Right now, it's apparently not possible to include a comma inside a product name in SiteCatalyst. I presume this is some hard-coded limitation inside the data processing component. The solution is to strip commas from the source data. This is a bad solution. Munging data is just a bad idea.
I can just see someone down the track, who doesn't know the story, doing some perversion or other with the data from SiteCatalyst and wondering why the numbers don't align. Then wasting a day or more before realising that the SiteCatalyst data, unlike the source data, doesn't contain commas.
In this instance, the data I'm talking about is movie titles. "Crouching Tiger, Hidden Dragon". "Synedoche, New York". Commas are pretty common in movie titles, and are really an integral part of the data.
Sure, I could use a unique identifier and then use SAINT classifications blah blah blah. Yet another automated upload process to monitor and make sure it doesn't break and stay in-sync. Sorry, but I've got enough of that pain already.
Easiest solution? Give us an escape character that's ignored all the way through until the point when the product string is split into an array. "Crouching Tiger, Hidden Dragon" becomes "Crouching Tiger,, Hidden Dragon" or "Crouching Tiger\, Hidden Dragon" or whatever escaping mechanism you choose. Just choose one and tell us!