One product can serve people with very different reasons for using it. That doesn't automatically mean they belong in the same ideal customer profile, or ICP. A broad profile can become difficult to interpret when every result needs a different explanation.

I would consider splitting it when the audiences need different discovery decisions. Different language, requirements, and replies are more useful signals than simply having two industry labels.

Compare what each audience is trying to do

Imagine a fictional scheduling product used by solo consultants and larger service teams. A consultant might care about avoiding back-and-forth booking emails. A team might care about allocating work across several people. The product category overlaps, but the situations can differ.

Write the problem, trigger, constraints, and useful first response for each audience. If those descriptions are almost identical, one profile may be enough. If they keep diverging, separate profiles may make the work easier to judge.

Find the rules that keep colliding

A useful sign is a condition that helps one group while excluding the other. Another is a feed you can only evaluate after sorting every post into a different mental category. I'd inspect whether the combined profile is hiding two distinct definitions of fit.

You don't need a separate profile for every minor difference. Each profile creates review work. Give a split a clear purpose before adding it.

  • Do the audiences describe different problems?
  • Do they need different capabilities or service scopes?
  • Would the same exclusion remove good matches from one group?
  • Would you write a substantially different first reply?
  • Can somebody review the results from each profile consistently?

Try one clear distinction

Within your ICPSignal plan's profile limits, describe each audience independently. Use concrete examples and exclusions. Review the profiles before activating them and inspect the original posts behind their matches.

Avoid changing the offer, source scope, and qualification standard at the same time if you want to understand the split. Record the conditions of the test. The aim is to see whether separate descriptions support clearer decisions.

Keep the split only if it helps

Compare relevance, review effort, and the kinds of conversations each profile produces. Keep track of repeated posts so you don't mistake overlap for additional opportunity. Judge the useful work rather than the number of profiles you've created.

If the audiences converge in practice, simplify. If they require distinct handling, keep the distinction and make the responsibility for each clear. I want the structure to reflect how customers buy and how your team works.