About 10 days ago seemed to be a low-point in the development and launch of Notify My Team.
I was stuck in subscription testing hell, essentially working through the various lifecycles of a subscription using a bunch of fake companies and our test Stripe credentials. If you want to get a feel for what I’m talking about, check out Stripe’s subscription lifecycle page. Subscriptions are endlessly complex compared to collecting a single payment for an e-commerce store.
I had 20 test companies at various stages in the process on accelerated subscriptions (the monthly plan billed out daily, the yearly plan every 3 days). An example would be: what happens if someone subscribes, upgrades to a higher tier during their trial period, makes a payment, switches credit cards, downgrades, cancels, and then decided to un-cancel prior to their period ending. Does everything work? Is the user experience good?
This is really important. Nothing can turn off a customer more than a bad subscription experience. If someone wants to upgrade/downgrade, cancel, change their payment method, etc, it’s critical that it works well and that it works with as little friction as possible. Especially for a B2B tool. And if scenarios like that seem far fetched, they’re not. I recall seeing everything I tested and more when the SportsLizard Price Guide was at its peak. I’m hoping to head off a lot of customer service and angry customers by building a much better experience this time around.
It’s also some of the more stressful testing I’ve done. Inevitably, each day for about a week I’d find a bug that I’d chase down for a few hours and couldn’t completely be sure I’d fixed until I set up the same scenario again and waited until the next day for it to bill. I’d be behind for the rest of the day. The difference between now and a decade ago when we launched LockerPulse (and more new things in general) is that I could make up the time at night or on the weekends. Now, with a family and the rest of the business requiring my attention, that’s just not possible.
The good news is that by early this week I had seemingly squashed most of the serious bugs. And yesterday we rolled out the service internally to our team, which felt like “launch day 1.0” because this was a product that we built first and foremost for ourselves. When our first snow storm hits, we’ll be ready!
There’s still a lot of work left to launch to the public, but this week felt like the project turned a corner. The ups and downs reminded me of a drawing I created a decade ago leading up to the launch of LockerPulse called The Web App Launch Emotional Rollercoaster. As true today as it was back then!