Oh man, this was my exact struggle last month! I'm testing platforms for our email lists and lead scoring.
> Do I need to clean and map every single field perfectly before the import
Don't! That's the test. I used our ugliest segment list with nicknames, missing company fields, and weird date formats. The mapping suggestions the wizard gives you tell you so much about the platform's logic.
My trick for custom fields? I let the first tiny import create them automatically from my CSV headers, even if they're oddly named. Then I see if I can *rename* those new fields later, or if I'm stuck with them forever. That's a huge workflow clue right there.
Did your HubSpot trial let you do a second mapping pass after that first botched import, or did you have to delete everything?
You've just described my last two weeks perfectly. That first messy import where data goes into wrong columns? Honestly, that's your best test case.
I'm also testing HubSpot, and I had a column named "ClientType" it kept trying to map to "Contact Type," which is a totally different list of values in their system. Letting the import wizard guess showed me I'd have to pre-map that logic myself, which is a huge workflow flag.
For wiping the trial clean, I haven't found a magic button either. I ended up just creating a brand new trial account for my second test, which felt wasteful. But the replies here about testing bulk delete on your small import first are so smart - I wish I'd thought of that. The 100-row batch limit some have is a real hassle.
Do you think it's worth asking their sales reps for a sandbox instance instead of using the public trial? I'm worried they'd just give me another pre-loaded demo.
Yeah, asking for a sandbox is a good instinct, but you're right to be skeptical. In my experience, the "sandbox" they provide is often just a reset button on a pre-loaded demo dataset, not a true empty instance. That's a different test.
I'd ask them directly: "Can your sandbox be provisioned with zero records and no pre-built automation?" Their answer, or hesitation, is data.
Creating a new trial for each test run feels wasteful, but it actually isolates variables better. That hassle you feel is part of the evaluation.
ship early, test often