CRM platforms. The technology allocation covers gaps existing tools cannot fill, not complete platform purchases. Organizations that audit existing capabilities before buying often discover they need less new technology than expected. The goal is using tools effectively, not accumulating tools.
How does 70-20-10 differ from enterprise AI budgeting?
Enterprise approaches often allocate 50-70% to technology and infrastructure, reflecting their scale, complexity, and dedicated AI teams. Mid-sized companies lack these resources and face different constraints. Following enterprise patterns leads mid-sized companies to overspend on technology they cannot fully utilize while underspending on the training and process work that determines success. The 70-20-10 framework fits mid-market reality and resource constraints.
What does the 20% process allocation cover?
The process allocation covers workflow mapping and analysis, process redesign to capture AI value, integration planning between systems, quality and verification frameworks for AI outputs, and performance measurement systems. This work determines whether AI tools integrate into actual operations or remain isolated experiments. Organizations that skip process work deploy AI but don’t change how work gets done, capturing minimal value from their technology investment.
How do you track 70-20-10 spending?
Create separate budget categories for people, process, and technology investments. People includes all training, workshops, coaching, and change management. Process includes workflow analysis, redesign, integration planning, and measurement systems. Technology includes software licenses, subscriptions, and infrastructure. Review allocations quarterly and adjust based on adoption metrics and value realization. If adoption lags, increase people allocation before buying more technology.
Can you start with lower AI budgets using 70-20-10?
Absolutely. The framework scales to any budget size. A $25,000 annual AI investment following 70-20-10 produces $17,500 for training, $5,000 for process work, and $2,500 for tools. This modest budget, properly allocated, often outperforms $100,000 investments that prioritize technology. Starting small with correct allocation builds evidence for larger future investments while avoiding the waste of technology-first approaches.
What if we already bought AI tools?
If technology investment is already committed, allocate remaining budget entirely toward people and process. Don’t buy more tools to match sunk technology costs. Instead, maximize value from existing purchases through training and workflow redesign. Many organizations with underperforming AI investments can improve results dramatically by shifting subsequent budget toward adoption rather than additional technology.
Does 70-20-10 apply to specific AI tools like ChatGPT?
Yes. Even when the technology is nearly free, the ratio applies to time and attention allocation. Free tools still require investment in learning effective use, integrating into workflows, establishing quality standards, and managing organizational change. The principle holds regardless of tool cost: people and process investment determines whether AI tools deliver value.
Who developed the 70-20-10 AI investment framework?
The 70-20-10 framework emerged from observations across AI implementations rather than a single source. PwC research establishing that technology delivers only 20% of AI value provides the empirical foundation. AI practitioners working with mid-sized organizations have refined the specific allocations based on what consistently produces adoption and ROI. The framework represents accumulated pattern recognition from thousands of implementations, not academic theory.
What Should You Do Next?
Audit your current AI spending against the 70-20-10 framework. Calculate what percentage currently goes to technology versus people versus process. Most organizations discover significant imbalance toward technology.
Reallocate future AI budget toward people and process investments. Protect training budgets when other priorities compete for resources. Measure adoption as the primary success metric rather than technology deployment.
Organizations that invest in people first build AI capability that compounds over time. Teams that understand how to use AI effectively apply that knowledge to new tools and new use cases. Technology purchases depreciate. Capability investments appreciate.
Get Your AI Readiness Assessment
AI Smart Ventures helps mid-sized organizations allocate AI investment for maximum impact. Our complimentary AI Readiness Assessment evaluates your current technology landscape, organizational readiness, and skill gaps to recommend the right balance of people, process, and technology investment for your situation.
The assessment takes 30 minutes and delivers practical recommendations for where your AI budget will generate the strongest returns, ensuring your investment builds lasting capability rather than accumulating unused tools.
Schedule your free AI Readiness Assessment to build an AI investment strategy that prioritizes people and delivers results.
This content is for informational purposes only and does not constitute professional business or technology advice. Results vary based on industry, existing systems, and implementation commitment.
About the Author
Nicole A. Donnelly is the Founder of AI Smart Ventures and an AI Adoption Specialist with 20 years of experience as a founder and CEO and over a decade leading AI adoption initiatives. She helps businesses integrate artificial intelligence with clarity and confidence, driving innovation and sustainable growth. Nicole has trained over 20,217 professionals in Applied AI, delivered 624 workshops, and worked with close to 1,000 organizations across diverse industries.
Expertise: AI Transformation, AI Strategy, AI Implementation, AI Adoption, Applied AI, Marketing, Business Operations
