Indeed. Part of the reason that satellites are so expensive is how light you have to make everything; there's a huge amount of engineering that has to be done in order to achieve mass goals, as well as make use of the most expensive materials on the planet. An example is the use of things like top-end Spectrolab multi-junction solar cells, which get the highest efficiencies, but since they're basically lab-scale production hardware they cost two orders of magnitude more than what slightly less efficient panels on Earth cost (about $400/W).
Beyond the simple cost effect on engineering, there's the size effect. Look at James Webb and the massive expense they had to try to make it "origami" itself to fit into smaller launch vehicles. Or how many parts ISS had to be built out of in space, dramatically escalating both ground engineering / production costs and in-space assembly costs (the latter costing nearly $10m per man-day). There are serious expenses to trying to compensate for a lack of space or payload capacity when you really need it.
Then there's the size of the market. As launch costs have been dropping, the number of companies looking to launch payloads has skyrocketed. The skyrocketing launch demand has been reducing average payload development costs, as designs get more reuse. Both of these trends will continue as costs continue to decline. Meanwhile, new markets will continue to open up. Space tourism has always been hindered by the absurdly high launch costs, limiting it to only the wealthiest individuals. While it will remain a "small" market for the forseeable future, it can expand by orders of magnitude with reduced launch costs. Which again makes more demand..
Lastly, there's economies of scale. It's generally recognized in the rocketry world that - at least up to a point - larger rockets get a better cost per kilogram than smaller rockets. So wherein you can launch a lot at once - multi-satellite launches, large geo launches, large interplanetary probes, fuel depots for tugs/boost stages, shielding mass for manned missions, etc - you tend to save a lot of money by going big rather than using multiple smaller launches. The caveat is that just simply going big is no guarantee of a price reduction. Just like you can make an absurdly-expensive smaller craft, you can also make an absurdly expensive large craft. And even a "moderately priced" large craft isn't generally a "win" - if you can sell payload space for $5k/kg on a 10 tonne payload rocket and for the same price per kilogram on a 100 tonne payload, the vast majority of customers will choose the former. But if you're a company like SpaceX that's been delivering cost reductions on the small scale, and you can carry it over to the large scale, it gives you the potential to take the cost reductions even further.
If you can pull it off.