They’re there so that the sales person can extract a variable amount of revenue from you based on how much value you’re getting from their system, and how much you can afford to pay.
Utility pricing assumes that service volume is high and can mostly be recouped at slightly above infrastructure costs. In that case, utility pricing is a win for most parties.
But in smaller services, where R&D costs are almost certainly more relevant than infrastructure, sales must recoup R&D. You may have a customer with 500 instances, but extracting low value per instance, paying the same as a customer with one instance, for whom the service is business critical. From the point of the provider, both are probably a win. But if they charged the customer with one instance 1/500th of the larger customer, their service might be unsustainable.