Where Operations Meets Go to Market

The easiest way to waste time on an AI project is to choose work that sounds exciting in a meeting but breaks down in day-to-day operations.

From Matthew Rubin’s operational lens and Brendan Ahern’s strategy lens, the strongest early opportunities sit in the overlap between two questions:

1. Is the workflow important enough that a business will prioritize it?
2. Is the workflow structured enough that a team can actually run and review it?

That overlap matters. A painful problem without a repeatable process usually turns into a custom rescue mission. A tidy process without real business value becomes a nice demo that nobody funds. The better starting point is a workflow with visible friction, a clear owner, and a straightforward way to tell whether the work improved.

A practical first candidate usually has five traits:

• one team clearly owns the workflow;
• the work happens often enough to justify standardizing it;
• delays, missed follow-up, or inconsistent handling create real business drag;
• a human can review exceptions before they become customer problems; and
• the first measurement is concrete, such as response time, completion rate, or handoff quality.

That is why operational readiness and commercial value need to be scoped together. Matthew’s side of the question is whether the work can survive contact with inboxes, handoffs, permissions, and real-world exceptions. Brendan’s side is whether solving it creates enough urgency and measurable value to deserve attention now instead of later.

For an early AI service, that shared standard is a useful filter: choose the problem that a team can both explain in a sales conversation and sustain in the workday. Everything else stays in discovery until the process and the business case are both real.