My name is randomperson and I am co leader of the mod team Generals British forces (GBF). This is a mod that is adding Briton as a playable side to Command and Conquer Generals. We recently went through a redesign and I thought I should share the knowledge that I have gained from that experience as I have learned so much and I think that people should understand how important design is to a successful mod. The pictures that I have included are of a cut down version of our real documents to show you what I mean without revealing too much.
You need to plan in the first place. It can not be emphasized enough, you need a design document. While it is so very tempting to rush off and start development it will slow you down later on in the project as you have to stop and redesign it parts of your mod. It doesn't need to be a 40 page long written document, we at GBF have created a mind map so we know where each unit fits in with what we want to do. What is also important is that it is kept up to date as if it is not up to date then there is no point looking at it. Lots of thought must also go into it as each idea that goes into the design document will hopefully go into the final product, so try and think of balance and gameplay when thinking of content. At GBF we did have a design document but (no offence to sparten07) it was unrealistic, incomplete and badly thought through as there were gaps in it, and uncodable and uncharacteristic units and structures. We then went off and started modelling all these new units on this incomplete design document, and now we are going to chuck away at least 3 fully modelled units as they don't fit into the redesigned mod. This doesn't do any good for the team as no one likes their work that they spend weeks on just chucked away. The second time that we did the design document we sat down for about 6 or 7 hours over a 3 day period and just hammered out what goes where in a systematic order. We had huge discussions but the result is a fully thought through mod that can be completed. Another example is a secret project that I am working on, I don't have a design document because I wasn't sure where I wanted this to project to go. This has lead to me wasting a lot of time as the code was a complete mess. I also hadn't though through several major features and just coded them and so I have had to go back through a redo them as they don't link properly with the rest of the game.
I think one of our original problems came from one person forcing themselves to think up ideas to fit into a structure. I have found that it is always better to talk about it with others in the team. This has several advantages as it allows for more ideas and also development, because not all ideas that you think of are going to work and you often need someone else to look at it and tell you. Sparten07 and I have spent the evenings of an entire week discussing each unit/structure/upgrade/power until we believe that it is workable and fits in. I would also encourage other people in the team to get involved. Maybe leave the development of ideas to 2 people so it doesn't turn into a endless battle, but let everyone submit ideas as it will increase motivation of people and because they might just have that idea that you have been lacking. It is also a good idea to brainstorm around one idea at a time. For example we started with the barracks units and then, after all of them had been thought of, we went on to the war factory.
When I first joined the mod team the biggest problem we had(and it is still going on now although getting better) is a lack of communication. No one knows what others are doing and this leads to fragmenting inside the mod team and makes the chance of the mod dying a lot higher. It is also nice to see progress, so what we at GBF are doing is to post the entire design doc and a spreadsheet with detailed progress for each unit onto our staff forums so that people can see how well we are doing and track the progress of the mod. I also allows for people to feel like they are part of something bigger.
This is where GBF has failed badly. You need a structure to your news posts. They don't need to on a set date or anything but the more you can do the better. At GBF as soon as a unit was done a screen shot was made and unloaded to the gallery without any kind of announcement at all. I believe that this generated less interest in it than if it have been in a properly constructed news post with another unit. Also, keep a track of what units you have posted and if they change or are deleted make sure that you update the gallery as this could be misrepresenting your mod.
The first ideas that you came out with might not have been the best ones. I would encourage you to change your ideas if you get a better one. Just be mindful that a change will set your entire development process back and so at some point (possibly a few months after you finish the design document) you should put in a feature freeze. This means that nothing changes unless there is a very good reason to (for example it is uncodable). If you have more ideas, make a second release of your mod, but get the first version out first.
If you are making a add-on like mod like us, then it is really important to look at the other sides that you are going to be integrating this side into. For example Generals has a very specific structure for the amount of units with some flexibility: each side has a light tank and a heavy tank, anti infantry and a special tank. There is normally a very good reason for this, the initial game company probably spent weeks just perfecting it. It also makes it better to balance units as they have units to compare them to. This means that we needed these things as well. For an even better example of a very well integrated mod look at shockwave. None of the units look out of place in it.