BCG's Build for the Future 2025 study — one of the most rigorous annual AI adoption surveys in existence, covering 1,250 firms across 25 sectors — landed a number that I want you to sit with for a moment.
60% of companies are generating no material AI value. Not "less than expected." Not "early stage." Zero material value. Revenue flat. Costs unchanged. Process impact negligible. Despite meaningful investment.
A further 35% are scaling up but admit they are not moving far enough or fast enough.
That leaves 5% — what BCG calls "future-built" — who are pulling away: 1.7x more revenue growth, 1.6x higher EBIT margins, 3.6x three-year total shareholder return.
Source: BCG Build for the Future 2025, n = 1,250 senior executives.
I don't find the 60% number surprising. I find it validating. Because it is exactly what I see in Indian contact-centre AI programmes every quarter.
What "no material value" actually looks like
It does not look like a company that has done nothing. Companies in the 60% camp are usually busy. They have a chatbot. They have a "GenAI task force." They have a vendor demo that everyone agreed was impressive. Some of them have twelve active pilots running simultaneously.
What they do not have is a business case that has closed. They do not have a P&L line that moved. They do not have a workflow that has been fundamentally redesigned — end to end — around AI.
The pattern I see most often: AI is owned by middle management. A digital transformation manager, maybe a COO minus one. Occasionally a well-intentioned CTO who is also responsible for keeping the data centre running. These people are competent. They are not empowered.
BCG is precise about this: nearly 100% of future-built organisations have deeply engaged C-suites. Only 8% of laggards do. That gap — 92 percentage points — is not a technology gap. It is a governance gap.
The Indian CCaaS pattern
In the contact-centre technology space specifically, the 60% problem has a characteristic shape.
A company — an insurance firm, a bank, a large BPO — decides to "do AI in the contact centre." The trigger is usually a vendor conversation. Genesys comes in with an AVA demo. Google CCAI runs a workshop. Someone has read about Cognigy and wants to explore it.
A proof-of-concept is agreed. A specific queue is selected. Results are encouraging — containment rate goes up, handle time drops, the team is pleased. The pilot is declared a success.
And then it stops.
The pilot stays a pilot for eighteen months. A second pilot starts in a different queue, run by a different team, with a different vendor. Neither integrates with the other. Neither connects to the CRM workflow that sits upstream. Neither changes how supervisors manage quality. Neither affects the training curriculum for new agents.
The contact centre has "AI" but the contact centre has not changed.
This is not a technology failure. The technology worked in the pilot. The failure is structural: nobody owned the end-to-end workflow redesign. No one had the mandate — or the P&L accountability — to make it happen at scale.
Three signals that separate future-built from laggard
After fifteen years in contact-centre architecture, here is what I look for when I walk into an organisation and want to quickly assess which camp they're in.
Signal one: Who presents the AI programme to the board?
In laggard organisations, it's the CTO or a digital transformation lead. In future-built organisations, it's the CEO or COO — and the framing is revenue and cost, not technology.
BCG's data shows future-built companies are 12x more likely to have C-level executives deeply engaged with AI. Not "informed." Engaged — meaning they set targets, they review progress monthly, they ask uncomfortable questions about why value hasn't arrived.
Signal two: How many active AI initiatives are running?
Counter-intuitively, the 60% camp often has more initiatives. They experiment widely. The 5% focus narrowly — two or three core workflows, pursued to full deployment.
BCG found that 62% of AI initiatives at future-built companies are already deployed or at scale, versus 12% at laggards. That is not a speed difference. It is a discipline difference. Future-built companies say no to most AI opportunities so they can say yes properly to a few.
Signal three: Is AI owned by IT or by the business?
This is the most reliable single predictor I have found. When I ask "who owns this AI programme?" and the answer is "IT" — or worse, a vendor — the programme is almost certainly in the 60% camp.
Future-built companies have joint business-IT ownership with clear decision rights and shared accountability. The business owns the outcomes. IT owns the architecture. Neither can proceed without the other.
In the contact-centre context: the operations director owns agent productivity targets. The technology team owns the Genesys or Cognigy configuration. When those two functions sit in the same room, with aligned incentives, things move.
What to do about it
If you recognise your organisation in the 60%, BCG's research and my own experience point to the same sequenced response.
Start with governance, not technology. Appoint a C-level AI sponsor with a specific value mandate — cost reduction percentage, revenue impact, containment rate improvement. Make that target visible. Review it quarterly in the same forum where you review the P&L.
Cut the pilot portfolio. If you have more than three active AI initiatives with no deployment timeline, you have a portfolio problem, not a capability problem. Choose the two workflows with the clearest value case — in a contact centre, that usually means intelligent self-service and agent assist — and drive them to full deployment before starting anything new.
Redesign the workflow, not just the technology. The contact centre that adds a chatbot in front of the same IVR, with the same transfer logic, with agents handling the same escalations the same way — has not changed. The contact centre that redesigns the queue architecture, redefines what escalation means, retrains supervisors to manage AI-assisted agents — that organisation will show value in nine to twelve months.
Measure the right things. Containment rate is useful. Cost-per-contact is useful. But the number that matters to a future-built organisation is the P&L impact: what did this workflow change contribute to EBIT? Until you can answer that question, you are still in the 60%.
The value gap is real and it is widening. Future-built companies are reinvesting their AI gains into additional capabilities, which accelerates their advantage further. Companies that stay in the 60% camp are not standing still — they are falling behind at an accelerating rate.
The technology is not the barrier. It never was.
Data in this article draws on BCG's Build for the Future 2025 global study (September 2025), covering 1,250 senior executives across 25 sectors and 68 countries. Courtesy: Boston Consulting Group.