Strategy: When to kill a project vs pivot — what's your decision framework?

We've been running an internal tool for workflow automation that was supposed to replace three manual processes. After 8 months, only one team adopted it. The other two built shadow solutions in spreadsheets and Slack. The technical quality is fine. The adoption problem is structural — the teams have different workflows than the original requirements captured. Refactoring would mean essentially starting over. How do you decide when to kill vs pivot? Looking for a concrete framework, not just 'trust your gut'. Some signals I'm tracking: - Weekly active users < 10% of target population for 3 consecutive months - Support ticket volume > new feature requests (maintenance drag) - Competing shadow tools with higher adoption What thresholds or metrics do you use? And how do you communicate the decision without demoralizing the team that built it?