Why AI Pilots Need Clear Direction and How to Decide What’s Worth Building 
If you already have AI initiatives running, your opportunity is no longer about technology.  Your opportunity is to decide which initiatives deserve investment and which ones would benefit from redirection or refinement.  

The opportunity cost of running everything 
Every team has ideas for AI. That doesn’t mean every idea deserves equal funding.  Mats Stuhrman, Data & AI Advisory Lead at Nexer Insight, sees organisations constantly running 10–20–30 AI initiatives in parallel without a clear framework for comparing their potential. 
This creates: 
– Dispersed investment across too many fronts 
– Teams needing clearer direction 
– Unrealised business impact potential 
– Security and compliance considerations that become more complex later 

And when leadership asks, “What are we getting out of this?” there’s often no data-backed answer. 

What successful organisations do differently 
Organisations that generate substantial value from AI excel at four key practices: 
1. They ensure every idea answers one question 
2. They validate feasibility before investing 
3. They compare all AI initiatives side by side 
4. They ensure business and IT decide together 

1. They ensure every idea answers one question 
“What will this change in the business?” 
If the answer is unclear, it’s not a priority. High-performing organisations are more than three times as likely to say their organisation intends to use AI to bring measurable change. They prioritise based on business impact and not technical novelty. 

2. They validate feasibility before investing 
Most AI initiatives that stall don’t fail because the concept was flawed. They stall because the delivery path wasn’t validated early enough. 
David Österlindh, CEO of Nexer Insight, often sees initiatives where: 
– A third party owns the data 
– The system is outsourced 
– The internal skills aren’t developed yet 
– Security requirements weren’t considered from the start 
– Compliance obligations, including the AI Act and GDPR, were not included in the design 

Unfortunately, these factors aren’t discovered until resources have already been committed. 

Before building anything, you must validate: 
– Do we have access to the data we need? 
– Can our systems support this use case? 
– Can we run this in production safely and compliantly? 
– Does this create security or regulatory considerations? 

If not, the opportunity should be redesigned or deprioritised. 

3. They compare all AI initiatives side by side 
When AI lives in silos, strategic value remains hidden. Marketing has its pilot. Operations has another. IT has its own roadmap.  Mats’ experience shows that value appears when organisations evaluate all initiatives together, comparing: 
– Business impact potential 
– Technical complexity and requirements 
– Security and compliance considerations 
– Resource requirements and capabilities 

This makes it clear where the highest-value opportunities are and where investment could shift. 

4. They ensure business and IT decide together 
AI initiatives struggle when business and IT pursue different objectives. 
Business seeks growth and innovation.  
IT prioritises stability, security, and governance. 
Both perspectives are essential. 

David Österlindh, CEO, Nexer Insight, explains, “AI succeeds when these groups make decisions together and share accountability.” 

How to move from exploration to value creation 
This is exactly what the AI Impact Lab is designed to support.  We bring your business and IT teams together in a structured workshop to: 
– Map all your AI opportunities in one place 
– Validate them against reality (data, systems, capabilities) 
– Prioritise by impact potential, feasibility, and risk 
– Identify security and compliance requirements early 
– Build the business case before you make major investments 

You leave knowing what to build first, what needs more groundwork, and what to defer, with evidence you can present to leadership. 
Then, when you’re ready to scale, the AI Strategy & Roadmap helps you: 
– Align leadership on direction and priorities 
– Build appropriate governance and security architecture 
– Create a phased implementation plan 
– Ensure compliance (AI Act, NIS2, GDPR) from the foundation 

This is how organisations move from experimenting with AI to generating measurable business value from it.