Executive Summary
Between 70% and 95% of AI implementations fail to meet their objectives—but the problem isn’t the technology. Drawing on research from Gartner, Kotter International, and neuroscience-based frameworks, Barry Eustance (The Sixsess Consultancy) and Mark Blackwell (Arkaro) unpack why AI adoption stumbles and how people-centric change management turns failure into momentum.
The core findings:
- AI projects fail on people factors, not models or mathematics
- “Corporate FOMO” drives poor implementation decisions
- Success depends on six human needs: self-esteem, purpose, autonomy, certainty, equity, and social connection (SPACES model)
- The PEOPLE framework provides structured guidance: people-centric leadership, empowerment, optimisation, purpose-driven vision, and learning
- ROI often appears first in reducing administrative load
- Organisations must solve real bottlenecks with their teams, not impose solutions around them
Watch the full conversation to understand how to move from AI hype to sustainable habit.
Most AI Projects Don't Stumble on Models or Maths—They Stumble on People
“Well, 70 and 90%, 95% of AI implementations fail,” Mark Blackwell states. It’s a figure backed by multiple research sources including Gartner and MIT. The headline shock isn’t the technology—it’s what happens when organisations forget that humans, not algorithms, determine whether AI delivers value.
Before we go further, let’s be clear: we are not anti-AI. As Mark emphasises, “We are not Luddites… This is something that organisations must have for the future.” Both of us use AI extensively. The question isn’t whether to adopt AI, but how to avoid becoming part of that 95% failure statistic.
The culprit? “Corporate FOMO”—fear of missing out on the AI gold rush. CEOs buy shiny new tools without proper strategy, persuaded by vendors and anxious boards. But as our conversation reveals, this “AI gold rush is a bit of a fool’s gold rush.”
The SPACES Framework: Six Human Needs That Predict Success or Failure
To understand why AI implementations fail, we need to start with how humans experience change. Hilary Scarlett’s SPACES framework—drawn from neuroscience—identifies six fundamental human needs: self-esteem, purpose, autonomy, certainty, equity, and social connection. As Mark notes, “These factors all come from imagining who we really are, which is running around the savannah plain two million years ago.”
When AI implementation tramples on these needs, resistance isn’t irrational—it’s predictable.
Self-Esteem: “You’re Asking Me to Put My Credibility on the Line”
The obvious threat is job loss, but research reveals deeper issues. Insufficient training leaves people feeling inadequate: they’re asked to use tools they don’t understand. Mark highlights a critical insight: “People are concerned… you’re asking me to put my credibility on the line and make recommendations from this AI tool, which I know is nowhere near as credible and reliable as I am as an individual.” When AI hallucinates, employees lose confidence—not just in the technology, but in themselves.
Purpose: Lost in Translation
“If we’re not aligning the potential of what AI can do to the commercial direction of where the organization wants to get, we’re missing the point,” Mark explains. When leadership can’t articulate why AI matters beyond buzzwords, employees don’t see how their work contributes to something larger.
Autonomy: Shadow AI and Pilot Hell
Heavy-handed mandates create “shadow AI”—employees secretly use consumer tools because official systems don’t work.
Certainty: A Tidal Wave of Change
AI arrives alongside countless other initiatives, creating “change saturation.” When people face too many simultaneous transformations, predictability evaporates and resistance hardens.
Equity: Winners and Losers
When AI benefits some departments whilst burdening others, resentment builds. Process changes that help one team create extra work elsewhere.
Social Connection: Isolated Implementations
When IT imposes AI solutions without involving the people who’ll use them daily, social connection fractures. Implementation becomes something done to people rather than with them, and the collaborative problem-solving that builds ownership never happens.
The PEOPLE Framework: Rebuilding the Playbook
If SPACES shows where implementations fail, how do we succeed? The answer lies in people-centric change management. Mark references the Arkaro approach: “Understand what people want, co-create solutions with people, enable capabilities, and sustain the change.”
People-centric Leadership treats objections as design inputs. “It spells people,” Barry notes. “It’s simple messaging.”
Empowerment brings process owners into problem selection. Let the people who live with bottlenecks identify where AI can help.
