Why Free Product + Paid Media = Death for Subscription Companies

The most common business plan that I see for subscription products is:

  1. Figure out the product
  2. Add a free tier
  3. Raise a big round to “scale marketing.”
  4. Make a lot of money

The big problem with this is that many companies die between steps 3 & 4 and struggle to understand why.

This plan makes sense in theory. We will build the product, attract more people with a free tier, and then attract even more people with paid media. Then we’ll have a giant business.

The problem, as I see it, the problem is that having a true free tier of a product and making paid media a dominant channel are incompatible with a subscription business.

Both paths are hard, and its very, very unlikely that you can make both of them work until you’re a giant company.

What Makes Free Products Hard

When I say “Free product” here, I am talking about a product that has a true recurring use case that someone can use for free. Think of the Spotify or Duolingo free products, where there is no limit on usage.

There is another type of “free” tier that has become popular that is really a trial in disguise. Docusign (contract signing) product has a “free” tier that allows you to send five documents for signature. This is not a “true” free product, so let’s hold this aside for now.

The common advice for free tiers of a product is to “think about it like a marketing expense,” which is technically true; however, this means it still has a cost.

Free products are not free for you to build or maintain. Also, freemium products typically convert much worse than trial or direct purchase-based products, costing you money upfront.

The reason that you build them is that they can grow the brand much faster as more people experience the product and allow you to (hopefully) displace incumbents who are charging for the same thing.

They hopefully lead you to the promised land of having a great brand name that allows you to acquire customers cheaper than everyone else.

That said, you are giving up money now for a better brand in the future. This makes them capital-intensive.

This is the path that Codecademy took; we gave away a really high-quality free product that led to a great brand and strong top-of-funnel traffic. We also had to raise ​$82 million dollars​ to do it.

For a free product to work, you need the following things to be true:

  1. You have a high-quality bar - Having a bad free product hurts you more than having no free product at all. This gets harder with time.
  2. There is a large potential audience for the product - The best freemium products have conversion rates of 5-7%, so you need a big enough audience to convert enough people and build a sustainable business.
  3. You are giving something away for free that others charge for - and its ideally better
  4. Each free user helps you spread word of mouth - For this to work, your users need to know other users.
  5. Supporting additional free users is nearly free - In business speak, the marginal cost of 1 more free user needs to be very low.

This means for the vast majority of companies, this is not a good idea.

Freemium works for broad appeal B2B or B2C consumer products with never-ending use cases (like communication, news, or entertainment) and doesn’t work well for products with small audiences or niche use cases.

​Equals​, which is a more robust spreadsheet product that allows you to integrate with data sources, wrote a great ​blog post​ this week about why they eliminated their free tier.

I am not that familiar with the Equals product (though it looks awesome); they pass most of these checks with the notable exception of number 3. Google Sheets is a high-quality free competitor, making it hard for them to get the full benefit of free traffic.

He shared today a ​post​ from other founders who were also burned by a free tier of their product.


Additionally, adding a free product raises the challenge of how you set up your paywall between free and paid, which is by far the hardest challenge in freemium products.

[Here is a ​deeper dive​ from a few weeks back if you're curious on that topic]

What Makes Paid Media Hard

Paid media is not conceptually hard to understand:

  1. Figure out the Lifetime Value (LTV) of your users
  2. Test Ad campaigns on a small scale until you can find a customer acquisition (CAC) cost that is 1/3rd of your LTV
  3. Scale those up and stay within the 3:1 LTV to CAC ratio

After you have one channel working, go find others and repeat.

As I ​wrote about a few weeks back​, the issue with subscription products is that LTV is really hard to calculate while the company is in a growth phase.

Because you are collecting money from your users over a longer period of time than most other businesses, it takes a long time to determine the impact of your changes on LTV.

You might be running a pricing test or attracting a new user persona for your product, which is a great idea; however, if you are retaining a user for ten months, you won’t know what this did to your LTV for ten months, which is an eternity.

You’re going to end up in one of two places:

  • The best case scenario is that you are underspending on acquisition to acquire users, which means that you are growing slower than you could
  • A much worse scenario is that you are overspending on these users, and you don’t realize you’re losing money for ten months.

The ironic part of this is that the better you are at retaining users, the worse this problem can be.

Why You Really Shouldn’t Do Both

Beyond the obvious advice that “doing two hard things is harder than doing one hard thing,” there are several other reasons.

1. Both of these are very capital intensive

Free products require more engineers, designers, product managers, and infrastructure. Setting it up is only the first part; you then also have to factor in the maintenance and the constant work to test exactly where your paywall should be.

Paid media teams need resources from marketing and creative departments and a lot of money to iterate on ads until they find a winner. Then, you need to float the cost of acquiring users until you have made back your CAC from the LTV.

Which, as noted above, you might have might be wrong about.

2. Both of these are very labor-intensive

Being great at your primary acquisition channel is what you should be shooting for early on.

Multi kinda effective acquisition channels do not equal a single dominant channel.

Regardless of your chosen path, you’ll need to staff a team focused on that channel, which means they need to be recruited, managed, compensated, etc.

Additionally, the company needs to stay committed to that path until you’re really good at it, which means attrition on that team might send you back to step 0.

3. Free Products Both Lower and Delay Conversion which Hurts Ads

Arguably, the biggest driver on the ROI of your ad spend is how well the website you’re sending people to converts to a sale.

Because the Freemium model converts at a worse rate than a trial or direct sale model, your ads aren’t going to be as effective as competitors without a free product.

This means they can both outspend you in those channels and that you’ll have a worse return on ad spend, which also likely means you can’t attract the same level of paid media talent to your company.

Additionally, what we saw at Codecademy was that:

  1. After seeing an ad, a portion of people just google the company, see the free option, and start there.
  2. It was very hard to determine how long someone would stay “free” before converting to paid (if at all)

Point #2 means that your ad payback period is going to longer, which means you’re spending capital less effectively and you need to raise more money to

So What Should You Do with This Information?

Be thoughtful in how you pick this path, then stay with it until you are great at it.

Once you think you can do both, wait another year or so.

If you start running ads, only sell your annual plan to help speed up the payback period.

Other Articles: