AI success will be determined by your operating model, not your AI strategy
AI is everywhere – new models dropping, new tools, and new (and old) service providers confidently selling AI solutions with ambitious performance claims. If you’re leading in an organisation today, AI is undoubtedly high on your agenda.
The pressure is on, and many leaders are rushing to create AI strategies. But while organisations focus on selecting the right tools and use cases, they're often neglecting something more fundamental: ensuring their organisation is designed to enable AI to deliver real value. AI success won't come from the best strategy – it will come from how well your organisation is structured to adopt, leverage and scale it.
Instead of asking, "What's our AI strategy?", companies should be asking, "How do we work in a way that allows us to unlock AI's value?"
The operating model gap & why product-led approaches matter
Through my work with organisations adopting digital and product ways of working, I've observed a consistent pattern: transformation initiatives tend to fail when they don't align with how work actually happens in a company. And this risk is particularly true with AI, where the gap between technical possibility and operational reality can be vast.
AI is challenging traditional approaches because of its rapid evolution. Unlike more stable technologies, AI doesn’t remain static – its capabilities and applications shift constantly – which drives the need for ongoing adaption. This means traditional planning cycles struggle to keep up, which makes large-scale transformation even more risky.
Organisations that successfully harness AI won’t necessarily be those with the biggest budget or best strategy, they’ll be the ones that have fundamentally rethought how they work. They'll have created operating models that can integrate new capabilities while adapting to rapid external technology change. A product-led operating model – characterised by continuous iteration, experimentation, and real-time learning – creates the right conditions for AI adoption by enabling organisations to:
? rapidly test AI applications in real business contexts
? learn from both successes and failures to refine their approach
? scale what works while quickly abandoning what doesn’t
? adapt to AI’s constant evolution with a culture of experimentation
The product-led approach
Product-led approaches create the right conditions for learning and adaptation in an environment where the technology is rapidly evolving and its applications are still emerging.
Product-led organisations won’t treat AI as a project to complete, they’ll treat it as a capability to develop. They’ll embed AI into teams that solve real business problems and continuously refine solutions based on real-world feedback.
This approach can embed AI capabilities within cross-functional teams who are empowered to solve specific business problems, whether in customer-facing products, internal operations, or support functions. Teams can quickly validate what delivers value before scaling, and leadership can provide air cover for experimentation while maintaining focus on outcomes.
Three pillars of an AI-ready operating model
Organisations preparing for AI success need to build their operating model around three critical pillars – all of which are tried and tested in the product-led model:
1?? Problem-First Experimentation
Core principle: Start with real business problems, not technology solutions. Identify specific pain points where AI can create clear business impact.
A UK-based law firm sought to reduce manual effort in contract review using AI, but needed to ensure legal accuracy wasn’t compromised. Rather than a full-scale rollout, they piloted the AI tool with 10 lawyers, refining it based on real workflow needs, resulting in major reductions in document review time, and firm-wide AI adoption.
This approach requires creating space for continuous experimentation – with small, focused teams testing multiple approaches in parallel, learning quickly from failures, and scaling what works. ?Leaders might champion a culture where rapid learning cycles are valued over perfect first attempts.
2?? Distributed Ownership
Core principle: Embed AI capabilities within business units and have lightweight oversight for governance, tooling, and best practices.
An insurance company’s AI claims processing tool was failing to gain adoption as frontline teams distrusted automated decisions. They pivoted away from a fully centralised rollout and gave claims teams control over AI decision thresholds, resulting in faster processing times and refined AI that better fits real-world cases.
Instead of isolating AI expertise in a central unit, the distributed model can ensure AI initiatives remain connected to business outcomes and support a wider cohort of employees and leaders to increase their AI-literacy.
3?? Outcome-Driven Measurement
Core principle: Measure AI's success based on business impact, not by the scope of deployment.
A manufacturer introduced AI-driven demand forecasting, expecting better accuracy. Instead, complexity increased without improving results, frustrating supply chain teams. Rather than persisting, the team pivoted to explore simpler automation.
Establish clear metrics tied to customer or operational value and continuously measure AI's contribution.
? Wrong Metric: "We deployed 20 AI models." ? Right Metric: "AI reduced manual invoice processing time by 40%."
This requires shifting from output-based metrics to outcome-based evaluation (impact on revenue, efficiency, or customer satisfaction). It might also mean being willing to pivot or abandon AI initiatives that aren't delivering meaningful results, regardless of the investment already made.
Making the shift: practical steps for leaders
Updating your operating model for AI success won’t happen overnight. Here are concrete steps leaders across functions can take:
?? For Executive teams
?? For Technology leaders
?? For People leaders
Challenges in the transition
Shifting to a product-led model for AI isn't without obstacles. Organisations often face resistance from teams comfortable with traditional approaches, struggle with legacy governance and measurement systems that prioritise outputs.
That said, the biggest blocker won’t be technology or talent – it’ll be mindset. Many leaders will expect AI initiatives to follow traditional IT rollouts, with detailed roadmaps, fixed timelines, and clear ROI projections before approving investments.
But AI doesn’t work like that. Early applications may not deliver immediate ROI, but they create learning value – helping organisations discover where AI genuinely adds impact. If leaders insist on rigid planning and guaranteed results upfront, teams face an impossible choice: make unrealistic promises or avoid AI experimentation altogether.
The critical shift for AI success is moving from ‘project completion’ to ‘capability building’, recognising that long-term AI value comes from embedding rapid learning cycles into the organisation.
The competitive advantage of organisation design
As AI technologies become increasingly commoditised and accessible, the real differentiator will be how well your organisation is set up to adopt and scale AI.
The most successful organisations in the AI era won't be those with the most advanced technologies or the most comprehensive AI strategies. They’ll be the ones with the right operating model, culture, and ways of working to unlock AI’s full potential.
At Kindred , we're helping organisations reimagine their operating models for AI. Because how you organise will need to continually evolve in response to whatever you deploy.
About Kindred
Kindred helps organisations achieve growth and impact by establishing the structures, practices and behaviours for people to do their best work.
We guide agility by tailoring solutions for our clients' goals taking incremental steps, not risky transformations. Our adaptable tools and support assess readiness, create common language, establish accountability and metrics, and sets the path to long-term agility. If this sounds good to you, let's talk!
CEO of HIVE Interactive | AI and Human Augmented Intelligence Expert. Learning, Tech and Entertainment Disruptor with a Focus on The Future of Humanity.
1 周Sorry to say this to you, but false. I’m coming from our team having worked to enable over 29,000 people on how to use AI Tools with everything from ChatGPT to Claude, Gemini, Co-pilot, beautiful ai and everywhere in between. Over 84,000 hours of instruction and 200% on average reported increases in productivity. The AI enablement, training on the tools and how humans skill up around communication among the lacking human skills, will be the only path. Tools will come and go, some good and some bad. Most are not intuitive in the slightest. They are terrible at model scope creep and complex. Happy to explain more but even you heavily AI written narrative (the emoticons give it away) tells me that you need to learn more about humans, not just the tech. I say this respectfully.
Maximizing success by leading teams from discovery of needs to measurable value
2 周Yes!
Managing Partner Dissonance Consulting. We develop strategies for growth and programmes for change. Co-Founder LDEC. Helping Londoners create an economy to deliver society's needs within planetary limits
2 周Absolutely spot on. In research we Dissonance Consulting recently conducted, leaders who are in a wide range of positions on the topic of AI adoption commonly suggested that readiness and operating model were two of their largest internal challenges. I know that Kindred is excellently placed to help organisations structure or shift their models to be ready for change.