Service and success are two doors into the same machine

A customer asks how to set something up. Or we notice they haven't set it up and offer to help.

Once they say yes, much of the job is the same. We need the right answer, the account context and permission to do anything we're offering.

I don't see the sense in maintaining two copies of that work because two teams start the conversation differently.

Customer success does plenty beyond this. Renewals, commercial relationships and advice about someone's business don't reduce to a help interaction. I'm talking about the proactive-help part of the job.

For that part, share the content and actions. Keep the teams if they have different jobs. They don't need different answers to the same question.

I'd start by looking at what exists twice. Compare the answers given on a success call with the ones in the help centre. Then look at setup questions reaching support from accounts already receiving outreach.

That last group doesn't prove anyone failed. Support might be the right place for them. It does show where to check whether either team knows what the other has already done.

The same causes of failure are worth checking at both ends. An unavailable action is still unavailable when we started the conversation. Perhaps the feature shouldn't have needed setting up at all.

Then there's the question of whether the outreach works.

If we choose which accounts get attention, comparing them with everyone else won't settle what the attention caused. We selected them for a reason. That reason can also affect the result.

My preference is a holdout. Take an eligible group and randomly leave some out of the optional outreach. Keep normal support available and compare a result agreed beforehand.

If it works, the team has a better case for doing more. If it doesn't, we can stop spending time on it. Either answer beats defending the programme because we've always done it.

Keep the reports separate, though. Useful outreach can create conversations. Useful service can prevent repeat ones.

Put both under a target to reduce contacts and someone will sensibly stop the outreach. We'd have given them a reason to undo the other team's work.

Outbound also interrupts people who didn't ask. They need an easy way to decline — and we need to remember they declined.

The limit on my argument is the size of this shared work. If most of the value is in relationships and commercial advice, sharing answers saves some maintenance. It doesn't justify reorganising the business.

A holdout won't suit every account either. Promised help has to be delivered, and a test on smaller accounts may tell us little about larger ones. Don't sell the finding beyond what was tested.

I'm still backing shared machinery for shared work. Ask the two teams what they maintain twice, then ask the outreach owner what evidence would make them stop.