Optimisation focuses on clean data and a “dual structure”—pairing operational hierarchy with volunteer innovation networks.
Purpose-driven Vision ties local wins to a unifying story connecting AI to commercial direction.
Learning closes the loop with short cycles and public celebration. “Build, measure, learn, pivot,” Barry emphasises. “Fail fast, fail cheaply.”
Embed ensures the change sticks. The new ways of working must become “how we do things around here” rather than this month’s initiative.
From Hype to Habit: Where ROI Actually Appears
ROI often shows up first in reducing administrative load, not headline transformation. AI that eliminates tedious report generation or automates routine approvals delivers immediate value that builds credibility.
Shadow AI becomes an asset when organisations create safe standards. Instead of banning consumer tools, provide enterprise alternatives with proper governance.
The path from pilot hell to scaled adoption requires “making sure this sticks.” Cultural change happens when “a small group of people doing things differently” demonstrates success. You can’t mandate culture, but you can cultivate the conditions where it grows.
AI Is Not the Enemy. Mismanaged Change Is.
As our conversation concludes, Mark articulates the central message: “Let’s bust this 95% and give it 95% success.”
AI implementations fail because organisations ignore human psychology, impose solutions without consultation, neglect training, and drown people in change saturation. Success requires treating the people side with the same rigour as technical architecture.
Start where the work lives. Involve process owners in selecting problems. Build capability through training. Create dual structures balancing stability with innovation. Connect AI to commercial purpose. Learn in short cycles, celebrate progress, and let culture shift naturally.
Build with people, not around them.
About the Speakers
Barry Eustance is Founder of The Sixsess Consultancy and creator of the PEOPLE Change Management and Leadership Framework. His work focuses on people-centric approaches to organisational transformation.
Mark Blackwell is Founder at Arkaro, specialising in change management, innovation, and commercial excellence for the agriculture, food, and chemicals industries. Arkaro’s collaborative “do it with you” approach works closely with clients through a four-step process: Understand, Co-create, Enable, Sustain.
Don't Become Part of the 95%
Your AI investment is too important to fail because of preventable people issues. Whilst your competitors struggle with resistance, shadow implementations, and abandoned pilots, you could be building momentum through people-centric change management.
The evidence is clear: technical capability alone won’t deliver results. The organisations that succeed treat change management with the same rigour as their AI architecture. They build with their people, not around them.
Are you ready to reverse the failure rate?
If you’re planning AI implementation—or rescuing one that’s stalled—let’s discuss how the Arkaro approach can help you avoid the 95% failure trap. We work with organisations in agriculture, food, and chemicals to turn AI from a boardroom buzzword into a competitive advantage your teams actually use.
Related Reading from Arkaro
On AI Implementation & Neuroscience:
- Why 95% of AI Implementations Fail | Neuroscience & Change – Comprehensive analysis of how the SPACES framework explains AI failure, with extensive research and data
- People-Centric Change Leadership and AI – Deep dive into the PEOPLE Framework for AI implementation success
On Change Management & Implementation:
- Why ADKAR and the Arkaro Approach Are Essential Partners – How collaborative change management creates lasting transformation
- Arkaro’s Approach to Culture During Change Initiatives – Building change roadmaps that incorporate cultural and behavioural aspects
On Strategy & Execution:
- Make Strategy Work: The Power of Integrated Business Planning – Moving strategy from slides to sustainable execution
- Strategy in a VUCA World: Emergent Approach Guide – Adaptive approaches for complex, uncertain environments
- Exposing Bottlenecks: The Key to Building a Great Strategy – Root cause analysis for better strategy design
External Resources & Further Reading
- The Sixsess Consultancy – PEOPLE Change Management & Leadership Framework
- PEOPLE Framework Course – Structured training in people-centric change leadership
- Hilary Scarlett’s Neuroscience of Organization Change – The S-P-A-C-E-S framework explained
- Julie Hodges on People-centric Organizational Change – Academic foundations for human-centred transformation
- Kotter International – Leading research on change management and AI implementation
Stay connected: Follow Arkaro Insights newsletter for regular perspectives on change management, innovation, and commercial excellence.
