Working and managing a contractor

Positives

• I personally really enjoyed getting to work with a contractor
  • Phil was great
  • Was nice to see my own design choices and convictions actually be implemented (on Manifold itself I feel like I sometimes have really good ideas which never get implemented, even if others agree they are good ideas).
• Things actually got done and properly optimised.
  • I’m pretty sure the optimisation of the bot is significantly better than a lot of new features we release on Manifold simply because Phil is committed to only working on the bot.

Could have been better

• We were very bad (both Phil and Manifold) at predicting how much work would be needed to create a Twitch bot.
  • To further exacerbate this we didn’t do a good job communicating to Phil that we initially just wanted a more proof-of-concept type of bot. It ended up just becoming a fully fledged high investment thing.
• In the future when contracting someone on a task-based contract the timeframe needs to be agreed upon in advance
• I should have implemented more regular “standup” like meetings/notes from the start. Luckily it didn’t affect this project too negatively but it definitely could depending on the contractor.
• Time zone differences caused some complications. In the future should ask the contractor what their work hours are and ask in advance if they can occasionally make certain times work. Phil was gracious to work in evenings so it luckily wasn’t a huge issue.

Timelines/deadlines

• I mentioned it above, but the bot did end up dragging on for ages and therefore becoming a much bigger investment than we were initially anticipating.
• Still unsure about the best process for trying to predict in advance how long different things will take
• Lots of random unforeseen things/features that required time
• Ultimately though very important to have an official launch date, even if tentative. Having the tournament happen was really good to make sure we finally got the bot into a working state.

Other