Selling expensive enterprise products does have a longer sales cycle than cheaper consumer products, and increments to your revenue come in more discrete blocks. But that doesn't invalidate pg's argument about growth: you just have to use other proxies to estimate your trajectory.

For example, instead of active users you track a number of leads and how interested they are. Even without a single closed deal, you can measure your sales pipeline. It is an imperfect proxy for potential revenue growth, but so is daily active users for consumer internet startups.

Thus, if you are planning to sell an enterprise solution costing $100000, and you think you are going to make these kind of deals a few times a year in the beginning, then you use your sales pipeline measurements to estimate your growth with a finer granularity than a big sale now and then.