AI accelerates workflow rebundling
Bob Moore, CEO and co-founder of Crossbeam, on ecosystem-led growth
AI favors companies that own more of the workflow, because a single assistant works better when the data, rules, and actions sit inside one system. That is the logic behind rebundling. The modern data stack split one analytics product into Fivetran, Snowflake, dbt, and BI, but newer winners are trying to pull adjacent layers back together so users can ask one interface to do the work instead of stitching tools together by hand.
-
Rippling is a clean example of rebundling in practice. It started with HR and payroll, then added IT, security, and contractor management on the same employee record. That lets one admin system trigger payroll, device setup, app access, and policy enforcement together, instead of buying separate tools for each job.
-
The same pattern is showing up in data. dbt grew as the neutral transformation layer inside an unbundled stack, but Snowflake and Databricks have moved into native transformations, while dbt has expanded into orchestration, catalog, and observability. Everyone is trying to own the main screen where teams define and act on business data.
-
Rebundling changes partnerships. When adjacent tools overlap more, they become frenemies. Crossbeam's own product exists for that world, letting two companies share just enough customer and account data by segment, geography, or use case to sell together in one area while competing in another.
The next phase is likely software that looks less like a stack of tabs and more like one operating surface per job. AI will accelerate that shift, because the best copilots need deeper control over the systems beneath them. The companies that win will be the ones that turn fragmented point solutions into one coherent workflow with one place to ask, decide, and act.