LOG-006 | Stardate 2026.05.18 | 5 min read | Status: Transmitted
Buyer literacy

Should you automate this? Be honest.

Most 'should we automate?' questions get a 'yes' before the question is fully formed. We are not against automation. We are very against the wrong automation. Five questions, one verdict.

The default-yes problem

When the answer is always 'yes, automate it,' you stop noticing the cases where the right answer is 'no, instrument it' or 'just write a Python script.' Most failed AI projects we audit started with a workflow that should have been left alone or replaced with rules, not with an LLM.

The cost of a wrong yes is not zero. It is six months of build, an ongoing API bill, a debug surface no one wanted, and a precedent that says 'this team ships things that do not need to ship.' The cost of a thoughtful no is one meeting.

Run the tree

Five questions. Pick the closest answer; you can always Back. The verdict at the end is one of seven, with the rationale and next steps built in.

Bias check

How you arrived at the question matters more than the answer. Two people can ask 'should we automate this?' and mean opposite things. Quick self-check:

Walking into this question, what's your default lean?

Neither end of this is right or wrong. But knowing your bias before you run the tree saves you from rationalizing toward a verdict you already had.

Three signals you are wrong about this

Even after the tree gives you a verdict, here are the three signals we see most often that the verdict is wrong, regardless of which way it leaned:

  • You cannot describe the workflow in five sentences. If the description sprawls, you have multiple workflows tangled together. Split, then re-run the tree on each.
  • The team running the workflow today is excited. Not skeptical, not nervous, excited. That usually means they want a tool, not an automation. Listen to that. Sometimes the answer is 'buy them better software, not build them an agent.'
  • You are pricing the build but not the operations. The build is one cost. The ongoing tokens, evals, monitoring, on-call, and prompt drift handling is another. If only the build is in your budget, your verdict was based on half the math.

The tree is a starting point, not a verdict generator. The real value is forcing the conversation to happen out loud, in front of a single concrete prompt, instead of in the swirl of a roadmap meeting.

O
The Orbit team
LOG-006 · Transmitted 2026.05.18 · 5 min read
Share

Get the next transmission

A new dispatch every couple of weeks. Short, interactive, no fluff.