One of the most common questions in subscription products is, "Should I have a free product?"

If you look at the current subscription companies that are worth billions of dollars, some have big-free products (Spotify, NY Times, Trello, Slack, etc.), and others don't (Netflix, WSJ, Financial Times, BarkBox, Your Utility Company, etc.).

You can also point to much larger, more enterprise-grade products that didn't start with a free tier, such as MongoDB, but later added one.  

Other companies, such as Palantir, never offered a free service tier.

So, if you're a start-up founder or early-stage employee, how should you think about whether to have a free product?

If you're going to have a free tier, how much should you give away?

Initial Decision - Your Skills & Your Market

To give a bit of a background, most consumer subscription companies end up building a free product to help fuel their user acquisition.

The traditional advice for investing in a free product is to "think of it like a marketing expense," which is sound advice.  However, remember that it is still an expense.

For most early-stage companies, the question is not "Should I build a free product?" but "How should I acquire users?". Free products are one of the ways.

You should pick the one that is not only the most effective but also matches the skills of the team that you currently have.

Early-stage companies are unlikely to be able to afford multiple acquisition methods at once, and this should be based more on what is easy for you to do.

If your team is:

• Mostly product development people - e.g., engineers, a PM, a strong designer - then building a free product is achievable.
• Mostly go-to-market people - such as product marketers, sales folks, and paid media people - then ads, direct outreach, and partnerships are going to be easier.

The smaller your team and the earlier your company, the more risk there will be in trying to bring in agencies or employees in a skillset the team doesn't fully understand.

Second Decision - Really Free or a Trial in Disguise?

The second decision is whether your product's free tier will be a full-fledged product line or a trial in disguise.

I define a "full product" as something a user can return and keep using effectively forever.

It doesn't mean that it has all the features they might want, but nothing prevents them from using it for a long time.

Spotify Free is a good example of something you can actually use forever. The ads get annoying, but nothing really prevents you from staying in that product.

A "Trial in Disguise" is a free tier that can't be used forever because of usage limits or similar restrictions.

• ​DocuSign​ gives you up to 5 docs that you can sign "for free," but when that runs out, you have to pay.
• ​MongoDB​ allows you to create a free database of unlimited size, but all the data is public.  Private data is a pretty hard requirement for app development, so you realistically have upgrade to use the product.

What has to be true for a free product to work?

For a free product to work in the long term, many (but not all) of the following things must be true.

1. You can keep a high-quality bar

Even if your product doesn't cost money, it still costs other things (like time, friction to get started, etc.).  Your product must be good and stay good for a long time to work as an acquisition tool.

When I was at Codecademy, we had millions of people sign up for our free product every year.  They only did this because it was better than programming at teaching you programming than the textbooks you could buy.

2. You have a large potential audience of users

To give you a sense of the math behind free to paid conversion in freemium products, good is about 3%-5%. Great is 5-7%.

For a free product to work, you need to be in a space with enough potential customers that you can grow a big business with 5

If you have a cloud storage company, such as Dropbox, and the whole internet-connected world could be a customer, you can build a giant company by attracting millions of people and converting thousands of people into paying customers.

If you are building a product in a niche space, as such as ​GeoScout​ does with making oil and gas exploration tools, there are probably only a few thousand potential customers.  They definitely can't build a material business by only attracting 20-30 customers.

3. You are giving something away for free that others charge for

For your free product to serve as an acquisition machine in the long term,  it should be better than what others are charging for.  The better it is, the more users you will acquire.

It's great and free; people will talk about it, and that's the best acquisition source.

4. Each free user knows a lot of other potential free users

You obviously can't spend a lot of money to attract more free users to a product, so it has to be driven primarily by word of mouth.

For word of mouth to work, your users must know other potential users who might be interested.

The more potential users that each of your users can talk to, the more word of mouth will spread.

Increasing the potential size of people who can talk about a product is one of the main benefits of a free product.

Consider two companies with the same number of paying users

• Company A has 100 paying customers
• Company B has 1,000 Free Customers and 100 paying customers

Company B, while spending more money to support a free tier, has 10x the word-of-mouth potential (assuming their product is good)

5. Zero (or Near Zero marginal cost of adding more free users

The final point is that if you aren't making any money from your free users, you can't spend much money to support them.

The free product isn't going to be free to build or maintain, but the incremental cost of adding one more free user should be very inexpensive, if not free.

There are notable exceptions, specifically companies that have to pay licensing fees for content (such as Spotify).  Most companies that fall into this tier end up monetizing the free audience with ads to get it to break even.

So What Should You Do With This Information?

While it's hard to predict when a free product will work, it is easier to see when one clearly won't work.

I wouldn't say you have to pass all of the above criteria, but I would be very skeptical if you pass less than 50%.

‍