If there is something I’ve learnt working with engineers is that having a good research question or a good idea in mind is not enough. You need to have a good method as well. “Method” comes from Greek and it means “following what way” you mean to investigate something. In all the proposals, you have to explain what method you will follow to achieve what you are proposing, but it's a very tricky point. The problem is: how detailed do you need to be? How much can you see into the future? There are two extremes: on one side, you can be very detailed and say what will be done at month 1, month 2 ... month 12 etc.; on the other, you can be quite vague, speak about the goal and be fuzzy about “how to” reach it. Some programs ask you to be very detailed, but this, in some way, clashes with a research activity, especially if it is a very innovative research. How do I know what I will do at month 24 if I am moving into unknown terrain and I don't know yet what I will find, in some way, by definition since I am researching? If I already knew exactly what to do and when, then it would not be research. On the other hand, being very vague gives the idea of not having a clue about what to do and this is no good, of course. There are only a few cases in which a vague idea may be ok: venture capitalists, for one, can appreciate a quite visionary idea that is still, somehow, undefined. But in cases like this, another important factor would come into play: who you are. If you are very famous, very credible, or if you belong to a very credible institution, your vague idea does have a chance of being financed. If you are completely unknown, it is of course more difficult. So, there is no single answer on what to do: it depends on the country, it depends on the research program. European calls, for example, do ask you to be quite detailed and you have to manage to be. The method issue goes hand-in-hand with how you sketch the temporal plan of the project. You want to do something super innovative and completely fuzzy: you are planning to create a teleportation machine like the one featured in the Star Trek series. How long will it take? Nobody knows. And still, if you are asking money for it, you have to say something about your temporal plan. Like for the budget, which we shall discuss in a minute, don’t ask yourself the question: “How much does it take to create a teleportation machine?”. Rather, see how much time is given to you in the program you are applying to and fit your schedule into it. They finance you for 12 months? Then ok, you can design a teleportation machine. They finance you for 24 months? You may get to a prototype. They finance you for longer? You may get to a beta version and even evaluate it with some brave volunteers. And so on. How much resources do you need, to carry on your project? This is one of the most relevant parts of a proposal. The reviewer is going to look at it for sure, so you want to pay the utmost attention to it. We will deal with it from the communication point of view. There are 3 main points you have to consider. Number one: you have to fit the proposal into the available budget. Although in principle you should decide what you mean to do, see how much it would cost and calculate how much you need, in practice (most of the times, of course) you would do the opposite. You have to see how much the available budget could be and decide what you can do based on that. Then, when you write the proposal, you kind of “go back to the theory”: you state what you want to do, you calculate how much resources you need, and that is your budget. Which, this way, corresponds to the available budget. Number two: pay attention to the “value for money” issue. If you ask too much for doing too little, you are not credible and you look greedy. If you ask too little for doing too much, you are not credible either. Just like as if a restaurant offered a Michelin starred meal at low cost: nobody would believe it. Make your request consistent with what you mean to achieve. Number three: mind that the operational plan of the project is actually consistent with the budget you are asking for. Remember that human resources are, usually, the highest cost. How many people will work at the project? Will they be working in parallel? For how long? Make your budget request consistent with your plan. Let us now talk about you. The proponent. This is one of the parts the reviewer will look at, like finances, for sure. May be not in details but as much as needed in order to be sure that you can do what you promise to do. Introduce yourself as someone credible: if you have done something that anyone can acknowledge as particularly good, mention it, even if not strictly relevant. Are you a Turing award? Do mention it, even if your proposal is about bioengineering. Introduce yourself, and your group or your lab, as expert or at least knowledgeable in the field of the proposal. List previous experiences, publications, anything that may prove it. I want to finish up with a tip. Write yourself your own review, in the proposal. You need to know or fathom the criteria the reviewer will have to evaluate your proposal based on. Even better if you know the key-words: if there are key-words in the call, do use them, not synonyms, the very key words. They will tell at first glance that you are in line with the call. Then you can, for example, highlight in bold a series of sentences that “answer” the reviewer’s questions. Is this proposal innovative? You can write and highlight “This proposal is innovative in these 3 respects…”. “Is this proposal sustainable?” You can write in bold or put in a box: “The sustainability of the project is based on…”. Etc. If you have the success criteria clear, you can “write your own review”, highlighting with different strategies a sequence of sentences and pieces of information. You may even pay attention to the fact that if you read them in a sequence, neglecting all the rest, they still make up a meaningful message. And the reviewer is going to be grateful. She may still make objections, of course, but you are making the basics clear.