The second issue, company names in reports not matching what you see on the deal, is usually historical. Reports can surface an older company record name if the company was renamed or merged, or if a different company was primary at some point in the past. The deal view always shows the current resolved association. Reports don’t retroactively normalize names the same way, because that would break historical aggregation consistency.
A few ways teams reduce this in practice. If the business question is “top companies by deal amount,” make Companies the primary data source and aggregate deals secondarily. That path is far more stable. Also, add an explicit filter like “Associated company is known” instead of relying only on “Primary,” and be cautious with renames or merges mid-period if reports are critical. HubSpot also calls out these aggregation limitations in custom reports here: https://knowledge.hubspot.com/reports/create-custom-reports.
So no, this isn’t separate databases drifting apart. It’s the difference between real-time record truth and aggregation logic optimized for reporting scale. Once you see that distinction, the behavior makes a lot more sense.

