How to Build an Effective Trial for Your Product

A “try before you buy” sales experience has become a pillar of the Product Lead Growth/self-service sales world, however getting a trial right is more nuanced than you might think.

We implemented a credit card upfront trial at Codecademy and had a huge conversion win, which taught me a lot about what makes an effective experience.

There are always people who will buy your product no matter what.

There are always people who will never buy your product.

A good trial experience converts the people who are hesitant but could be convinced.

How Product Get Sold Online

For 90% of products that are sold online, there are 4 core sales motions.

  1. Direct Purchase - e.g pay now for access
  2. Direct Purchase + Guarantee - Pay now for access, we’ll guarantee that this will work for you or your money back
  3. Trials - Experience the paid product for x days, y uses, etc then decide if you want to buy
  4. Freemium - Certain things are free forever, certain things are paid forever

In reality, these models are mixed and matched, as you can have a freemium product that allows people to have a trial of the paid tier, but for clarity's sake, you can think of them independently.

The further you move down this list, the less risk that your customer feels but also the slower that you'll collect cash from sales.

From the company's perspective, you want to be as high up on this list as possible.

From the customer’s perspective, you want to be as low down this list as possible.

The goal of testing your sales flow is to see where you have to meet in the middle. You want to give away as little as you have to.

The Goal of a Trial

The job of a trial is to allow the user to experience the paid product for just long enough to see the value and no longer.

How long it will take to see this value depends on:

  • Who your customer is
  • How often the customer experiences the problem you solve
  • How much they trust your product & your company

If you’re building a database solution for enterprise customers, they might need 180 days to get up to speed and see the value.

However, if you’re building a running companion app for serious athletes, they might be able to see the value in 72 hours.

It all depends on the context of the user.

Additionally, as companies mature, they start to offer different trial experiences for different customer types.

The Key Decisions of a Trial

To set up a trial, there are 2 key decisions that you’ll have to make

What is the “cost” or friction that the customer needs to pay?

To make a trial effective, the customer has to “pay” with something or overcome some level of friction to get started.

In the consumer space, the best practice is currently to have them input a credit card to begin the trial.

Effectively all consumer-focused product that has a trial using this model (such as ​Headspace​, ​Peleton​, ​Spotify​, etc)

If you’re building a consumer product right now, don't overthink this part and just use a credit card trial upfront.

For business-focused (or B2B) products, the best practices are a lot less clear.

What makes B2B different is that the user might not end up being the buyer of the product due to how budgets/purchase approvals work.

Therefore, your user, who is interested in your product and wants to check it out, might not have the ability to put in a credit card.

Products like MongoDB have helped popularize the Product Led Growth movement, where they let a user get ​started for free​ with no credit card required.

This works because:

  • Setting up a database is a lot of work, so onboarding has built in friction
  • Actually buying the product for large companies is likely in the 5-6 figure range, so that’s beyond what most people will charge on a credit card.
  • The sales cycles can be long, so having users already on the product will help them move the sale process forward.

Miro combines ​all​ of these tactics but in a subtle way. They have a free tier, a direct purchase tier, and a credit card upfront tier.

They are (likely) only offering a credit card up front trial on their business plan as that's the plan they actually want you to buy.

How long should it be?

Building on the above advice, you’ll need to make this trial long enough so that the user experiences the value, but no longer.

The hardest part about managing a trial is trying to determine how long that period or usage is.

The best advice to early-stage companies is to default to being generous here. Start with a longer window and then test making it shorter.

As we talked about a ​few months back,​ most major companies start with more generous trial windows and pull them back with time:

- Netflix cut their streaming service trial from 30 days to 14 days, to 7 days, to no trial at all.

- The NY Times used to offer 10 free articles per month, then cut it to 5 articles, then cut it to 3 articles.

- Miro once upon a time gave away unlimited boards, then cut it to 10 boards, then to 3 boards.

The more that you give away initially, the more people will experience your product and the faster your brand will grow. However, this comes at the expense of your immediate revenue and conversion.

Additionally, the length of your trial can be used as both a tool against your competitors and to help mitigate other business constraints.

So What Do You Do With This Information?

If you are implementing a trial for the first time, I would start with a relatively basic model, you can always iterate on this later.

Additionally, I would watch out for the following “gotchas”

  • Longer trials slow down the payback cycle on paid ads - So if your company acquires primarily through paid media, how fast you can grow is partially dependent on how fast you collect money and re-deploy it to buy ads. The longer the trial, the more delay there will be.
  • The longer your trial, the fewer shots you have at improving it each year - Skillshare (back in the day) ran 90 90-day free trial, which apparently was the best at converting users, however really slows down product development. A trial this long means that they only had 3-4 tests they could run per year to improve it.
  • If you have a material cost to provide your product (such as streaming music rights, physical products, etc) your trial user (can) become a material expense. Remember to factor this in and don’t go broke :)

Trials can be a huge step forward for a company if implemented correctly and they are worth re-testing the model every few quarters to see what additional conversion you can find.

Other Articles: