[saturdayplace]

I have no experience, but here's my reasoning:

At the moment you don't have a lot of traffic, so you likely don't have a lot of users, and you don't really know if you have something they'll pay for. You don't really know what problems they experience using your app. You don't really know where else in the process they might fall out of the funnel. Putting the CC requirement at the beginning will even further restrict the flow of users through your funnel and make it harder for you to learn where your stuff is broken.

When you do have traffic, A/B test to find out which one works better for you BTW: there is likely more than one axis with which you'd measure what is "better". You'll have to decide for yourself what those are. Is it better than you have less tire-kickers and more seriously committed users? Is it better to have more people in the funnel in total so you can optimize the trial experience? Is it better to have pre-authorized the credit card and remind them that you're going to start charging them tomorrow if they don't cancel their account?

[ramenable]

These are excellent points, especially about validation.

Question to OP - Why not have a freemium model - where if the user does not pay then his account is downgraded to free plan. This generally has the benefit of keeping those users interested who may pay for the service sometime in the future.

As an example - I used Zopim on my website for a whole year for Free before converting to a Paid user. Had Zopim not had a free plan, I would have tried other competitors and eventually become a paid customer of theirs. By keeping a free plan, they retained me as a customer.

[iosnoob]

Good question - we see a lot of value in our tool, hence are leaning towards paid rather than free plans ala basecamp or zendesk. We're also bootstrapped so really focussed on bringing in revenue earlier..

[lifeisstillgood]

That's an excellent solution to an awkward problem.

[iosnoob]

Good question - our definition of best is the one that would maximise total revenue coming in (not so concerned with the number of accounts). ie- weighing the negatives of dropping more people at the start vs increased revenue of the remaining people who do sign on and give credit card when they remember and are interested in the service..