AI columns

AI columns turn your own criteria into structure. Each column asks the AI one question about every prospect in the list, and files each one under one of the values you defined. An “Industry” column can sort a whole list into Communication / Non-profit / Other; a “Relevance” column into Hot / Later / No.

Creating a column

A column is three things:

• A name — what the column measures.
• Values — the closed set of possible answers (up to 24), including a default for the cases the AI cannot decide.
• Rules in plain language — how to decide, written as you would brief a colleague: “companies that sell services to other businesses count as B2B”.

No formulas, no formal syntax — the rules are natural language.

Running the analysis

You launch the analysis on the list; progress shows in place, and already-analyzed prospects are never re-billed. Columns run on AI credits, with a choice between an economical model and a higher-quality one depending on the column.

What columns are for

• Slicing a list — filter by column value to focus a campaign on one segment.
• Cleaning — combined with the assistant, columns make removals verifiable: you see exactly who matches before confirming.
• Conditional messaging — campaign messages can vary by column value, so one sequence speaks differently to each segment without splitting the list.

For the people who must never be contacted at all, use Exclusions instead — columns organize a list, exclusions protect the whole workspace.