You know the feeling. The dashboard was green for six months. The steering committee meetings were polite updates about "foundational architecture" and "vendor alignment." Then, in a single week, the status flipped to yellow, then deep crimson red. Now you are six months past the original go-live date, the budget variance is negative $2.4M, and the board wants to know why the "game-changing" ERP migration hasn't processed a single invoice.
You are not alone in this trench. According to McKinsey, large IT projects run an average of 45% over budget and deliver 56% less value than predicted. But averages hide the real danger. The true killer is the "fat tail" risk: 17% of IT projects go so badly they threaten the very existence of the company. We call these "Black Swans," but in the current enterprise landscape, they are becoming common pigeons.
For the Enterprise CIO or VP of Engineering, the problem isn't usually the code. It is rarely that Python scripts failed or the API didn't connect. The problem is a governance vacuum filled by scope creep, sunk cost fallacy, and vendor mismanagement. You are trapped in a cycle of "one more sprint" to fix a structural failure. To break out, you don't need more developers; you need a triage surgeon.
When a project bleeds, the instinct is to apply a tourniquet of cash—hiring more contractors or buying bolt-on tools. This is fatal. You must pause execution for a 72-hour "Audit & Triage" cycle. Your goal is to categorize the initiative into one of three buckets.
These are projects where the technical thesis holds, but the timeline was delusional. McKinsey data shows that every additional year a project runs increases cost overruns by 15%. If the ROI is still valid despite a 20% budget hike, you re-baseline. You strip the scope to the absolute Minimum Viable Promotion (MVP) and ship.
Often, 80% of the complexity comes from 20% of the requirements—usually custom edge cases demanded by a single stakeholder. I recently advised a Fortune 500 logistics firm where a $12M migration was stalled by a bespoke billing module used by three customers. We cut the module, offered those customers a manual concierge service, and launched the main platform in 60 days. Enterprise software gets stuck in committee because we try to automate exceptions rather than managing them manually.
The hardest category. If the market has moved (e.g., you are building a proprietary LLM when GPT-4 just commoditized it), or if the technical debt is structural, you must kill it. A Gartner survey reveals that only 48% of digital initiatives meet their business outcome targets. Better to write off $3M now than $10M next year. As an operator, your credibility soars when you shoot a zombie project in the head rather than letting it eat the P&L.
Once you have identified the "Cure" or "Cut" candidates, you need a recovery protocol. Standard agile sprints are too slow for crisis management. You need a War Room.
Recovery is not about working harder; it is about reducing the surface area of the problem. Your job isn't to be the hero who codes all night. It is to be the leader who makes the hard decisions to cut scope, fire underperforming vendors, and focus the team on the single metric that matters: shipping.
