How to do a trial offer?

We'd like to offer 30 day trial for a subscription service.

Has anyone experimented with that?

We will probably do a bunch of testing with it but I'd really like to hear your experiences on:

1) Length - 15 or 30 days?

2) Do we get the customer's credit card info at signing up or after the trial, to reduce friction at signup?

3) Is the trial supposed to be free? Or do we charge for it if the customer doesn't cancel?

3) Do we charge when the trial starts or when it ends?

While BasicallyMoney.com offers a good consumer's perspective, keep in mind that you're in the business to make money. Everything single thing that you can possibly do (even, say, give away free $100 bills on the street) will turn people off.

Don't be afraid of turning people off. That's called qualifying, and it's a critical process to making sales. But the difficulty for you is that the answer to every single question you're asking is "who knows, it depends."

1) Length - 15 or 30 days?

How long will it take to evaluate the service? Hoover's offers a twenty-four hour trial. Microsoft Servers are often 120 days.

2) Do we get the customer's credit card info at signing up or after the trial, to reduce friction at signup?

Someone who gives their credit card number is a much more series prospect than someone who just fills out a form. Who do you want in your sales pipeline? What's the conversion rate difference?

3) Is the trial supposed to be free? Or do we charge for it if the customer doesn't cancel?

It doesn't have to be. For services that need longer commitments (6 - 12+ mos), it may make sense to allow people to try it out for just a single month at 1/12 the cost of a year subscription.

3) Do we charge when the trial starts or when it ends?

If you're going to charge for a trial, I'd charge at the start. If the service was a poor fit, it's better to have the payment for that service be as far away from the realization of the poor fit as possible.

Basically, there's no way of knowing the answers to these questions until you try them out.

If you ask for a credit card number up front, you will turn people off. I guarantee it. If you charge if the customer doesn't cancel, i.e. "negative option billing" for your trial, you will turn people off. I guarantee it. If you charge when the trial starts, it isn't really a trial then, is it? You will turn people off. I guarantee it. Make it easy for your prospective customers to try your service. Bend over backwards to help them understand what your product is all about. If they get it, they will open their wallets, and spread the word. If you try and force their wallets open, you'll get nothing but bad press and chargebacks.

Last, go at least for 30 days. People are busy nowadays. It's easy to sign up for a trial and forget about it for two weeks.

We've gone a couple different ways with our apps. And it's really depended on who your target customer is. In one of our apps we offer a 45 day free pilot, no credit card upfront, etc. But this app's primary customers are large businesses whose employees don't have the authority at the "let's try this out" point to pay with a credit card.

With another app we've built, and sell (tgethr ), those customers are small businesses and families. There we offer a totally free plan, and the other paid plans do ask for a credit card up front. This is a great way to protect yourself from people abusing your most resource expensive plan up front. With some luck you'll get some press soon when you launch, and if your application can get broken up into plans, you'll have a few more hosting headaches if you let anyone and everyone sign up for a plan that maxes out resources on your server(s). Resources like email. Asking for a credit card to make people more serious about trying your resource intensive plans helps reduce that headache.

Remember that the goal of a TRIAL should be to address a customer concern/objection.

Some folks erroneously think that a free trial is an actual marketing tool. A trial is seldom useful enough to cause someone to learn how to use the software. Reducing the cost of something (even to free) does not increase the demand for it, it merely removes one objection ("it costs more than I want to spend/than it's worth").

And often, a Money Back Guarantee is a better way (than a free trial) to address those concerns.

So... if you're offering a trial, what objection are you trying to overcome? How long a trial do you need to accomplish that. Some objections that a Trial can overcome:

• Will this run on my computer?
• Is it as easy-to-use as you say it is?
• Does it have feature XYZ?
• Will this work for my needs?

What a trial will not overcome:

• This costs more than I want to spend.
• I don't have time to figure out what I want to buy. ( A money back guarantee is often better in this case b/c a free trial adds an unnecessary step).

I thought Seth Levine had a great discussion on how long a trial period should last. The author argues that 30 days is way too long for most startups. He argues:

1. You're probably doing something wrong if you can't show the product value within a couple of days
2. Customers who will pay you after 30 days are likely to pay you much faster, so you're leaving money on the table when you have a long trial period.