This is something I've experienced first-hand, and an extremely valuable lesson for people thinking about starting a startup. At a low average monthly cost (say less than $30), it's really damn hard to get enough people into your funnel for a reasonable cost of acquisition. And you need to reach a LOT of people for your service to be a significant source of long term income.

If you're about to start a startup, I'd strongly encourage you to come up with some hypothetical figures for these metrics:

* Number of monthly uniques

* Percentage that will convert to trial

* Percentage that will convert to paid

* Average monthly subscription cost

* Churn rate

* Cost of acquisition

You'll be way off, but run the numbers you've made up past as many experienced founders as you can, and you might be able to come up with a model that's semi-realistic. Of course it will evolve over time as you get some real data and more experience, but the main point of the exercise is to then ask yourself questions like: "How can I get X0,000 uniques to my marketing site on a $0 marketing budget?", and "What can I change about my product/positioning/target market so I can triple my average monthly subscription cost?".