Use marketing as a means to gather more data and test. Marketing is one of the most underrated and undervalued skills in startups.
Startups don't struggle with competition but getting attention. Marketing drives more data into the product, thus you can validate your theories.
Pre-product market fit marketing should drive acquisition and customer development.
Every startup needs to do some kind of marketing to get net new users to talk with / test product features with, etc.
Marketing pre-product / market fit is usually done to learn faster, not to scale the business.
So, yes - if you're running marketing hopping that it will fix lack of traction, it'll usually fail.
The engine needs to be running and well tuned if you want to pour more gas into it.
Not necessarily, if at all. I'd say "optimizing" marketing will be difficult, but marketing and acquiring users is totally doable. The problem will be retaining those users, but that's ok too.
When your in the "hunt" (ala the search for product market fit) you need prospective customers to come and test your product in order to get feedback to find out what needs to be modified/created to hence find product-market fit... right?
So, spend some money marketing to get some users, hopefully you've got pseudo-product-market fit so customers will like what you have just enough to give you feedback to make it better, then iterate, market again, and see what they say. Eventually you can optimize marketing for cost and quality.