OK, then set up 15 recurring $5 monthly donations. Nothing to track. 15 sub-$1 Amazon purchases is a great way to get the interest rate lowered or the thing shut down entirely.
I appreciate your train of thought and concern, but where's the line? If 15 sub $1 purchases are "bad", then why would 15 $5 purchases be ok? The bank could deem them "inappropriate" as well. So should it be $10? $20? Or is it not determined by dollar amount but rather location? If we look at your "15 meals for $5 each suggestion", then it might cause an issue if I like to eat lunch at the same restaurant everyday (we are creatures of habit, after all). The bank could deem that inappropriate as well. Do I need to create an algorithm that will provide me with a random frequency of locations to spend money with an equally random purchase amount to ensure I'm not putting the program at risk? All of this is speculation.
This is what I
know: The current method I've decided to use is the most efficient method possible within the "rules" I'm currently operating under. Isn't being financially efficient something we all strive for? I'm not going to try and speculate what the bank is "thinking" and use a less efficient method under the assumption that I'm not putting the program at risk, because I have precisely zero evidence of that being the case (or not the case).
Now, if the bank sends me a message like Elle8, then that's fine. At that point in time I'll change my strategy to find the most efficient method under the those newly defined rules.
At the extreme end, if this very rare [here's me speculating now ;)] strategy that a select few of us use at different banks in different locations with different sets of operational guidelines somehow causes the complete breakdown of these programs, then I'll eat my crow with a huge side helping of humble pie. But until then, I'm going to keep on keeping on.