I think there's room for both. Going on a $ per kg metric is drastically oversimplifying the other factors that go into launch vehicle selection. SpaceX ridesh

I think there's room for both. Going on a $ per kg metric is drastically oversimplifying the other factors that go into launch vehicle selection. SpaceX rideshare is great if you have a mission that you want to launch to a popular orbit or an orbit near enough to a popular one that you can close the gap with on-orbit propulsion. However, the cost savings come with disadvantages. You don't get to pick your launch date, and if you miss your launch, SpaceX isn't waiting for you. It could be a long time before the next rideshare that's going to the orbit you want is available. If you want an oddball orbit, or you want the flexibility to set your own launch date and maybe delay it if you have program delays, the smallsat launchers could make a lot of sense even if they are 5-10x the cost on a per kilogram basis. Space missions are really expensive, and it's not the launch that's the driver of that anymore - it's the engineering, customization, and other overhead involved with developing and operating a spacecraft. You can (and should) use off-the-shelf hardware such as a finished satellite bus from somebody like SSTL or Blue Canyon, but you still have to develop the concept of operations, requirements, payload, operations plan, operations infrastructure, and staff the operating team (which is often a 24/7 endeavour, or at least an 8/7 endeavour with on-call night staff). To put it another way: if your total lifetime mission cost is going to be $50M-75M (not an unrealistic figure for a novel SmallSat mission) and you're looking at $1M for SpaceX versus $7M for RocketLab (to use the numbers from the article), the answer may not be immediately obvious. There may be serious advantages to a dedicated launch at the cost of growing the program budget by about 10%. It's totally situational.