Why do IT projects (almost) always fail?

Back in 2008-10, during my Computer Science studies at university, one of our professors shared startling statistics: more than 50% of IT projects historically failed, either partially or entirely. What was even more surprising was that, thirty years on, this ratio remained almost unchanged. Over half of the IT projects were still failing – failing to deliver, failing to work as expected, or dramatically exceeding the budget.

Fast forward to the present, did the situation improve? The answer is disheartening.

• The Standish Group’s CHAOS 2020 report reveals that 66% of global tech projects (based on 50,000 projects) result in partial or complete failure. Large projects succeed under 10% of the time.
• Additionally, they found 31% of US IT projects were outright canceled, and 53% showed worrying performance.
• 2020 McKinsey research shows that 17% of large IT projects pose a threat to the company’s survival.
• BCG in 2020 found 70% of digital transformation efforts missed their targets. A 2020 CISQ report showed failed US development projects costing $260B, and operational failures from poor software quality costing $1.56 trillion.

70y after the invention of FORTAN – the first widely used high-level general-purpose language by IBM – it seems we still haven’t learned to properly analyze, plan, create, and deploy software.

During my software building journey, I’ve identified several sources of failure:

Lack of Right Question Asked
Suppliers often don’t challenge customers, even when the proposed build doesn’t make any sense from the outset. It’s akin to allowing a customer to buy a sports car to run a moving company.

I don’t believe customers pay us to blindly build software for them. In this day and age, AI like ChatGPT can write code more efficiently than many programmers. I firmly believe that customers hire us to become their technology partners upon whom they can rely.

Wrong Technology/Tools
Tools are just that – tools. There are more than 700 different programming languages (some sources even list up to 9,000), but if we narrow it down to the “meaningful” ones listed by the TIOBE index, we end up with 274 official languages.

Technology changes quickly, so the temptation to go with the newest “sexy” technology is high, but it doesn’t necessarily mean it’s always the best choice.

Wrong Project Management
There’s not much to add to this. There’s no budget that can’t be wasted without proper management in place.

Prioritizing Price
Interestingly, about 50% of our current projects are considered “emergency” projects. These are situations when a customer didn’t initially choose us due to a more attractive price offered by someone else. However, a few months later, they return to us because their project ended up in… surprise, surprise – failure.