LinkedIn
You’ve probably talked about AI transformation in boardrooms, strategy off-sites, and networking events. The discussions often focus on automation, efficiency, and gaining a competitive edge. But here’s the tougher question: have you thought about the fact that the success of AI implementation usually isn’t just about the technology? More often than not, it’s about people.
When implementing AI into business strategies, it is important to understand that AI transformation is an organisation-wide shift rather than a technical upgrade at the individual departmental level. What would it mean to have a framework that could provide detailed steps for executives to redesign processes, create role-specific AI fluency, and develop targeted metrics for adoption? Would it be an opportunity for executives to empower AI to create and capture value and to select technologies that extend human judgment?
This guide outlines governance guardrails, pilot-to-scale tactics, human-in-the-loop AI transformation, and change-management moves that accelerate adoption. Read on for a practical guide that turns AI from an experimental expense into a lasting business benefit and provides clear ROI.
What is the 70-20-10 Rule in AI Transformation?
The 70-20-10 rule now guides organizations that want to succeed with AI. Most people think AI success depends on technology, but this framework shows a different reality.
Most companies see AI transformation as just another tech project, and most promising AI projects never grow beyond testing. The ideal ai transformation framework for implementation needs to be the following:
- 70% - People and Processes: This covers workflow redesign, leadership support, team training, and new behaviour patterns. Companies need to change how they work, what they reward, and how they measure success. Even the best AI systems fail without this cultural foundation.
- 20% - Technology and Data: Constructing proper data foundations and technological infrastructure in support of AI initiatives is important, but the building blocks alone will not differentiate your organization.
- 10% - Algorithms and Models: Transformative AI algorithms lose their potential value if they operate in silos, away from the people, processes and data ecosystems.
Companies implementing the 70-20-10 principle avoid the lure of having no clarity around new models (''shiny toy syndrome'') and steer clear of the testing limbo (''pilot purgatory''). High-performing businesses encourage environments where AI utilization is driven by the goals of the business. They let the right people outline the problems to be solved before selecting the right solutions. Training investments help, the ''no blame'' culture of failure is adopted, and adaptive processes and digital transformations are coupled. This balanced approach makes AI transformation truly powerful, and it becomes a force that makes human capabilities better.
Why Human-centric AI and People-First AI Matter
Boardroom discussions about AI transformation often focus on algorithms and computing power. This misses what truly drives successful AI adoption: “people”.
While the implementation of the strategies generated by the algorithms might be valuable, if algorithms are the only focus, the psychological and operational realities of implementation are ignored. Employees may have worries regarding displacement, become overwhelmed by additional tools, and find it difficult to envision how AI fits into the new reality of their everyday tasks. Initiatives can be sophisticated and expensive, but without certainty to alleviate fears, resistance will undermine the initiative.
Human centered-AI transforms and focuses directly on this friction. It positions AI as a complement to human capability, not a substitute. The elimination of repetitive, routine tasks will allow teams to avoid the tactical and operational tasks and focus on the mental, creative, and relational. This model of partnership will build trust and accelerate adoption, resulting in more sophisticated solutions. Human-centric AI transformation matters for several key reasons:
- Trust and adoption: Companies where leaders promote AI see better teamwork and staff participation. AI should be positioned as a tool that enhances human capabilities.
- Improved capabilities: AI handles routine tasks well, which lets people focus on creativity, emotional intelligence, and strategic thinking.
- Cultural transformation: Organizations that smoothly implement AI encourage their staff to try new tools.
- Balanced outcomes: Human-focused methods create a "double bottom line" that balances profits with social impact. This ensures AI serves broader company goals instead of just chasing efficiency.
People-first AI transformation helps break down departmental barriers, too. AI works like a "cognitive exoskeleton" that increases employee knowledge by sharing relevant information across teams. This creates connections that promote innovation.
The difference between technology-first and human-first approaches is clear. Companies that force their processes to match technology see poor adoption. However, when we design systems based on human needs, intent, and natural work patterns, we can overcome technology barriers and accelerate real change. So, success with AI transformation requires putting people first in your strategy.
