Well the end finally came for the old guard of engineers at Appcues, the ones who built the place. That isn’t quite right or fair, many others have come and gone over the years either voluntarily to an exciting new opportunity or involuntarily due to a sequence of layoffs over the last few years. For a long time, though, a core team endured and helped shape the engineering team into what it was for a while and still is today.
My former colleague Brett Crowell quite neatly summed up many of our relationships to the job in his own writeup where the primary characteristic of our time was a love of the people: the team we built and got to work with every day for the better part of a decade (and in some cases, longer). In attempting to pen my own words I struggled to find another conclusion, it really was a team that made it easy to get up in the morning and solve problems with.
If I could add something, though, it would be not just the team but that Appcues true enduring legacy will be the lessons we learned in our time there that we will carry with us for the rest of our careers. We made a lot of mistakes but such a massive part of what we did was learn and grow and challenge each other to improve upon those mistakes. With the pressures of an early startup comes the desire to deliver, and deliver quickly, and in service of product market fit (apologies for using business jargon unironically) we made compromises that we would later need to rectify. A slipshod NoSQL as a service data store backing all of our customer data eventually became a SQL store with an API in front of it. Janky auth, account, and user management became a system we could be proud of. SDK delivery and end user reactivity went from costly, error prone, and slow to cheap(er), resilient, and scalable. We learned how to save tens of thousands of dollars a month on user profiles. Our event pipeline went from lambdas, unstable services, and a lot of shuffling objects around in S3 to a kafka pipeline and sturdy fleet of different consumers. The analytics offering went from unstructured data in a service we were costing significantly more money than we were paying for it to a columnar database that could get customers answers in milliseconds instead of minutes (or an unfortunate timeout). Every single lesson was hard earned, and made it so the team didn’t just grow in size but in ability: the school of hard knocks is the best teacher after all. Personally, that is what makes me grateful for my nine years before the mast at Appcues.
Appcues will continue, there’s still quite a few talented people working there, but for the rest of us, the war is over.