Every CEO eventually sits in the sunk-cost meeting
Every CEO eventually sits in the sunk-cost meeting. It opens with a sentence that sounds like an argument and is not one: "We've already spent $300k."
That sentence is information about the past. The decision in the room is about the next $300k, and the first one has no vote.
The only questions that matter are forward-looking. What does it cost to finish from here, honestly, with the risk included? What does it earn once finished? What does the best alternative cost, starting today, and what does it earn? Compare those. The $300k already spent appears in none of them.
This is easy to say and brutal to do, because the money spent has names attached. The manager who championed the project. The vendor relationship. The board update from last year. Killing the project feels like admitting the $300k was a mistake, so the company spends another $300k to avoid the feeling.
The cheapest way out is to separate the two questions in the meeting itself. First: was the original decision reasonable with what we knew? Often yes. Say so. Then: knowing what we know now, would we start this project today at the remaining cost? If the answer is no, you have your decision, and nobody had to be wrong to get there.
The past is paid for. The future is the only thing still for sale.