Upgrade Q - Number of predicted changes in Upgrade preview vs Actual number of changed records?

[Suggy]

Has anyone compared the "Number of predicted changes in Upgrade preview" vs "Actual number of changed records" post upgrade?

How close will the numbers be?

[Ravi Gaurav]

Hi @Suggy 

Good question This comes up often during upgrade planning.

Below details are my Personal Writeups for other clients:

Upgrade Preview – “Predicted changes”

• The preview scan looks at:

  • Which out-of-box (OOB) files/records will be replaced by the upgrade.

  • Which customizations will be skipped or marked as “potential conflicts.”

• The “number of predicted changes” is a static estimate based on the delta between your current build and the target build.

Post-upgrade – “Actual changed records”

• After upgrade, the actual count of changed records is influenced by:

  • Which update sets/patches your instance already had.

  • Which store plugins/features you have installed.

  • Whether some changes were skipped (manual compare, retained customization).

  • Any post-upgrade tasks or patch fixes that overwrite additional records.

How close are they?

• Generally close, but not identical.

  • Preview gives you a good directional estimate (within ~5–15%).

  • Actual numbers may differ because:

    • Some changes are skipped due to conflicts.

    • Some records may have already been updated in hotfixes.

    • Customer customizations may suppress certain updates.

    • Patches delivered after preview scan can slightly shift numbers.

Why use preview?

• The main value of preview is identifying conflicts, not guaranteeing an exact record-change count.

• Think of it as “here’s the scale of what will be touched” so you can assess risk and testing scope.

 The numbers won’t match exactly, but they’ll usually be in the same ballpark. Use the preview count as an estimate, and rely on the post-upgrade skipped record report for the actual truth.