No matter what you do, I recommend you have a clear grandfathering policy in place.<p>I am particularly generous in that department "Should you decide to return, we will reactivate your account at last billing tier even if that tier is no longer published on our site."<p>There are some conditions to that, such as material change of control events (all bets are off if you have been acquired).<p>Failure to do so results in upset customers and negative press coverage. Competitors love that. If you decide to lower prices, be sure to apply decreases to your existing customers. That's positive press.<p>I am experimenting with pricing as well and prefer my customers to be protected from effects of such experiments.<p>Annual subscriptions work great when they are a discount over monthly billing. The pattern tends to be that the user subscribes for a few months of month-to-month billing, realizes she is in love, and renews at annual level to save money.<p>I haven't seen the research, but I think you will find more annual subscriptions tend to be renewals rather than initial purchases. An annual subscription for $50 is below my pain threshold.<p>Let me illustrate my point.
I pay some small amount for <a href="http://eventid.net" rel="nofollow">http://eventid.net</a>. I don't ever remember what it is, but I pay it annually. It's $29 and I had to look that up. It's just not material to bother with quarterly $12 payments when I can simply pay $29 and ignore this for the rest of the year.<p>Eventid is a useful service and all my subscription does is speed up Microsoft knowledge base searches. When I am billing my client for time spent figuring out the problem, the speed of research matters and so I pay for this service.<p>Another example:
I pay for <a href="http://usaip.eu" rel="nofollow">http://usaip.eu</a> on a monthly basis $9. This actually just got lowered to $7.99. I primarily use the service to watch Formula 1 on BBC live and without commercials, but it also comes in handy for on-demand VPN connectivity that ALWAYS works when dealing with insecure WiFi networks (have you heard of firesheep? :)). I used to pay by the day, then by the week, now I pay by the month, and pretty soon I'll just pay for it once a year ($74.99) and again be done with it. That service also happens to bill me for European Union VAT, which I am technically not subject to, but again the cost is just not material enough for me to bother with.<p>I personally do not have free accounts at all. I have plenty of competitors who have free option. The fact is that should the customer decide to pay my competitor, they will likely re-evaluate the market anyway. My software integrates with too many things and simply delivers too much value. You get either 100 transactions or 14 days, whichever comes first. For most people, it pays for itself in one extra transaction as what my software does is maximize revenue per staff member. I do not yet have transaction-based pricing unlike say Recurly, but I will likely add that after we sign up enough customers.<p>I am guessing this is in regards to your water skiing log site? Determine the value and price accordingly. From what I know, water skiing is not a sport for poor people, so you may have price elasticity due to customer affluence. My snow ski equipment is quite expensive.<p>I'd probably keep the site free with a limit on how many entries can be made in a year and make a paid mobile interface as a subscription. As a skier, I would probably want to enter my data, complete with picture attachments, right after I am done with my run instead of when I get home. In essence, charge for usability. :)<p>One last thing - check out Wepay for subscription billing. Much cheaper than Paypal. They do not advertise that feature yet, but I spent some time with the team at the hackathon and was pleasantly surprised it was there. :)<p>Then again, I am just beginning with offering a SaaS solution. I could very well be wrong. I am easy enough to find on Twitter if you want to chat more. :)