Also, I agree, our pricing is too low. Our focus right now is on developers who can use Codiqa for work purposes. That has a higher value than $10/mo.

We are going to experiment with some of these ideas and blog about them. Thanks again.

Does anyone else have any other opinion on this? I am currently trying to decide whether CC should be a mandatory field or not for my app (which will have a 30-day free trial). What kind of impact on conversion are we talking about?

I understand that have data on this and the customers who converge outnumber the ones that you loose, but it might diver per demographic. Paying with credit cards is less common in Europe than in America, so that might play into it. What might help is get a seal of approval from some organization that you are safe. If you have social proof it's even better.

EDIT: I might also be worth considering if non-paying customers are really just a liability to you. I understand the bottom line is what counts at the end of day, but free customers might recommend your service to someone else or they might come back later when they really need your service.

I for one, subscribe to a lot of SaaS but would never put in my CC number just for a trial.

And what about corporate customers? I think many subscribers can come from corporations where someone somewhere tried the product and liked it, and then made sure his company paid for it. If he had to enter his CC# before he could even try he would probably not do it.

Data points of HN readers who would never do X or would be annoyed by Y in the process of evaluating your software.

Fixed typo.

I think the target population matters. If you're selling to teachers or dentists, then you should speak to teachers or dentists to learn what they will do or not do.

If you're offering a jQuery framework, then HN seems pretty relevant as far as the people susceptible to subscribe to your services.

I get a notification when anyone unsubscribes and, if it looks sort of borderline (e.g. sent one test phone call on the first day, never used the service again, and then canceled on day 33) I offer to refund their money. I get about 50% uptake on that. This makes me sort of anomalous in the industry (take a look at e.g. 37Signals policy on refunds -- "We. Don't. Refund." -- which is much more common) but it works at my scales because AR doesn't have so many accounts at the moment that sending an email or two a week is a big tax on my time. (And, of course, I don't lose sleep at night over $9 or $29 or $79 in revenue. It is literally rounding error for me -- if the yen hiccups that is much more painful.)

At the risk of stating the obvious: the business model is not hoping that people will forget about their subscription to Appointment Reminder. The business model is in convincing them to log into the site every day, run their business off AR, and call my cell phone at 3 AM in the morning to scream at me if anything goes wrong because that way I can get them to pay me thousands of dollars.

A note, in our case we used PayPal for subscriptions.

However, as a way to weed out non-serious users, I can understand the rationale (but then again you would let go of publicity effect).

You're not trying to get people to forget about the service and pay for one month then cancel. That's not a long-term success for the business. You're trying to extract a commitment to pay money because extracting a commitment to pay money automatically increases their perceived value of your service, their engagement with it, and their likelihood of paying for it.

Forcing the decision-point until when the trial is over will cause them to discount the value of their software because a) they've already consumed it or b) they haven't had time to look at it yet but have had time to totally forget your sales copy about it. You want the big decision point to come at their moment of maximum perceived desire for the software: when they've just come to your website with a headache and you've just told them you're selling aspirin, the magic cure for headaches. At that moment they burn with desire for aspirin, so get agreement in principle that if the aspirin cures their headache you get paid, rather than saying "Take this aspirin and then let's talk about how much you value not having headaches four weeks from now."

After you get people to make the decision, all of their future actions will tend to reinforce that they previously made the correct decision. (People hacking 101.)

These companies ranged from 2 people to 400+ people. Most SaaS offerings are under $50, which seems to be something that most business people won't hesitate about spending money on and certainly wouldn't require someone to go to someone else to approve.

If this is the case in your company you're wasting a lot of time!