Hillside ponds can be very complex, labor intensive, expensive or very easy and inexpensive. The difference is in the design. For example, if a person wants a 20'x20' pond on a steep slope that certainly can be done given enough concrete, engineering and money. It's done all the time. Gallon for gallon the most expensive pond that can be built.
Or a person can work with what nature has provided. The key is doing exactly the same as you would see in nature. Water in nature will erode and cut valleys, but water also follows the coutours of the land. We can do the same.
When we follow the contours we are actually building on basically flat ground. How wide we can build will be determined by the contours on the slope. Some sections will be wider than others. That makes it look natural. The more you force the hill to change the less natural it will look and the more work you will have to do and the higher the erosion risks will be. So it's win-win to work with what you have.
Best place to start is the bottom. This is where the pump will go.
How wide you make the pool will be determined by how steep your slope is and how large a retaining wall you are willing to make.
Retaining walls can be very complex and expensive. I mean retaining walls that actually retain. Many people build "retaining" walls they hope will retain. Some even do. You will always find a line of people ready to tell you examples of "we have a 6' high rock retaining wall that's been there for 100 years" to demonstrate that civil engineers are all wrong with their fancy and expensive designs. Well that certainly is true, a rock retain wall can last 100 years. What's harder to point to are the thousands of other rock retaining retaining walls from a 100 yeas ago that are no longer around because they failed.
Now I'm just relaying info. I'm not saying you shouldn't build a vertical rock retaining wall. To the contrary. I build stuff like that all the time. It's great fun. But what I try not to do is bury my head in the sand and just fantasize and listen to people cheer me into a big problem. I just happen to think that understanding what risks I'm choosing to take on let's me make better choices. I may not mind if a rock wall falls over, but I don't want it to fall over on a person, my house, my car, etc. A little understanding keeps me out of serious trouble. But that's just me.
Here for example is how a rock retaining wall can be made safer, actually easier imo, and last much longer. The example on the left is a true retaining wall. No question that soil is going to want to collapse so the rocks have to keep it back. Well a small child could tip that pile of rocks over.
The example on the right isn't really a retaining wall, it's erosion control. What we've done by cutting into the slope is make a steeper slope. The biggest near term risk is erosion. Greater the slope the greater the erosion. We can reduce erosion with a layer of rock. Erosion risk can be lowered more by laying down geo fabric first and even more importantly contouring the top so water coming down the hill is diverted around our cut.
The example on the right can still fail, to be sure. It depends on soil conditions and how well we diverted water. But the risk is way less than the near vertical wall.
One last plus. Say the example is 6' high. Say Grandma is over with the family to look at your wonderful pond. Bright sunny warm spring day, birds chirping. Life is good. Where's Grandma? Grandma walking down the hill didn't see the cliff and took a 6' header into the rocks below. Now we're off to the hospital. Now you understand why we have building codes. And all those people on the internet that were cheering you on and saying it'll be fine, you don't have to do all that code stuff? Why they will very nice and write very sincere condolences...and go right back to telling other people building codes are a waste of time. That's their joy in life.
The example on the right is much safer (not perfectly safe of course). It's easier to see, the fall would be less severe.
Adding plants on a controlled slope is much easier and effective than a rock retaining wall.
Easier to build, lasts longer, safer and you can add plants. Maybe there is something to a little understanding?
Finally the fun stuff.
Most people don't like the stuff above. Bah, bah, bah. And kind of a bummer. So this section is the fun stuff. I made the title bold so people can just skip all the bah, bah, bah above.
The "pond" becomes a series of streams, rapids, small waterfalls and pools. All together it can represent a surpring volume of water, a large pond.
The course you follow and placement of waterfalls is more art than science. But basically you go slow. Just try to dig down a little. Even set liner down and test the water course. A water level is a great tool for figuring out if you're on course. Here's a tip: A stream does not have to be sloping down. A 100 mile long perfectly flat stream will flow. However, in this case we are trying to get down a hill so we do want the stream to head down slope wherever on the slope it makes sense.
When you get to the edge of the property, or wherever you like, you make a switchback. Here we have no choice but to turn down hill. So we add a waterfall here to drop down. Now this doesn't have to be a single plunge. You can step it. And it doesn't have to be straight down the hill. Following the contour. Let the hill design the shape of the waterfall.
Drops can also be done in the middle of the hill. Basically some rapids.
The variation of waterfalls, rapids, stream and pools can create a very natural looking pond with very little digging (less than a standard pond) and at a very little cost. Plus it can be done in stages more so than a standard type pond.
Bottom Pool Size
Since you're still reading I can give you one more bah, bah, bah boring detail. The bottom pool, where the pump goes, has to be a certain size. Picture your hillside all done, the bottom pool is full of water and you switch on the pump. The first thing you will notice is it will be awhile before you see water coming out of the pipe at the top. The second thing you'll notice is as water comes out of the pipe the water runs into the stream and it takes a while to see water moving. The last thing you'll notice is water flowing over the bottom most waterfall and into your pond. At least I hope you notice this process.
There is a time delay from when you turn on the pump to when water starts flowing back into the bottom pool. That can be a surprisingly long time. Until water goes back into the bottom pool the pump will be happily pumping water out of that pool...and the water level will drop...a lot. Some people freak out and thinks there's a leak. It's called drawdown and is normal. The question is will the bottom pool run dry before water starts refilling it? If the pool runs dry you're a little screwed.
The idea is to make the bottom pool large enough that it doesn't run dry, that the water level just goes down. Then you can just top off the pool with more water (while the pump is running.
I'm a mind reader...you're thinking "how big should the pool be? Unforfunately that's impossible to tell. The bigger the stream system the more water it takes to charge. The bigger the pump the faster the drawdown. Experience is a great help but if your reading this that's probably not an option. So just make the bottom pool as big as you can and then use work arounds as needed.
Filling the bottom pool to it's max first and leave the hose running. The hose may not keep up with the pump but who knows it might be enough to keep the pool from running dry.
Pre-fill the stream with the hose. Basically fill the bottom pool and then move the hose to the top and run the hose until water starts flowing into the bottom pool and then turn on the pump. You have to be a little careful water from the hose doesn't go into the pipe and back down into the pond (see check valve below).
Putting a check valve right after the pump on the pipe going up hill to the top will stop water from flowing in the wrong direction. Without that when you turn off the pump all the water in the pipe, and it can be a lot, will flow down backwards thru the pump and into the bottom pool increaing flooding risk and probably stirring up dirt in the bottom pool. A check valve stops that and is pretty cheap. You want a "swing check valve". It just has a flap that is easy for the pump to push open. Other types have springs which the pump has to work against so you get less water flow for the same $$$.
The pump has to have enough lift to push water up the hill. So a 1000 GPH pump may only do 50 GPH if it has to push 20' vertical. Or zero GPH. The easiest thing is to buy the right pump. That can be expensive. You can slpit up the system. In the pictures above there is one meandering stream coming down the entire hillside. But that same system could be 2 streams, or 3, or as many as you like. Each horizontal section could have it's own pump. That way the lift needed would be like any other pond, 1-3'. In this scheme these are separate streams, separate water, separate pumps. They can be built close together to appear to be all one stream. Kind of like Escher.
I know I've forgotten some things, but that's the basics.
The info above has more to do with how DIY people want to build and less to do with proper building techniques. What I've shown above is great as long as you don't cut too deeply into the hill because that requires an actual engineered retaining wall. What I've shown above is very much like some soil erosion control techniques, minus the water in the stream of course. The rocks divert and slow water coming straight down the hill. Howver, you still have to inspect for erosion and fix any problem areas because these things will change drainage on the hill.