Different people tend to have different processes for planning an adventure game. Even professional companies do it differently.
But to generalize it though:
- First you need to explore ideas,
- then you need to choose an idea and expand on it,
- then you need to work out a story,
- then divide that story into a series of events,
- then divide those events into goals,
- then figure out what needs to be achieved in order to fulfill those goals,
- then continue to subdivide these goals until you have an appropriate level of complexity,
- then finally start making the artwork and implementing it all, making adjustments to the plan as you go along.
As an example, let's say I want to make an adventure game.
I might have a few ideas, a sci-fi story, a high fantasy story, and a horror movie story.
I choose the sci-fi story, and decide that it's going to be about a guy trying to get drunk.
I then come up with a premise, it's about a guy named Bobby Bloke, who visits a space bar in order to get a beer, but he can't because he's underage in that part of the galaxy, despite being 46 in human years.
The events are: visit space bar and try to order beer, and go on adventure which results in getting beer.
The goals are therefore to: find a bar, prove you're not underage, and finally get a beer.
To find a bar, you would need a map. To prove you're not underage, you'd need a fake id. To get the beer, you'd need money.
To get the map, you need to steal one. To get a fake id, you need to apply for one. To get the money, you need to get a job.
This continues until you have a full puzzle dependency chart, and then you just need to write down a layout of the rooms, how they interconnect, what needs to be in them, and of course start drawing them.
Lastly, you don't want to stick too rigidly to the plan. It should always be evolving. Also, you don't want to spend too much time planning. A bit of planning can go a long way, but there is a point where you start to hit diminishing returns.