Skip to content
Governing agentic GTM
ReadLesson 1 of 7Set the standard

Writing rules agents can actually enforce

Learn what separates a rule that can be evaluated from a policy that can only be quoted, before you author any of your own.

12 minAdvanced
Part 1 of 2

Why rule-writing comes before everything else

Governance in this model is not a review layer bolted on afterwards; it is the rules, the routing and the evidence trail, and all three depend on rules that can actually be evaluated. Almost every organization already has documents describing how things should be done, and almost none of them contain rules in this sense.

The gap is specific: a policy states an intention, and a rule states a condition a record either meets or does not. "Opportunities should be kept up to date" is an intention that has probably been in a handbook for years without ever being enforced, because there is nothing in it to check. Turning that into something checkable is the skill this guide teaches and the next lesson makes you practice.

Part 2 of 2

Read it as preparation for authoring

Read this one differently from a normal guide, because the next lesson has you write five rules of your own and its quality depends entirely on this. Pay particular attention to how a rule expresses its consequence and its exceptions, not just its condition, those are the parts people skip, and skipping them is what produces a rule that fires constantly and gets switched off within a fortnight.

It is also worth separating two things that feel connected and are not. Enforceable does not mean strict. A rule can be generous, with a long grace period and wide exceptions, and still be perfectly evaluable. The property that matters is whether the condition can be tested, not how demanding it is.

  • The condition — what a record must show for the rule to be met
  • The consequence — what happens when it is not, and who hears about it
  • The exceptions — expressed as part of the rule, not as informal tolerance
The RevTech rule library for CRM data standards.
The Sales Process rule library, where the data and process standards agents enforce are defined and maintained. Screenshot of the RevTech application; sample data.
Open exceptions raised against the standards. This is the output your written policy has to produce. Read it as the shape a rule has to reach before anything can evaluate it.

Now read the guide

Writing rules agents can actually enforce

This lesson is the orientation. The material itself lives on its own page, kept in one place so it stays current everywhere it is referenced.

Key questions

You should be able to answer each of these from memory before opening it. Recalling the answer is what makes it stick; recognizing it when you read it does not.

What separates a rule from a policy?

A policy states an intention; a rule states a condition a record either meets or does not. "Opportunities should be kept up to date" has nothing in it to check, which is why it has never been enforced.

Covered in full: Writing rules agents can actually enforce

Does enforceable mean strict?

No. A rule can be generous, with a long grace period and wide exceptions, and still be perfectly evaluable. The property that matters is whether the condition can be tested.

What to take away

  • Read it as preparation for authoring. The next lesson asks you to write five.
  • Have a candidate standard in mind before you start, and test it against the guide's four parts as you read.

Teach this lesson

The argument in 4 slides, for presenting it to your team
Slide 1 of 4

Most process documents are not rules

They are statements of intent. "Opportunities should be kept up to date" cannot be evaluated against a record, which is why it has never been enforced.

Slide 2 of 4

What to look for

  • The test that turns an intention into a condition
  • Why a rule needs a stated consequence, not just a threshold
  • How exceptions get expressed without hollowing the rule out
Slide 3 of 4

You are about to write five of these

Read this as preparation, not as background. The next lesson is authoring, and the quality of your five depends on this one.

Slide 4 of 4

Enforceable is not the same as strict

A rule can be lenient and still be evaluable. The property that matters is whether a machine can tell if it was met.

Demo

Try the demo.

See agents carry the repeatable work of GTM across sales, marketing, customer success, and RevOps. Every action prepared, reviewed, and recorded. Fictional data, real product.

Explore the demo