“Initially we were excited by the prospect of a single data model for all our integrations” Jake McCloskey, co-founder and CTO of Unthread, says. “We were using the CRM integrations of Merge.dev and could ship a first version pretty fast. But this was when we started to see that a rigid, unified data model might not be the right choice for us.” Unthread helps companies run customer support on Slack connect channels.
They integrate with CRMs to link conversations in these Slack channels with CRM data, which enables all sorts of helpful automations and enhancements for their customers.
“The unified data model Merge provided us was missing key information our customers needed in Unthread”, Jake elaborates. They managed to get the data with passthrough requests, but doing so meant re-building a lot of the integrations infrastructure in Unthread. It also lead to a lot of API-specific code in Unthread’s application.
“In addition, we discovered that even with a unified data model, there were still differences in behavior between the CRMs our customers were using. We may get the data in the same model, but our code was still sprinkled with Ifs for different CRMs and handling different cases for each. It got to the point where we working as much around the platform as we were leveraging it, which felt wrong to us.”
“On top of the technical challenges, we were facing a dilemma with Merge’s pricing”, Tom Bachant, co-founder and CEO of Unthread explains. “Our customers were asking for more, and deeper, integrations with more systems. Unlocking these was key for our success, but the high per-connection price we were quoted by Merge just made this unfeasible with our business model.” Armed with these insights the team started looking for solutions that fit better with their needs.