Update Cadence is something that has gone through a number of revisions since the game's initial launch.
In the first month, we put out updates each week at the end of the week. Actually, in the first week we put out a few updates - hotfixes for specific bugs found at launch. We continued to put out updates each week but it was exhausting & those updates really didn't have much in them other then bug fixes & minor polish.
For a few months after that, we switched to bi-weekly updates and also moved to targeting the update for Wednesday. This allowed us to get a little bit more into each update and prevented slipping the update to the following week as it turned out whatever we were trying to accomplish always came in hot.
Then we went on a short vacation and had to skip an update - landing the next update at about a month since the previous. At that point, we switched to a Monthly update cadence with the update scheduled on the last Wednesday of the month.
The monthly update cadence has worked very well, allowing us to plan bigger changes and get a lot into each update. With enough room in a monthly iteration, we are able to burn down polish and deferred maintenance tasks as well as investing in less flashy things like telemetry.
Over time, Ship Day has become Ship Week. Major changes stop at the end of the previous week ideally, allowing us to spin off a build and test it thoroughly as well as sharing that build with beta testers and localizers. The first few days of Ship Week are dedicated to testing and working on messaging for the update - something that we previously tried to shove into the schedule without budgeting much of the time it took, which was one of the reasons update slipped. What time remains in that week gets allocated to project planning and web presence updates.
You can see game update posts here:
Along the way, we've paid attention to how the rest of the industry deals with updates. Of course it depends quite a bit on the scope of the project, the size of the team and the state of the game once it's live.
Common consensus seems to be that an update cadence much more frequent than a month doesn't give the players enough time to anticipate or digest an update. Small, frequent changes tend to blend together or get ignored and are better suited to tuning and polish than introducing new features or drawing old players back to a game.
Introduction of bigger features works better on less-frequent updates where there's enough in the update to warrant the attention and enough time has passed for the players to have digested the previous changes. Big changes can also be an opportunity to do timed-events, combined with a marketing push to bring in new players and renew interest in players who have put the game on the shelf.
It's probably no coincidence that Steam won't let you schedule a sale any sooner than 6 weeks after a previous sale - one of the things they recommend is you pair sales with major updates. Steam is a company that has a ton of data about what works and what doesn't work in such cases, their recommendations usually correlate strongly to the more successful patterns.
Honestly, something like quarterly updates may work better for those big feature updates. For now, we're sticking to a (mostly) monthly cadence for updates of Prismatic Maze, but we'll continue to tweak this going forward.