I ran Accenture projects in Asia in the early “focus on quality” days. One of the clients was AS Watson in Hong Kong. The projects worked, the people were good, and many of the Accenture leaders I met then became long-term friends across the business.
Back then, Accenture was not yet the mass BPO and outsourcing machine it later became. That business was being invented while I was there. I still recall the first outsourcing expert hired from EDS, and the executive strategy sessions with him on how to build a BPO business and take it across key clients.
So the latest headline lands personally.
Accenture is reportedly exploring the sale of its Greater China business — mainland China, Hong Kong and Taiwan — in a process that could value the unit at up to $1.5 billion. JPMorgan has reportedly been hired, Accenture may retain a significant minority stake, and no transaction is guaranteed. (Reuters · FT)
Financially, I understand the logic. Strategically and emotionally, it looks like giving up.
The firm that scale built — later
Accenture’s modern empire was built on a simple proposition: we can put more capable people on your problem than anyone else.
Today that means approximately 814,000 people, around $74 billion in annual revenue, and roughly 9,000 clients. (Accenture Fact Sheet)
Mass BPO and outsourcing became the engine over the last few decades. Accenture became better than almost anyone at taking over complex operations, industrialising them, and running them at global scale. In the latest quarter, managed-services revenue was $9.40 billion, against consulting revenue of $9.28 billion — operations is roughly half the firm. (Accenture FY26 results)
Accenture Song, the creative and innovation business, is a strong idea. Helping clients invent the products and services of tomorrow is a genuine need. But Song cannot simply substitute for the economics of an 800,000-person operations company. And fusing Song to the old engine room is not the answer — the engine room, in its current form, is dead.
The tide has turned
AI agents attack the unit the outsourcing model was priced on: the billable human hour.
Agents now resolve customer contacts, process documents, reconcile exceptions, update systems and escalate the cases that need human judgement. The work does not disappear — but the number of people required to deliver the same outcome falls.
Accenture knows this, and to be fair, it already cannibalises its own model in part: it reported approximately $11.5 billion in cumulative advanced-AI bookings across roughly 11,000 projects by the end of FY26. (Accenture FY26 results) But it could do far more, far faster.
There are not enough genuine subject-matter experts to reinvent thousands of companies at once: people who understand credit operations, retail supply chains, claims, payments, core platforms and how to rebuild them around AI agents. Accenture has scale in people. The market now wants scale in judgement.
That pressure is sharpest in Asia. This region has an affinity for faster, more rapid innovation — Greater China most of all, but the rest of Asia too. When clients innovate quickly, firms with real experts gain leverage; firms selling scale lose it. Add slower China growth, data-security regulation, political scrutiny and clients choosing domestic providers, and the old integrated global model starts to bite. Other Western consultancies have already cut back in China. (Consulting.us · SCMP)
The replacement is not a smaller Accenture
This new expert-driven, AI-first world needs a different structure.
First, capability must be built in-house. Data, agents and the decision layer cannot remain rented permanently from vendors and integrators. If someone else owns the intelligence, they own the learning — and eventually they own the margin.
Second, companies will need niche firms with real subject-matter experts. Not armies of generalists, but small senior teams who have run the bank, the retail platform or the core system themselves, and can transfer that judgement into the client’s own organisation. That is the space firms like mine, Pivot Digital, were built for.
Third, IT has to stop behaving like a budget to be spent and start behaving like P&L to be invested. Too much enterprise technology is still justified by annual budgets, vendor roadmaps and project completion dates. The better questions are: what return does this create, over what horizon, and what capability remains when the external firm leaves?
Increasingly, IT can also ask:
“What features can our own team generate — this quarter, on our own data, inside our own decision layer?”
Build capability once and it compounds. Rent it forever and it does not.
At Pivot we summarise the model as: rent less, own more, grow new.
Time for tough calls
In my view, Accenture — and every large services firm — now has to:
- Excel in innovation and creative. Help clients strategise the business of tomorrow. There is a true need, and it is worth more than delivery capacity.
- Provide true expertise where it matters: strategic IT effectiveness, the approach to agentic AI, cybersecurity, strategy and operating models. Depth, not coverage.
- Cannibalise harder. Accenture already does this in part. Outcome-priced agent services must replace FTE-priced operations, even when it hurts this year’s headcount story.
- Innovate the inside of the firm itself. Consultants should work with new agentic tools, plugged into clients’ own agentic platforms — not coordinate armies of people doing work agents now do.
- Integrate differently, and measure differently. Embed expert teams with clients, share risk, and judge consulting experts on client outcomes — P&L impact, risk reduced, cycle time cut — not on how many people they sold.
- Don’t leave Asia. Change the business model. Asia innovates too fast to abandon. But the model must become expert-led and locally rooted, not scale exported from headquarters.
- In China: reinvent the business, don’t just keep a flag. A minority stake in the old model is financial engineering. A reinvented, local, expert-led innovation business would be a strategy.
The winners in this era will not be the firms with the most people. They will be the companies that build intelligence in-house, surround it with genuine experts, and treat technology as a long-term investment in the P&L rather than a budget to be consumed.
Time for tough calls. Not by putting more bodies against the tide.
—
Axel Winter, CEO of Pivot Digital