Hi Ardent_Formulations. I agree with your thinking. The idea of targeting keywords with less competition yet decent search volume is certainly a good strategy if you want to increase your chances of starting out profitably and with the best ROAS.
I’d start by categorizing keywords into two groups, those that are super relevant product (e.g. ‘best robot vacuum’), and those that are more vague/browse terms (e.g. ‘best vacuum’ - which could mean handheld, upright vacuums etc).
Next, build your campaigns around the super relevant terms, and save the more vague terms for later on when you’ve exhausted the super relevant ones. I suggest starting in small batches of 5-10 keywords, in the order you suggested (least competition, decent search volume), then only graduating to the next batch after testing the former.
Use all 3 match types. The reason is that when you build out your campaigns initially, it helps to think of it as mere data collection. We don’t yet know the actual CPCs and advertising cost per sale by keyword (average # clicks per sale x CPC). Only when you collect those data you’ll know which keywords you can afford (i.e. those that have a positive ROAS, and those with a negative). Only then will you also know how much you should be bidding.
Tip: To set your bids, work out your breakeven ACoS, so you’ll know which keywords you can afford, and which are too expensive. Those that are over your breakeven ACoS you want to bid cap, and those below, you want to push your bids to at most breakeven.
To answer your question about minimum searches, I suggest not going lower than 100, though there’s nothing wrong with experimenting. In terms of bidding, your mindset should be to try to push bids as high as possible as long as you stay within your profit margins.