Startup Lessons: From Free Product to Charging a Fee

Launching a fee for a service that has been free for users is scary. Here’s what I learned.

It’s a fairly common scenario among startups: launch a free product to test and iterate, and attract as many people as you can, then start charging. It’s never easy to take the next step and charge customers, even if a fee model was always the plan. One can’t avoid the risk that customers will leave. We know companies eventually need to make money, but we all love a free lunch.

Have you ever been the user of a service that started charging you, seemingly out of the blue? No one likes change, especially when it means you’ll pay for something you used to have for free. It’s even more of a challenge if you’ve used the service for free for multiple years.

That was the case for users of my client this year.

I had the pleasure of being part of a small task force attempting to predict the results and consequences of my client launching a service fee for the first time. The lessons were far more than these, but I still haven’t managed to articulate them all.

Here’s what I’ve learned so far from the process of charging a fee for a previously free product:

Launch the fee sooner, don’t wait.

• The users who had the service for longest were most irritated about the fee. Their messages were angry, insulted even, and said they’d never recommend the company again. Those ones did in fact churn, and likely won’t be back.
• Some of these users had used the product for three full years. That’s a long time to get used to not paying.
• The least inflammatory comments and customer support messages came from users who had started with the service more recently, even as much as a year earlier.
• In comparing the responses of newer (less annoyed) and older (extremely annoyed) users, I believe that we could have softened the blow if the fee were introduced sooner in the product’s lifetime. Waiting so long only made it more difficult for many users to adjust to paying for something they associated so strongly with ‘free.’

Nail the user types before adding a fee.

• This sounds elementary. But my clients don’t have a really solid sense of their customer types yet, and I don’t think they’re alone. It’s really tricky to launch a fee when we don’t fully understand which customers give the company disproportionate value, and which ones we can actually let go without much harm.
• When the fee was launched, our lack of knowledge about the user types meant we had to play catch-up, fast: I dug through Facebook comments, customer support messages and user data trying to piece together profiles of these users on the fly. We needed to understand who was churning, and how much it should hurt.
• We got lucky: many users remained loyal, and even publicly defended the company’s need to earn revenue to pay its hard-working employees. But some of those longest term users did leave. They were possibly the ones worth two or three or ten times as much because they recommended the service to many others. This is still something we can’t fully quantify, and measure the impact of.

Use qualitative input to prep the Customer Success team.

• We didn’t do any qualitative work in our risk assessment study. We spoke to no users ourselves to gauge emotional responses to a fee, to understand how this would impact them, or collect questions they might ask of us.
• Our quant focus helped us estimate how many customers we risked losing, and whether we could acquire new users to cover the loss. But it didn’t help us prepare for existing users’ questions and confrontations regarding the fee.
• This left the customer success team and management scrambling to respond to users one at a time, with no ability to anticipate the topics that arose. It wasn’t efficient or effective, and wasn’t their fault. Our task team should have helped them prepare by using qualitative work to better anticipate users’ questions.

Prepare for hate mail, and prepare a human response.

• I knew there would be people angered by the introduction of a fee on the service they’d used for three years without paying for. But I wasn’t expecting so many to seem personally insulted. On one hand, it showed how strongly connected users felt to this brand and its values. On the other, people were quick to change sides when they felt the company no longer had their best interest in mind. There were a lot of strong words, and situations bordering on harassment by users. Management was shocked.
• We didn’t speak with users about the fee ahead of time because the team was afraid of letting the news out too early. Internally, many imagined a ‘leak’ from a tester would drive hate mail before we even launched a fee.
• In the end, the strong reaction they feared happened anyway. We would have benefited much more from daring to interview select users ahead of time, to understand their feelings, and have more time to prepare a considerate way to respond.

Add something, anything, so the user sees added value.

• One question from existing users came up a lot in those early days: “What are you adding for the amount I’m paying every month now?”
• It’s tough to argue that anyone should start paying for the exact thing they’ve had for free.
• Interestingly, some users even shared that they tried to convince themselves to stick with the service: they believed in the company’s values, and the service had been a big help to them. But paying for what had been free made them feel too much like “fools,” they said.
• With hints of upcoming product improvements in some comments, the mood lightened and many users became more positive. Compared to the idea of paying for zero product enhancements, the potential for product upgrades seemed to change the minds of some. No one wants to feel like a fool for paying for the same “free” product. Giving them something to pay for seemed to help them convince themselves.

Expect users to change behaviour within your service

• If charging users depends on any specific usage (using a specific tool, doing a task more than 2x per day), you will see drop-off, and usage patterns will change as users find workarounds.
• Some of the things we thought we knew concretely about usage patterns have shifted in a noteworthy way, potentially impacting anything from revenue estimates to how we will design in the future.

The lessons from such a big shift are many, though these ones rose to the forefront as I tried to communicate to teammates what we could have done better. The overall launch of the service fee was considered a success, but there’s always room for improvement.

From a user research perspective, our study could have done more to thoroughly understand where our users were coming from. In this case, business targets and timelines took over, and we didn’t do things quite as I or my teammates initially proposed.

Next time, I hope I can heed my own advice. Hopefully these lessons can help anyone — on the research side or business side — on the journey to charging for your service, and better serving your customers at the same time.