Overnight migration is a feature now

For most of the last 15 years, the strongest thing a help desk had going for it wasn't the product. It was the 8 years of tickets sitting inside it.

Switching meant a migration project. Meant a services engagement, or a spreadsheet, or a junior person copying macros by hand for 3 weeks. Meant losing your history, or paying someone to not lose it. Every incumbent in every category knows this and prices accordingly, whether or not anyone says it in the room.

We pull a customer's entire Zendesk history in overnight. API in, everything mapped, and in the morning the new system knows more about their customers than the old one did.

That's not a clever thing we invented. It's what happens when ingestion stops being a schema-mapping problem and becomes a reading-comprehension problem.

Why it was hard before

Migrating between two help desks used to be an exercise in making two data models agree. Custom fields that don't exist on the other side. Statuses that mean different things. Macros written against a workflow engine that isn't there. Attachments, threading, merged tickets, the customer who emailed from 4 addresses.

You could write the mapping. Teams did, for years. It just cost more than most companies were willing to spend on a lateral move, and that gap is exactly where incumbent pricing power lived.

The fields didn't have to match to be useful. That's the change. The system reads the ticket, understands what happened, and stores what matters, instead of requiring that column A on the left equals column B on the right.

What it's worth on a sales call

More than anything else we've built, honestly.

The objection that used to kill deals at the end was never price. It was "we've got 6 years of history in there and I'm not doing that project this quarter." That's a real objection and it's the reason a lot of teams stay on tools they complain about weekly.

Now that conversation takes 20 seconds. We ingest it, you keep it, it's searchable, and the AI is better on day one because of it. The buyer's risk goes from a quarter-long project to an evening.

I'd guess it's worth more to us than any 3 product features combined, and it's not a product feature at all in the way anyone means that phrase. It's a consequence of the architecture.

The wider point, for anyone not in support

The moat in a lot of b2b categories is a data-shaped inconvenience.

Your CRM. Your billing system. Your ATS. Your project tracker. In every one of those, some meaningful part of why customers don't leave is that their history is in there in a format that was annoying to get out. That annoyance has been quietly load-bearing for a decade and a half.

If that's the shape of your defensibility, I'd take a hard look at it this year. Not because someone's coming for you specifically, but because the cost of reading and re-understanding a competitor's data has dropped by something like an order of magnitude, and the people building against you know it.

The inverse is also true and it's the more useful read if you're the challenger. Categories that looked closed because of switching costs are open again. Not all of them. But more than were open in 2023.

What it doesn't solve

Migration being easy gets you a fair evaluation. It doesn't get you the customer.

They still have to want to leave. Something has to be wrong enough with the incumbent that a good team spends a week of attention on an alternative, and "wrong enough" is a high bar for software that mostly works. Most of our wins come from teams who are already unhappy about pricing or already burned by an AI layer that lied to a customer. The overnight migration removes the last excuse. It does not create the first reason.

So we didn't build a growth engine here. We deleted a stop sign.

Distribution is still the hard part. Always is.