From Vision to Execution: The AI Transformation Roadmap
Effective AI transformation begins by aligning AI initiatives with core business goals. Successful organisations prioritize their main objective: the overall business strategy, instead of getting caught up in conversations about algorithms or models. A successful AI transformation roadmap typically follows these critical phases:
Define Clear, Measurable Goals:
Your first step is to identify specific problems or opportunities that AI transformation can address within your organization. These goals must be precise and measurable, improving operational efficiency by a certain percentage, enhancing customer response times, or increasing sales forecast accuracy. Capital One’s approach included significantly improving fraud detection accuracy and cutting down on manual processing time. They focused on meaningful outcomes rather than pursuing technology for its own sake. ( Source)
Assess Data Quality:
The results from AI depend on the quality of the input data. It's essential to assess data quality based on accuracy, completeness, consistency, and relevance. Optimized data pipelines and adequate storage solutions help data flow efficiently into AI models.
Build Skilled Teams:
These teams consist of data scientists, machine learning engineers, software developers, and project managers who have experience with AI transformation. Organizations that promote innovation inspire employees to embrace change, explore new ideas, and take part actively in the adoption process. Financial services firm Citi built a network of around 4,000 internal “AI accelerators” and champions who help colleagues adopt and scale AI tools within business units. (Source)
Implement, Test, Scale:
Start with pilot projects to assess capabilities in low-risk environments before full deployment. Models should undergo thorough testing using separate validation datasets before implementation. A solid scalability plan allows systems to handle growing data volumes without performance loss.
Establishing Governance and Decision Rights
Define decision rights at the strategic level for use case selection, the tactical level for model architecture, and the operational level for monitoring. Clear AI data governance prevents paralysis and ensures oversight.
Organizations successfully implementing AI transformation keep an eye toward "the art of the possible" with a growth mindset throughout this process.
One of the world’s largest retailers faced constraints due to on-premises infrastructure that could not support advanced analytics or scalable AI tasks. Tredence worked with the client to create and implement a hybrid analytics platform on the Google Cloud Platform (GCP). This setup allowed access to customer data and advanced machine learning capabilities across the enterprise. This case illustrates how thoughtful AI implementation, supported by modern infrastructure and user acceptance, changed analytics from a reporting task into a business advantage. (Source)
Change Management and Cultural Shift for AI Integration
The implementation of technology such as AI can be mandated, but cultural change cannot be mandated. AI adoption is primarily focused on cultural change, which centers on the willingness of people to allow data to drive their decisions, accept recommendations from an algorithm, and view AI transformation as an enhancement to their work.
Understanding and Addressing Resistance
Resistance is predicted at every layer of the organization. The frontline fears job loss. The middle levels fear loss of decision-making authority, and the top level fears loss of control and ROI. The MIT Center for Information Systems Research found that, in the research of the 61 AI implementation failures, the main cause for failure was employee resistance, not technical challenges.
Communication Strategies That Build Understanding
Specific and bilateral communication determines effective change management, and this communication must be continuous. Avoid generic reassurances and clearly communicate the impact of the business change. Don’t just tell employees, but show them the systems and their role by developing AI showrooms. Users should be included in the communication loops that control workflow to influence system changes.
Upskilling and Capability Development at Scale
The AI literacy drive must cover all teams, and executives need to appreciate AI at the conceptual level. Business domain experts need to appreciate AI at the functional level. Technical experts need to appreciate AI at an advanced and functional level. Frontline workers need AI literacy at the operational workflow level.
As evidenced in a McKinsey study on AI use in the distribution operations, set control towers in the distribution networks, in this case, distribution networks of a supplier of building materials, and manage the stock levels in the warehouses, take control of the issues, and encourage collaboration in decision-making. (Source)
10% Technology: Applying Machine Learning and AI Technologies Effectively
Leading organizations establish regular review cadences, quarterly for stable domains and monthly or weekly for rapidly changing environments, to incorporate new data, test alternatives, and adjust to evolving business requirements.
Operationalizing the 70-20-10 Framework at Enterprise Scale
AI transformation across an enterprise is a theoretical exercise and transcends the construction of a new process. Organizations foster sustainable behavioural modifications through a three-stage learning process. The first stage of the process consists of acquiring and understanding primary concepts and frameworks. The second stage involves learning through direct practice at the workflow level. The final stage is reached when habitual new behaviours are firmly established. Most traditional training stops short at this level.
The enterprise-wide implementation of AI needs to be role-based and address four different AI archetypes for enterprise-scale implementation to be effective.
- Shapers: Executives setting the organization's AI vision.
- Leaders: Managers creating conditions for AI to grow.
