Aight, generally I start with a base idea. Normally, this is things that I have lifted from other places.
As an author, you know damn well theres no such thing as an original idea. What we do now is more our own unique autistic takes on what someone has done before, probably better. So, lift those ideas, but don't make it too noticeable. Mostly, I go more for themes than a one to one lift, generally taking the core idea of what I want, and then warping it to actually fit in with the world I want to make. So, I want a robot army, and I like the Vex from destiny. But I don't need radiolarian, or the timetravel, and I think that the look of the Vex is fake and gay.
Make the robots drones, add in some heavy support, and an overarching machine intelligence. Add more as I go along.
Spreadsheets help.
The key to being an author is that you need to know everything. Basically, if it affects the story, you need to know it.
However, if you try to put that everything in the story, your gay and retarded enough to deserve a thread. Think of world building like an iceberg, if your on the surface like a reader would be, you should only ever see that tip poking out above the waves.
You can also thing of successful worldbuilding like a series of webs. At the very heart of the web is the story. And each time an aspect of the worldbuilding intersects the story, it makes it own micro-web of ideas branching off. The reader should only see some of these ideas, not all. What you are looking to do here is to create the illusion of a real, breathing world, not an actual breathing world. Where you can, leave things shrouded in mystery so as not to waste the readers time with unnecessary shit, but at the same time, make sure that the readers understand something is there. In the real world, people go through their days all the time and find things that they don't understand. You have the answers for your reader if they ask, but otherwise, they don't need to know.
Research is key. Seeing how other people have done this via history, or how other authors have done this is both valid methods. The internet is out there, use it.
If there are things that I like to focus on, its probably logistics.
I actually went into the stock market and used one of those tree graphs that divided every industry in the world up and conveniently labeled them. This gave me basically all the categories my budding civilizations needed to be successful. Taking these categories and putting them in a list, I went through them one by one, answering the questions that they raised as I went along. As I did this, I also considered natural advantages and disadvantages. If you have say, infinite clean energy, then there is going to be more electrical cars than there is petro-fuels so long as the grid can handle it. However, further out, where there is no grid, you might have petro-cars. This naturally makes a divide between the people who have reliable acecess to the extensive grid, and the people who don't. These fracture points are places you can make distinct subcultures around: IE, the smooth, polished city slickers, and the hicks in the country.
Things like rivers are going to change the trade routes of a country. More goods going by boat than by truck for example.
Cities tend to be industrial centers. Raw materials flow into them, finished products come out. This means that for each city, you should have a web of small towns that supply the city with the resources.
Stuff like this is useful for when you write wars and conflicts because you understand where the enemy could hit to do damage. They don't actually have to bomb london-town, they can just have their tanks roll Shit-ville and suddenly London-town doesn't have the raw iron ore they need to make swords. Or they can embargo the trade route between London-town and Shit-ville.
Figuring out the logistics of a place gives you a real insight of what the culture you are trying to make is going to consider important. If they are smack dab in the middle of an ugly desert, then you are probably going to see lumber as a luxury item. Really nice lumber would be to domain of rich guys and nobles. If they are by a river, they are probably going to have a bunch of river dieties, and the river is going to have a place of religious importance because of how vital it is to everyones survival. If they are in the land of japan, with really shitty iron, they are going to have Damascus steel to get those impurities out.
This gives you alot of really small, but really useful tidbits for you to just drop in wherever you want. Its small stuff like this that makes a world really feel alive.