Fixed milestones and experimental development are not good bedfellows.

Milestones want to talk about the relationship, they want to know where we’re going with this, they want communication, they want commitment. Experimental development doesn’t want to be tied down, they want to play the field, try new things, see what works, keep things casual. And this can make your scope super creepy.

Creepy scope is a big problem, if you try to do everything in one go, you risk seriously delaying release. When there’s always a new thing to move on to, bug fixing can easily become future you’s problem.

We found the best thing for us was to group REX in to sub-products and then prioritise the bugs as we went along, agree, fix and release that MVP.

Our thoughts on creepy scope:

• Set clear timed objectives and stick to them as if you're being held externally accountable.
• Don’t get carried away with what that product might be, focus on what it needs to be (I say from the REX Island I have in my head).
• Find a way to organise bugs and feature requests from test users as they come up.
• Find the right tracking software for you and your team, just because it works well for someone else doesn’t necessarily mean it will float your boat.
• As well as functional, performance and usability, find you own bug categories, eg: critical, blocker and acceptable, again categories can be personal to a product.