It's a classic teacher move: you ask a question you already know the answer to. The whole thing is rehearsed. EdTech companies do this constantly with districts.

Dylan Hoffman, our VP of Partner Success and Operations at Subject, noticed the pattern early and decided to stop performing. When he built Subject's customer success function from scratch in fall 2023—no team, no process, 387 accounts staring back at him—he made one foundational decision that changed everything: treat every district like the expert in the room, because they are.

Net retention went from below 60% to above 140% in under two years. How did we get there? Well, none of it is complicated. It's just that Most EdTech companies have decided the juice isn't worth the squeeze.

EdTech's favorite bad habit

There is a standard script for EdTech-district relationships, and it goes like this: vendor trains district on the platform, district uses the platform, vendor checks in at year-end to see if the contract is still breathing. Everybody plays their assigned role. The vendor is the teacher. The district is the student who already Googled the answer but sits politely through the presentation anyway.

Nobody questions the script. Questioning it would require admitting it isn't working... and who has time for that in Q4?!

Dylan questioned it. He came in with a VC background wired for portfolio thinking, not account management. Plus, what he found looked less like a partnership and more like a one-way street where only one party owned a car.

The data that should have been shaping Subject's product was sitting unused in support tickets. Rich, specific signals from teachers and administrators—people who had survived no fewer than eleven "next-generation" EdTech platforms since 2015—were never reaching the people making product decisions.

"There was a lot of unmined, really rich data in our customer support channels," Dylan says, "but it wasn't even making its way to the product at all."

Most EdTech companies struggle with retention rates below 30% after the first year. Our Subject team was already outpacing a very low bar but still trending in the wrong direction. Dylan's diagnosis was that the product was fine, but the relationship model around it was broken.