The first constructs that caught our attention were Layer 2 State Channels.
Even though Sidechains were equally popular back then, Sidechains were more complex and often needed custom protocol extensions on the sidechain to work their magic which was a strict no-go for us.
Also, sidechains opened a whole new can of worms with propagation and consensus related issues and this level of complexity is best avoided unless there is no other option.
We were slowly gravitating towards developing a Cold Wallet & Hot Wallet solution for Bitcoin and Layer 2 State Channels seemed like a good fit for this as they were private constructs between two or more cooperating entities and were also simple to approach as a technology.
Every week, we would come up with a new State Channel idea and debate its pros and cons. And we rejected any idea that was too complex operationally or internally. Nobody would trust a solution that they cannot understand easily or so was our rationale.