First thing to do is start with one problem you want to solve. You can add more problems later, but pick one to get started on.
Next, narrow down the problem. For example, "education" is too broad. Narrow it down to something you can measure.
And narrow it to something you think feel passionate about that you think "they" should fix it, e.g., something so broken that somebody should do something about this!
In my opinion, some of the best goals are those that affect you personally so that you can see the impact in your own life as well as others. For example, I was bothered by junk faxes so after a 3 year crusade, my efforts led to the collapse of the world's #1 junk faxer. I now get a lot fewer faxes and so does everyone else in the country.
For example, "the high school drop out rate in major cities is about 50%.... that is abusrdly high!"
Now, narrow the problem even more and make it as local as you can because trying to fix that everywhere is hard to do! Small problems are easier to fix and you can then leverage that fix later.
Example: "the high school drop out rate in Oakland, CA is over 50%. Someone ought to fix that"
Now, narrow it even further. Go as far as you can. Example: "the drop out rate at Castlemont High School in Oakland, CA is 63%."
Now that you have a narrow problem that is inspirational to you, set a big hairy audacious goal to solve the problem.
Example: "I want to cut that drop out rate at Edison High School in Oakland, CA in half in 4 years."
Excellent. If you can achieve that in that school, then you can go on to bigger problems like how to apply the solution you just got in that problem to a wider area, e.g., to all Oakland schools. And if you can achieve that, you can then expand to all California schools. Then to all schools in the US. In short, solve one problem narrowly and then expand the solution outwards.
Now, the "local community" approach doesn't always work, e.g., if you are involved in nuclear disarmament, you can't start locally; you have to start nationally. Air quality problems caused by fuel emissions probably can't be solved locally. You probably have to attack it at a state or federal level, or start with a specific car company.
And some education problems lead to solutions that are only possible on a national scope, e.g., if the solution to the drop out problem at Edison High is the lack of national standards that can only be imposed by the federal government, then you need to acknowledge that and put that in your plan.
However, in our example, because other schools in the state are doing well, and it is only the inner city schools that are doing poorly, chances are excellent that the problem can tackled on a purely local level and then those solutions can be exported to other inner city school systems within the state (and then within the country).
OK, so you have a problem that is possible to solve.
Next, recruit a small panel of 3 or 4 subject matter experts (or do it yourself if you are so inclined) to do the homework and research what others have done. Has anyone solved this problem before? How did they do it? Can it be replicated here? This is really important. It is stupid to solve a problem that there are already good solutions for unless those solutions are inadequate for some reason.
If no one has solved it, then your experts need to come up with a business plan for a pilot project that might have a chance at achieving the goal you defined.
In either case, your panel creates a business plan for how you are going to achieve your goal.
Then you have a panel of other experts as a sounding board.... not the same group, but an independent group. Does the plan make sense? Does it have a chance of working? If so, you then execute the plan which may be a feat in itself!
Execution may involve lots of your time, or you may be able to persuade (or pay) others to manage the execution for you. Your choice.
And presumably that plan may involve cash investment, like some of your money to incentivize the behavior required to achieve the goal (i.e., behavioral changes that are required by the plan) or to simply provide the investment capital required to implement the plan.
So your contribution is a) you came up with the goal b) you recruited the team (or hired someone to recruit the team or started a foundation that recruited the team) that came up with the plan and c) you helped fund some or all of the plan, and d) you managed the implementation. You can do any/all of a, b, c, d.
Not much different than starting a company.
Some people might focus their charitable dollars on the research and planning stages and turn it over to others to provide funding. Some people might do the whole thing themselves, funding the plan and funding implementation. The choice is yours sometimes. Other times, the burden will fall on you to fund the implementation of your idea if you want to see it succeed.
Here are a couple of examples from my life.
In my day-to-day experiences, I often run across things that need to be fixed. For example, last night I learned that the charter school that my kids attend is forced to raise 40% of their budget from parents because the law doesn't allow parcel tax funds to be allocated to the charter school. This was clearly an oversight in the law that should be fixed. But there is no bill in the legislature to fix that inequity probably because the teacher's union would be against it (since it would mean less money for them). So you have this one big constituency that is against such a bill and a small number of people who are adversely effected being in favor. How do you get politicians to vote for something that is clearly the "right thing" and "fair thing" to do? That would be a wonderful cause for someone to take on.
Another example is medical research. I set for myself a goal of curing all major diseases. Of course, that is way too broad, so I picked a specific disease to focus on: glaucoma. Next, I gathered experts together who advised that an interdisciplinary approach to the problem might be a new approach worth trying since other approaches had not been successful. I hired a medical specialist to manage this and she came up with a collaborative model for medical research and carefully selected medical researchers that could put their self-interests aside and work as a team. She also arranged a partnership with the Glaucoma Research Foundation so that we didn't have to supply all the funding ourselves. The collaboration has been very successful and has resulted in new important understandings of how glaucoma works. It may lead to a cure.
I had over $5M in capital to invest, so I started a supporting organization at my local community foundation. The advantage of a supporting organization is I got a lot of assistance that I otherwise would not get if I started a private foundation. There are additional benefits as well that make a supporting organization superior to a private foundation. The only downside is you don't have absolute control, but to me that's a benefit because my foundation won't let me do anything stupid. Having the supporting organization means that I can say "this is what I'd like to see solved" and then there is a whole staff that is dedicated to getting it done.
The choices are yours to make. If you want to solve specific problems the advice above may be helpful. But you could ignore all of it and just invest in existing causes, e.g., you might choose just to fund science fairs all over the country or just give it all to the American Cancer Society. It all depends on what problem(s) you choose to want to solve.
Keep in mind that there are no right answers; there are only reasonable directions to take. It's your money. You get to decide how you want to spend it.
Have fun and make a difference.