The Software Crisis Is Real -- Lessons from 5 Multi-Billion-Dollar Failures

5 Massive Failures, 1 Industry-Wide Wake-Up Call

Hey Friend,

Welcome to the Scrum Master Toolbox Podcast - Software Leadership Newsletter: Where we learn to create better software, and better software teams!

In this newsletter edition, I have for you a reflection that I believe we should undertake as software practitioners. I list 5 incredible software failures that illustrate something deeply flawed in how we organize and execute software.

When five countries lose billions building software—and people go unpaid, systems go dark, and doctors beg regulators for help—it’s not just failure. It’s a failure to learn. The real cost of software failures is starting to affect our societies, not just companies!

These aren’t isolated disasters.

They are a pattern.

A warning. And a call for our community to do better, together! 👫👫👫

These aren’t just technical failures.
They’re multi-billion-dollar collapses! Costlier than new schools, hospitals, road systems, and the salaries of thousands of teachers, nurses, police officers, and public servants combined.

And yet… they happened.

Let’s take a closer look, starting with one that shook Finland’s healthcare system to its core -- and some would argue, still is.

🇫🇮 Apotti (Finland): When Design Ignores the Front Line

💸 Cost: €626 - €950 million (1), (2)

The Apotti project set out to revolutionize healthcare IT in Finland by unifying patient records and social services for over 1.7 million people. Instead, it became one of the country’s most criticized and costly digital missteps.

Despite the ambition, clinicians were left out of critical design decisions. What followed were years of confusion, mounting frustration, and even threats to patient safety:

• Medication lists that didn’t work as expected

• Basic tasks requiring dozens of clicks

• Lost or misrouted test results

• Integration failures with national systems

In 2022, over 600 doctors petitioned Finland’s healthcare watchdog to abandon the system altogether. The country’s Safety Investigation Authority launched a formal probe after multiple incidents endangered patients. Apotti’s own CEO later admitted: “The goals were partly unrealistic, and user satisfaction is not at an adequate level.”

📘 Immediate Lesson for Us:

If users aren’t part of your design process, they’ll be part of your failure story.

→ In your projects:
Test early, test often... with real users. If your software creates friction for the people delivering care, teaching students, or protecting communities, it’s not a system, it’s a liability!

🇸🇪 Millennium (Sweden): When the Launch Comes Before the Readiness

💸 Cost: €500+ million (3), (4)

Sweden’s Millennium EHR project was intended to unify digital records across major healthcare regions using Oracle Cerner’s platform. But what was supposed to make handling information more efficient, turned into a chaotic three-day rollout that sent hospitals scrambling, and ultimately forced the system offline. You read that right! They went LIVE, only to switch it off in 3 days!

The problems were immediate and severe:

• 📝 Staff reverted to pen and paper after losing access to critical records

• Training was incomplete, and many users hadn’t even seen the system before launch

• A KPMG audit found the organization of the Millenium project “ill-prepared and lacking leadership” (they would not need to pay KPMG for this conclusion, if you ask me)

Medical staff protested in waves. One doctor described it as “a hamster wheel of hell.” The system was quietly paused, with hundreds of millions already spent.

📘 Immediate Lesson for Us:
You don’t “fix” your way into a running system. You MUST START with a running system that does something valuable, even if small, and grow it from there.

→ In your projects:
Ask this question: “why can’t we release today?” Whatever that is, FIX IT NOW! Don’t wait for the release date to do that. Do this every day, and soon you will be ready to release. Never stop being ready to release!

🇬🇧 NHS National Programme for IT (United Kingdom): The Vision That Outgrew Reality

💸 Cost: Over £10 billion (~€11.6B) (5), (6)

Launched with the promise of transforming healthcare across England, the NHS National Programme for IT (NPfIT) was billed as the most ambitious civilian IT project in history. Instead, it became a case study in overreach and inflexibility.

What went wrong? A lot:

• Over-centralized decision-making with little regional autonomy

• Repeated vendor conflicts and high-stakes contract disputes (this one is right there in the Agile manifesto as a stark warning: when we prioritize contract negotiations, we kill collaboration!)

• A vision that couldn’t evolve with new insights or on-the-ground realities (Waterfall anyone?)

Despite years of investment, the program was formally dismantled in 2011. One parliamentary report called it “the biggest IT failure ever seen.”

📘 Immediate Lesson for Us:
Big visions need agile minds, and small increments! No amount of ambition will be a substitute for adaptability.

→ In your projects:
Plan to change the plan! In classic Agile manner, make sure adaptability is part of the base way of working! Build in feedback loops—early, often, and across every level. Otherwise, you risk locking into the wrong thing, at the wrong scale, for far too long.

Before going on to disaster #4, please take a moment to consider the human potential, and the many schools, roads, waterways not maintained or built, and the amount of value-adding jobs to society that these IT disasters evaporated! And that’s just 3 of the 5 I have to share with you today.

Strap on, there’s an even bumpier ride ahead... 🎢🎢🎢

🇺🇸 Kmart IT Overhaul (United States): When Technology Becomes the Scapegoat for Strategy

💸 Cost: $1.4 billion + ~$600 million follow-up effort (7), (8)

In the early 2000s, retail giant Kmart attempted to leap into the future with a massive IT modernization—its largest tech investment ever. Instead, it accelerated the company’s collapse into bankruptcy.

