It's not that hard - 9 year olds aren't stupid. You just have to show them the change in pain a couple of times for them to "grok" it.
The pain, in Factorio terms, is having to sit there not doing anything. Boring. Once you set up an assembler to do it for you, suddenly you can fill that time with whatever you want to. The simple problem/solution/reward loop Factorio presents around automation works remarkably well.
It's probably more along the lines of incentives, trade-offs, and pragmatism. If you can survive and have fun while making everything manually, why bother to automate? If suddenly, the proverbial bear is gaining on you such that you realize that the status quo won't cut it down the line, then the cost of automation becomes worth it.
If anything, it probably means that the nine year old is faster at making a known factory manually than the parent.
And in this case, the "bear" is biter evolution. The biters (enemies in the game) get stronger over time. At the beginning they're trivial to kill with your basic weapons or turrets, but as they get stronger you'll need walls, AP ammunition, and then finally laser turrets (in increasing numbers) to fend them off. If you aren't automating construction of this stuff then you'll fall behind and get killed.
I discovered this in my first play-through because I was prioritizing high tech research over military research, and thus didn't get laser turrets until well after I needed them to take on the stronger biters. I had to roll back an hour and change my research priorities to get them in time to prevent getting stomped.
The pain, in Factorio terms, is having to sit there not doing anything. Boring. Once you set up an assembler to do it for you, suddenly you can fill that time with whatever you want to. The simple problem/solution/reward loop Factorio presents around automation works remarkably well.
Optimization, however, is harder to teach.