Some advice from a prospective customer: if you can't/don't want to offer free plan, at least give me a trial that's long enough to actually try out your product. Two great examples that come to mind:
1. Mixpanel vs. Kissmetrics: Mixpanel gives me a free account that I can use for as long as I want to evaluate all the stuff I can do with it. KM gives me a 14-day trial. I'm the only developer working on this project which means that 14 days is not enough time to implement the changes I need and actually start to see the results. Going with Mixpanel for this one was obvious. I will reach their free 25k data points way before the first month so I have to get a real paid plan or put a banner on my website. In any way, they won already. Even if I wanted to switch to KM for the lower prices, exporting the data I currently have would be a hassle.
2. The New Basecamp: Although they don't offer a free plan anymore, they do have a 60 days trial. They don't really lose much by offering those extra days and after two months I'm already hooked to their service, I know how everything works, and most importantly, they have my projects tied down. The work of those 2 months is already there so even if I don't like the product, I'll probably pay for it so I don't have to export it. Joel Spolsky has a great article that talks about this retention thing (his specific example was with email providers) but I can't find it right now :(
Rob Walling on Startups for the Rest of Us had a different take on it. One of the things that dropping to a shorter (14 days vs 21 vs 30 vs 60) is that it lets you iterate your marketing A/B tests much faster.
Which matters more, iterating tests faster, or getting to a better result? In my books, a better result.
Here we have an actual customer here saying that uselessly short free trials will drive him to a competitor that offers a useful free trial. And the customer had an actual example to point to of it happening. If this is a general trend, then it is impossible for the best optimized short free trial to match the best optimized longer free trial.
And yes, long conversion cycles are a PITA. However they are sometimes a necessary PITA. Furthermore a long conversion cycle (if needed) is a useful barrier to entry for competitors who will need more float to survive, and who will learn more slowly.
Absolutely love Rob, he's helped us out at www.serpiq.com quite a bit. We moved from freemium to a 7 day trial with the launch of v2.0 of our product, while at the same time raising our prices. While we did lose some customers due to switching to Stripe instead of another payment gateway, our metrics are looking really healthy and we have the opportunity to iterate a lot faster with a 7 day trial - then again, our product doesn't require a lot of time to realize value.
1. Mixpanel vs. Kissmetrics: Mixpanel gives me a free account that I can use for as long as I want to evaluate all the stuff I can do with it. KM gives me a 14-day trial. I'm the only developer working on this project which means that 14 days is not enough time to implement the changes I need and actually start to see the results. Going with Mixpanel for this one was obvious. I will reach their free 25k data points way before the first month so I have to get a real paid plan or put a banner on my website. In any way, they won already. Even if I wanted to switch to KM for the lower prices, exporting the data I currently have would be a hassle.
2. The New Basecamp: Although they don't offer a free plan anymore, they do have a 60 days trial. They don't really lose much by offering those extra days and after two months I'm already hooked to their service, I know how everything works, and most importantly, they have my projects tied down. The work of those 2 months is already there so even if I don't like the product, I'll probably pay for it so I don't have to export it. Joel Spolsky has a great article that talks about this retention thing (his specific example was with email providers) but I can't find it right now :(