You're asking the wrong question. You're not choosing between a fast car and a comfy couch. You're choosing which kind of headache you want.
Ease of use is a trap sold by marketing teams. It's a shiny dashboard that falls over the moment you try to connect your actual data or have more than five people use it. Performance isn't a luxury; it's the foundation. A slow BI tool will die in adoption because no one will wait 45 seconds for a chart to load.
Start with your data volume and your team's SQL comfort. If you're pulling from a massive data warehouse, prioritize the engine (like BigQuery, Redshift, or Snowflake-native tools). If your "data team" is one sales ops person who knows Excel, then maybe you suffer with something like Zoho Analytics or a simpler connector. But understand: picking ease today usually means a costly, painful migration later when you hit the performance wall.
CRM is a necessary evil
> Ease of use is a trap sold by marketing teams.
This is a great point, but I think the trap is slightly different. It's not that ease of use is fake, it's that it's non-transferable. The "easy" tool's magic only works inside its walled garden. The moment you need a custom calculation or to join across weird sources, you're coding anyway, but now in their proprietary language that's slower and has no ecosystem.
Performance *is* ease of use, just deferred. A fast, ugly SQL editor that gives you a chart in 2 seconds is easier to live with daily than a drag-and-drop UI that spins for a minute before timing out. People confuse "easy to learn" with "easy to use at scale." They are opposites most of the time.
Seen it a dozen times with BI tools. The "easy" one gets demoed and bought, then 18 months later I'm getting paid to migrate them to something that can actually query their data lake without melting down 😅