YOURAIEFFICIENCYGAINSAREBEINGCOMPETEDAWAY

The HBR article mentions a powerful analogy. When the camera was invented, portrait painters could not adapt by simply deploying the new technology to paint faster. Photography had commoditised the entire offering; customer willingness to pay collapsed and the value pool dried up.

AI is doing the same thing to efficiency-based advantages.

If your AI strategy is to automate reports, speed up document processing, or reduce response times, you are painting faster. The problem is not the quality of your brushstrokes. The problem is that the value of the painting itself is shrinking. When every competitor uses the same AI tools to achieve the same efficiency gains, those gains no longer differentiate. They become the minimum requirement to stay in the game.

To be clear: efficiency matters. Of course it does. A company drowning in manual processes absolutely should automate. We build these systems daily. But efficiency as the endpoint of your AI strategy is where the danger lies.

What happens in practice: Company A invests in AI-driven automation. Costs go down by 20%. Margins temporarily improve. Then Companies B, C, and D adopt identical tools from the same vendors. Within 12 months, the entire sector has lower costs. Prices adjust. Margins compress back to where they started. The efficiency gain is real - but its captured by customers, not by the companies that made the investment.

The BCG Henderson Institute data across 800 companies shows this pattern already playing out. Sectors with the highest automation potential are not translating that potential into profit. The gains are being redistributed, not retained.

For the past two years, the dominant AI ROI conversation has been: “How much time did we save?” That question is almost expired. Now we must ask “What new revenue line did we build?”

Choudary and the BCG Henderson researchers argue that durable advantage comes from using AI to create new categories of value, not to speed up old ones. This means identifying new frictions in your market, building novel business models, and repositioning your company within an AI-powered ecosystem where new value pools are forming.

The 5% of companies that BCG identifies as “future-built” share this exact characteristic. They moved beyond efficiency. They used AI to fundamentally change what they offer, how they operate, or how they create value for customers. They allocated over 80% of their AI investment to reshaping key functions and inventing new offerings, not to incremental optimisation.

Value creation through AI is not abstract: it’s specific to your business, data and operations. And this specificity is exactly what makes it so defensible.

Consider this example:

Efficiency play: You use AI to generate sustainability reports 40% faster. Your competitors adopt the same tool. Speed is no longer a differentiator.

Value creation play: You build an AI system on 15 years of proprietary reporting expertise and 750+ completed reports. The system does not just generate documents faster. It embeds domain knowledge that generic tools cannot replicate, and creates a new category of service that didn’t exist before.

Generic tools commoditise. Custom systems on unique data compound.

If efficiency has become the new norm, where does actual competitive advantage come from? Based on what we observe by building AI systems across different industries, three patterns stand out:

1. Proprietary intelligence layers:

   Your business generates data that no one else has. Customer interactions, operational patterns, service histories, pricing dynamics. AI built on this data creates insights and capabilities that are unique to you. A competitor using off-the-shelf tools cannot replicate what your data teaches a custom model.

2. Workflow integration, not workflow acceleration

   The most impactful AI systems do not sit on top of workflows. They become the workflow. When AI is integrated into how decisions are made, how operations run, and how customers experience your service, it creates switching costs. The system becomes inseparable from the value it creates.

3. New service categories

   AI enables services that were previously unattainable due to cost, complexity, or scale. Real-time route optimisation for waste collection, intelligent claims processing that reduces 15 manual actions to 3 checks, AI phone assistants that ensure zero missed customer calls. These are not efficiency improvements on existing services. They are entirely new value propositions.

Most AI strategies are built around reducing cost. That is easier to quantify, approve, and measure. It also has a ceiling.

The harder work is asking: what can we offer that did not exist before? What decisions can we make that were previously impossible? What business model becomes viable when AI removes a constraint we always assumed was fixed?

These questions require deeper thinking than selecting an AI vendor or piloting a chatbot. They require understanding your own operations, data, and market position at a level that most generic AI tools simply cannot reach.

This is why custom AI built on proprietary data and workflows consistently outperforms off-the-shelf solutions. Not because the technology is more advanced. But because the system is built around the specific problems and opportunities that define your competitive position.