The effort was plagued by:

• A “big bang” rollout that overwhelmed internal systems

• A dysfunctional supply chain platform that underperformed immediately

• Leadership churn—five CIOs in eight years—and a critical vacuum at the top during the crisis. It’s almost as if no one wanted to take responsibility for fixing. Or perhaps, nobody knew, how to fix it.

In January 2002, Kmart filed for bankruptcy. Analysts pointed directly to the failure to modernize IT effectively as a key factor.

📘 Immediate Lesson for Us:
Big bang modernization of IT systems is unsustainable (aka “rewrite projects”), and no amount of IT can replace a bad strategy. If anything, it accelerates the catastrophic consequences of a bad strategy.

→ In your projects:
Start small, aim to grow big. Solve one problem at a time, build systems that work together, but not as a BIG BANG, rather start with a small subsystem, deliver every day into production (or a production like system), and grow from there.
Oh, and if your strategy is unstable, your tech will only make it more visible… and more costly. Fix strategy first!

🇨🇦 Phoenix Pay System (Canada): When Cost-Saving Becomes Catastrophic

💸 Cost: Over C$3.5 billion as of 2024 (9), (10)

Originally intended to simplify payroll for Canada’s federal employees, the Phoenix system became an enduring disaster. And it had real impact, for real people: it left tens of thousands of public servants unpaid, underpaid, or paid incorrectly for years. Now that’s a case of “IT messing with people’s livelihoods” if I ever saw one!

What went wrong?

• The system went live despite unresolved bugs and minimal testing. Sound familiar?

• Cost-cutting was prioritized over training and verification. I’ve heard this one before, have you?

• Backlogs of over 400,000 unresolved pay issues stretched on for years. That’s a REALLY BIG backlog! #NoBacklogs

• Auditor General’s quote: “An incomprehensible failure of project management and oversight.” They forgot to specifically call out “IT and software” in that quote, but I’ll forgive them for that.

Phoenix didn’t just fail. It broke trust at a national scale—and Canadian taxpayers are still footing the bill.

📘 Immediate Lesson for Us:
What might have worked for other industries does NOT WORK for software: Cutting corners isn’t efficient—it just delays the cost until it’s bigger, messier, and more visible.

→ In your projects:
Going live MUST be part of your incremental strategy. We don’t just develop incrementally, we MUST take systems into use incrementally. BIG BANG or Waterfall don’t work for software. Start small, and grow from there also when it comes to deployment!

🧭 So… Where Do We Go From Here?

These stories aren’t about pointing fingers. They’re about finally paying attention to what we already know: we are in the middle of a second software crisis!

We can’t afford to keep building software systems the same way and expecting different results. As a global software community, we need to:

• Reflect on these patterns

• Share our stories

• Build new ways of working—together

That’s why we created the Global Agile Summit. It’s not just a conference. It’s the start of a long-overdue conversation—and a movement to reshape how software is built across sectors, borders, and generations.

🔗 Get all the Global Agile Summit videos, and learn from the best in the industry!

Let’s stop repeating the past—and start building the future of software. Together!

Reference Material

References in case you are feeling too cheerful today... (read at your own risk)
(1) “Doctors file complaint with health watchdog over Apotti data system” https://yle.fi/a/3-12609291
(2) In Finnish: “The total cost of the problematic system Apotti already reaches an astounding figure.” https://www.kauppalehti.fi/uutiset/kohujarjestelma-apotin-kokonaishinta-yltaa-jo-huimaan-lukemaan/60638ea4-9221-4fd9-9cb3-e661fe23b7ef
(3) “Scrapping the millennium: introduction of a health record in Sweden fails” — https://www.heise.de/en/news/Scrapping-the-millennium-introduction-of-a-health-record-in-Sweden-fails-10323142.html
(4) “KPMG audit: Millennium rollout was unprepared and unsafe” — https://www.heise.de/en/news/Scrapping-the-millennium-introduction-of-a-health-record-in-Sweden-fails-10323142.html#:~:text=An%20external%20audit%20by%20the,of%20500%20million%20euros%2C%20altogether
(5) “Abandoned NHS IT system has cost £10bn so far” — https://www.theguardian.com/society/2013/sep/18/nhs-records-system-10bn
(6) “UK Parliament Public Accounts Committee Report on NPfIT” — https://publications.parliament.uk/pa/cm201314/cmselect/cmpubacc/294/294.pdf
(7) “IT difficulties help take Kmart down” — https://www.computerworld.com/article/1351181/it-difficulties-help-take-kmart-down.html
(8) “Risky IT Projects – NC State ERM Initiative” — https://erm.ncsu.edu/resource-center/risky-it-project/
(9) “Canada’s $1bn Phoenix payroll IT fiasco torched by auditors” — https://www.theregister.com/2018/05/29/canada_phoenix_payroll_system_audit/
(10) “Phoenix pay system - Wikipedia (as of 2024)” — https://en.wikipedia.org/wiki/Phoenix_pay_system