- Transformers: Team leads reworking workflows around AI tools.
- Frontline Contributors: Individuals using AI in their daily work.
The construction of role-based embedded fluency is a key driver in building momentum throughout the enterprise. The greatest positive disruptions due to AI in organizations are when leaders act in the ways they want to see their teams behave.
In the process of reengineering related workflows to leverage AI, organizations need to understand where the opportunity for high-value disruptive interventions exists across different teams in the organization because most organizations exist at multiple points along the maturity continuum simultaneously.
Success metrics should go beyond just measuring how well algorithms perform. Organizations gain by monitoring how tools are used, the range of their use, how often experiments are conducted, improvements in behaviour like more effective prompting, and business results such as time saved, fewer mistakes, and quicker decision-making.
An organization is ready for AI when business goals steer the use cases and the right people identify problems before picking solutions. Reskilling employees becomes crucial, failures become chances to learn, and both workflows and roles change with technology.
Governance and Measurement for Sustainable AI Transformation
AI governance involves much more than writing policies; it must be practical and applicable. For companies utilizing AI in their operations, they must implement AI governance to measure and affect AI transformation’s impact throughout its lifecycle.
The Risk of Operating AI Without Governance and Measurement
In the absence of AI governance, initiatives will be implemented in ad-hoc manners across the entity, leading to stalled / incomplete initiatives (aka technical debt), compliance violations, and wasted spending. Without the proper measurement, value can’t be derived from initiatives. Meaning, without defined measurement, all that business value cannot be achieved, and initiatives will plateau. Transformation in AI will not be achieved if there are no governance, measurement, and controls in place.
Establishing AI Principles, Governance, and Responsible Practices
Organizations are forced to put ethical points of construction (EPC) in their construction and deployment processes. These principles will include fairness, bias, transparency, data and system safety, accountability, and the mechanisms of the system's construction. An advanced organization of these principles will include, as far as governance is concerned, not being hypothetical praxis. These will include impact assessments (if they exist) and bias assessments (if they exist), along with the processes that respond to conscious human oversight in high-stakes situations and the processes that respond to incidents.
Defining Success Beyond Model Accuracy
Accuracy alone does not equal impact. Effective measurement frameworks track four dimensions: technical performance, business outcomes, adoption rates, and organizational capability growth. If users do not trust or actively use AI systems, even highly accurate models deliver limited value.
Continuous Improvement and Long-Term Value
AI transformation is an ongoing capability, not a one-time deployment. Organizations that sustain value invest in continuous feedback, retraining, skill development, and governance refinement. This ensures AI evolves alongside strategy, regulation, and workforce readiness.
Conclusion: Making AI Transformation Stick
The 70-20-10 model provides a useful counterbalance to other considerations outside of equations, models, or other technical mechanisms. Businesses that look to other third-party stakeholders, as opposed to technology, experience a higher degree of success. The fastest, most effective, and most profitable AI-focused process digital transformation initiatives are those that include people. As a learning opportunity, technology, along with process automation and digital transformation initiatives, should shift workflows.
If you’re looking to shift your organization from isolated AI initiatives to collaborative, system-wide AI initiatives, Get in touch with us. Tredence can assist in the development of a strategy for AI transformations that is people-focused, scalable, and practical.
FAQs
Q1: What is the 70-20-10 rule in AI transformation?
The 70-20-10 rule focuses on people (upskilling, culture) 70%, processes (roadmaps, change management) 20%, and technology (ML tools) 10%, emphasizing human-led scalable growth over tech hype.
Q2: Why is people-first, human-centric AI important for enterprises?
People-first AI causes 89% higher adoption, bias risk mitigation, and ROI improvement by offering data teams clear models and avoiding 70% of cultural resistance failures in complex analytics deployments.
Q3: How can organizations integrate AI into existing processes effectively?
Evaluate workflows, launch high-value use case AI systems with people included in the process, redesign through cross-functional pods, and continue loops of iterative feedback to grow the use of AI from only 10% of processes to enterprise operational systems, all without causing disruptions to the existing supply chains.
Q4: What role does AI strategy consulting play in AI transformation?
Consultants like Tredence provide customised 70-20-10 strategies and roadmaps for governance and the centre of excellence, leading to faster ROI, resolving the gap from the C-suite strategic vision to practical implementation, and achieving 25-40% efficiency in ML-driven business transformation.
