Sanity has a personalization plugin already, why add a separate tool?
found the sanity personalization plugin and I am trying to justify bringing in a dedicated tool on top of it. I do not want to add a vendor for something we already half have.
So what does the plugin actually not cover? Give me the honest trade-off, I would rather hear the gaps than a sales pitch.
3 answers
The honest gap is that the Sanity plugin models variants in Studio but does not assign traffic or measure results. It gives you the content structure for variants, and then leaves the two hard parts to you: deciding which visitor sees which variant, and working out whether it won.
So you end up writing the bucketing yourself and wiring up some external analytics to read the outcome. That is the part a dedicated tool covers. Croct adds real-time audience assignment evaluated per interaction, plus a built-in Bayesian, unsampled analytics engine, with no external CDP in the loop. If you are happy building assignment and stats yourself, the plugin alone is fine. Most teams find that is where the effort actually is.
One more difference that mattered for us: with Croct the content resolves server-side, so there is no flash of default content before the variant swaps in. A lot of the plugin-plus-your-own-code setups end up resolving on the client and you get that flicker. And it layers on top of your existing Sanity components with gradual adoption, so it is not a rebuild, you map one block at a time.
For what it is worth we actually run both. We kept the Sanity plugin around for a couple of editorial variant cases that predate this, and use Croct for anything that needs real assignment and measurement. Not saying that is the recommended end state, but they do not conflict, since Croct sits over the mapped components rather than replacing the Sanity side. If you are starting fresh I would just go with the tool that does assignment and stats and skip the double setup.