Kyndryl

IBM announced in October 2020 that it would separate its managed infrastructure business. The unit was named Kyndryl in April 2021 - kyn from kinship, dryl from tendril, a construction that drew a certain amount of press amusement - and the separation completed on 4 November 2021, when it began trading on the New York Stock Exchange with about 90,000 employees. It was, and remains, the largest technology spin-off by headcount.

The scale is easier to grasp from the other side. The Financial Times described it as IBM shedding a quarter of its business, and the unit taking with it the bulk of what had been IBM Global Technology Services: roughly 4,400 customers including around three quarters of the Fortune 100, operations in sixty-three countries, and some four hundred data centres.

The reason usually given is decline, and it is true but incomplete. Inside IBM the business had struggled through the cloud era, because customers were moving workloads to hyperscalers and few had any appetite left for the long, large outsourcing contracts the unit was built around. Shedding a shrinking business to concentrate on hybrid cloud is a legible strategy and it is what the coverage led with.

The more interesting reason is structural, and it is the one worth taking away. A managed services business owned by a cloud vendor cannot credibly recommend a competitor's cloud. Whatever the engineers actually think, the advice arrives from a company whose parent sells the alternative, and the customer discounts it accordingly. Independence was not the consolation prize for being unwanted - it was the thing that made the business sellable again.

The evidence is in the calendar. Kyndryl announced a partnership with Microsoft in November 2021, the same month it separated, and with Google Cloud in December. Neither was available on those terms to a division of IBM.

That is the same argument this timeline records elsewhere, in a different setting. CompTIA's certifications are valuable precisely because they belong to no vendor, and the open question there is whether that survives private-equity ownership. Kyndryl is the mirror image: a business whose advice was worth less while it was owned, and worth more once it was not. In both cases the asset is neutrality, and neutrality is a property of who owns you rather than of what you know.

Martin Schroeter, who had been IBM's chief financial officer and later a senior vice president, left the company in June 2020 and returned in January 2021 to lead the new one. The market's first verdict was cool: the shares closed their opening day at $26.38, down about seven per cent, and slipped further after hours.

And it is the third time IBM appears on this timeline handing something over. It bought Lotus in 1995 for the Notes technology, sold Notes and Domino to HCL in a deal announced in 2018, and separated this business in 2021. A company that spent decades acquiring is now most visible here for what it has let go.

Founding stories

2021

New York · Founders: IBM

Not founded so much as separated: IBM Global Technology Services, given a name in April 2021 and its own listing on 4 November. What made it unusual as a new company was that it began with a complete customer base, a complete cost base and no ability to choose either.

The timeline

The announcement
October: IBM says it will separate its managed infrastructure business.
Named, listed, partnered
Named in April, separated 4 November, and into hyperscaler partnerships within weeks.
The three-A's
Alliances, Advanced Delivery and Accounts: grow revenue through hyperscaler partnerships, automate delivery so people can be redeployed rather than replaced, and fix or exit the customer relationships carrying substandard margins. An unglamorous three-part programme, and the whole of the turnaround.
The thesis, measured
Fiscal 2025 produced $1.2B of revenue tied to cloud hyperscaler alliances - more than double the prior year, and past the target. That figure is the neutrality argument expressed in money: it is the business the company could not have written while IBM owned it.
Signings above revenue
Trailing-twelve-month signings of $15.4B to $15.6B against revenue of about $15.0B, hyperscaler revenue heading for a $1.8B target, and Kyndryl Consult growing in the mid-twenties per cent while approaching a quarter of the business. In the third quarter alone, eleven contracts over $50M each.

Flagship products and solutions

Managed infrastructure servicesThe original business: running mainframes, networks, storage and data centres for organisations that will not run them themselves. Unfashionable, enormous, and the reason three quarters of the Fortune 100 were customers on day one.
Kyndryl ConsultAdvisory and implementation, and the deliberate move up the value chain - now around $3.6B of revenue, growing in the mid-twenties per cent, and the part of the business that decides whether the company is a contractor or an adviser.
Kyndryl BridgeThe operating platform: the delivery telemetry from thousands of managed estates turned into a product customers see. It is what an outsourcer has that nobody else does - observed data about how large estates actually behave.
Hyperscaler practicesDedicated practices for the major clouds, staffed and certified separately. The point is that all of them exist at once, which is the entire proposition.
Security, resiliency and network practicesThe specialist lines sold across the same customer base - the pattern of a services firm growing by depth rather than by new logos.

Key innovations

Neutrality as a balance-sheet itemThe argument that an independent adviser can recommend any cloud is easy to make and hard to prove. Here it is measurable: hyperscaler-linked revenue more than doubled in a single year and is running toward $1.8B. That is not brand positioning, it is business that could not previously have been written.
Automating delivery to redeploy rather than to cutThe Advanced Delivery initiative freed staff to serve new work and backfill attrition rather than being made redundant. Whether that framing survives contact with a downturn is untested - but a labour-intensive business that automates its own delivery has to decide what the savings are for, and this one said so out loud.
Fixing or leaving unprofitable contractsThe Accounts initiative meant renegotiating or exiting business the company had inherited at bad margins. Shrinking revenue on purpose is difficult inside a public company and nearly impossible inside a division of a larger one, where the revenue line belongs to somebody else's story.
Selling the operations data backBridge productises what the company sees across thousands of estates. An outsourcer's real asset is not its people but its observations, and turning those into something the customer can look at is the difference between selling hours and selling knowledge.

Main markets

Large enterprises and governments in more than sixty countries, at around $15B of annual revenue - the organisations whose technology estates are too large, too old or too regulated to move wholesale, which is a durable position for as long as those conditions hold.

It competes with the global integrators and consultancies, with the hyperscalers' own professional services, and with the offshore service providers on cost. Its distinguishing claim is the one the separation created: it can be credibly indifferent about which cloud a customer chooses.

Analyst standing

Signings running above revenue is the measure worth watching in a services business, because it is the only leading indicator of a company whose revenue is contracted years ahead. On that measure the trajectory has been positive since the separation.
The verdict on the original question is now largely in. Five years after a cool opening, the fastest-growing lines are the ones that depend entirely on being independent. Whether the business would have declined more slowly inside IBM is unknowable - but the business it has written since leaving is not business it could have written while it stayed.

Kyndryl ↗

From the company

Official site ↗

Sources