Executive Summary
Every vendor in the category is now an AI company, which makes it unusually hard to work out what has actually changed. Some of it is real: summarising a long account history, drafting a follow-up, spotting a deal that has gone quiet. Some of it is a feature that existed in 2019 with a new label on the pricing page. This hub is about telling those apart, and about the second-order question that gets less attention: what an AI layer does to a business whose underlying data was never modelled coherently in the first place.
The uncomfortable dependency
AI in a sales organisation runs on the same records everything else runs on. If a group of trading companies cannot agree what an account is, or which entity owns a relationship, then a model summarising that data inherits the confusion and states it fluently. The question of whether a group needs one CRM or one per company sits upstream of any AI decision, because a model trained on badly separated records will confidently blend accounts that should never have been merged. Fluency is the problem. A wrong number in a spreadsheet looks like a wrong number; a wrong number in a generated summary looks like an answer.
That is not an argument against the technology. It is an argument about sequence, and it is the reason the least glamorous work, deciding what your entities are and who is allowed to see what, has become more valuable rather than less. Getting that structure right, particularly the question of who should see what across group companies, has to happen before an AI layer is worth trusting. The same discipline applies to sharing leads between companies in the same group: an AI tool that surfaces a lead to the wrong brand is not a technology failure, it is a permissions failure wearing a technology costume.
What this hub covers
How AI changes forecasting and pipeline review, where automation genuinely removes work and where it moves it, what to ask a vendor whose roadmap is suddenly all AI, and how any of it applies to a business running several companies at once rather than one.