Day 1 of Inbound - Hubspot Data Cloud+AI

I was at the Inbound event today and they demo’ed a list upload use case. I loved it! It’s something we built a while back. Customers love it, it solves a real problem, it’s tractable and solvable.

What did I see that impressed me? I love how you can invoke a prompt to clean the data in a list (something allGood does). And I appreciated how you can restrict what data is provided to the prompt, something we’ve found to be key to reducing hallucinations.

‍

And of course being able to upload the list of leads seamlessly into Hubspot is such a no-brainer for an admin. It’s one of the first actions allGood took in our list upload use case. There was some real good product thinking that went into this launch.

We’ve spent a lot of time with customers on the list upload use case to move from a demo-ready to production-ready product. The area that’s been the trickiest is supporting all of the corner cases that inevitably emerge. Things work perfectly if the spreadsheet uploaded follows a standard format. But what happens if the header row is not the first row of the spreadsheet but the third row? What if there are multiple tabs in the spreadsheet. Which tab do you use? And worse yet, if you need to use all tabs, but several of them have identical leads but different event statuses. Which record takes priority?

We’ve even seen lists where the rows are color-coded red for users who don’t opt in to marketing emails. There are lots of corner cases. That said, we’ve worked hard on solving these edge cases for our customers. But what I’ve come to appreciate is the edge cases customers call real life.

Congrats to everyone who worked on this launch!

‍