Even though the term ‘Freemium’ got introduced into business lingo recently, the central idea behind it is not something new: give a product/service for free to the customer with the hope to influence her to buy another product. Gillette pioneered this idea in the modern industrialized world in late 19th century by giving razors for free but charging customers for disposable blades to recover the cost of razors. And, thus built a brand with a powerful moat that has survived for more than 100 years.
Nowadays, you can see numerous examples of the same business model all around where a paid product is cross-subsidizing the free product: get $700 worth of phone for free by signing a $60/month cellular contract for 2 years; get a Square Card Reader for free and pay a fixed %ge on each transaction.
This strategy of cross-subsidization worked well in digital world too when software was sold on CDs in physical stores. It works even now in the gaming industry where game publishers pay Microsoft and Sony a fixed percentage on each sale of game CD sold that runs on PS4 or Xbox.
But everything changed after the invention of internet. It flipped the assumptions around how software products are delivered and consumed as marginal costs — the costs incurred in selling an additional unit of product or service — came down significantly.
- Cost of Goods Sold is zero as software is just a piece of code that can be infinitely replicated.
- Distribution costs — costs of getting the product in the hands of the customer — also became almost zero as software can be delivered over the internet v/s buying CDs in stores.
- Transaction costs, the costs of executing and supporting a transaction, also became very efficient. You don’t need to go to a physical store; don’t need a sales person to show you around; don’t need a cashier to help with the transaction. All you need to do is enter your credit card details and you’re all set. You might never interact with any personnel from the company in your life.
As all of the these variable costs mentioned above have come down significantly, it has become even more tempting for software businesses to offer a free version of the product with the hope to acquire a significant number of customers that they can convert to paid later.
But it’s easier said than done. History of software products is littered with examples that couldn’t succeed at this approach; Companies that didn’t think deeply enough about basic principles and or even if they did, by the time they realized it was too late to make any course correction.
What are those principles/conditions that create the environment conducive for freemium to work? Without further ado, let’s dive in:
It’s free, but still Good
As i said earlier, the core objective of freemium in the software world is to acquire as many users as you can with the hope to convert them to paid users sometime later — you can afford to do so because of very low marginal costs. But users will come and stick around only if they can derive value out of it that they can’t get for free elsewhere.
Providing a sub-par experience in free version and hoping that customers will convert to paid, better version will do your business more harm than good as customers will shun your product at a faster rate than they came in. Not only will they leave but will also discourage other customers from trying out the product.
Therefore, no gimmicks; no sub-optimal product; no bad user experience, please. Customer must be able to derive sufficient value from the free product even if they never decide to upgrade.
Through it’s free version a user can send emails to upto 2,000 contacts or a maximum of 12,000 emails/month. Product works the same way for both free and paid versions. To add to that, since the launch of free version, it has consistently increased the value in the free product by increasing the number of contacts you can send email to. It started with 500 contacts, increased it to 1,000 to finally 2,000.
On the other hand, i hate Apple’s iCloud Storage. Not only is the overall product buggy and non-intuitive, but the free version is just awful. You get a paltry 5GB of storage for free. And that hasn’t changed ever since the launch in 2011, even though the number of images/videos you take on your phone and the quality/resolution of those has changed drastically.
iCloud Photo Stream is a half baked attempt at solving this problem as it keeps your recent photos/videos in the cloud for for 30 days before deleting. In comparison, Google’s new phone Google Pixel allows unlimited storage of photos and videos taken through the phone. If I ever switch to Android, iCloud would definitely be one of the biggest reasons.
Very large market with some degree of virality
According to Tomasz Tunguz, conversion rate for freemium products normally varies between 2 to 4%. A simple math exercise shows that to achieve $100 million in Annual Recurring Revenue (ARR), a company needs 1 million customers who pay you $100/year. Assuming an average conversion of 3%, company needs to have 33 million customers, both free and paid. That’s quite a lot actually and implies that to be successful:
- The product needs operating in a market that is huge.
- Company needs to figure out a way to acquire these customers very cheaply, as paid acquisition channels are not a option; business economics will not make sense at all. Therefore, product needs to have the right hooks/incentives in place to go viral and bring customers into the fold, either organically or through referrals.
Dropbox is one such rare company that everybody loves to talk about. It has 500 million total users and only 2% are paying customers. Dropbox has been able to achieve such a feat because:
Network Effects — Every time a user shared a file in Dropbox with somebody who didn’t have Dropbox, other person was inclined to try it out. And as more and more people got on the Dropbox and shared files, value of Dropbox increased for everybody.
Viral Referral Growth — Dropbox famously hacked its way to hundreds of millions of users by incentivizing customers with free storage for every new user they referred to Dropbox. With this simple hack users became their evangelists in droves.
Weebly, a do-it-yourself (DIY) website builder struggled to achieve the scale needed to become a sustainable freemium model and was finally sold to Square at a lower valuation than their last funding round.
Just to be clear, selling your startup for hundreds of millions of dollars ($365 million to be exact) is no small feat and is not a failure by any standard. I just want to highlight that getting millions of SMBs on board to use your product — even for free — is very difficult, if not impossible.
First, Weebly didn’t have right ingredients in place to go viral. Free websites prominently show ‘Powered by Weebly’ advertising somewhat implying that the business might be cheap. No customer looking for a professional web presence would like to be associated with such image. Plus, most of the competing website builders offered a similar free product. There wasn’t anything unique that would set them apart and gain virality. Second, it couldn’t outspend Wix or GoDaddy in brand dollars to acquire free customers. A graceful exit through acquisition by a large company made so much sense.
Costs to serve customers needs to be very low
So far, the company has created a valuable free product and has figured out a way to acquire customers cheaply. But that’s not enough. To become a sustainable business, the costs to serve/support customers should be very low. In the case of Dropbox, only 11 million out of the total 500 million are paying, yet the company is free cash flow positive.
Typeform, a survey/form builder tool, is another good example to illustrate this point. Like MailChimp their free version is limited by number of free surveys per month or total number of questions that you can send in a survey. A person with rudimentary understanding of the online world can create and start sending surveys without any personnel help, whatsoever. In addition, the cost to Typeform for sending/managing surveys (bandwidth and storage costs) is almost zero.
Whereas Spotify, a music streaming app, is the antithesis of this principle. At the time of recent filing of its IPO, it had 158 million monthly average users and 70 million paid subscribers, implying close to 90 million free users.
Spotify pays between $0.0006 to $0.0084 royalty fee to rights holders for each stream of music played. Assuming a free user plays on an average 5 songs/days each month, at a minimum, these free users would be costing Spotify ~$100 million each year. That’s a LOT of money to support free users. There are various industry dynamics responsible for such high costs. I’ll try to cover that in a separate blog post, sometime later.
Strong incentive to upgrade from free to paid
Even if a company swiped right on all the points covered so far, all the effort will go down the drain if there doesn’t exist a compelling reason for customers to upgrade to a paid offer. For that, not only you need to be judicious in deciding what features to keep in the free version, but also you need to be innovating to consistently improve the paid version.
Wix, a DIY website builder that lets you host your website for free, has 130 million registered and 3.5 million paid users. Free users get a domain with *.wixsite.com in the suffix, only 500MB storage & 1GB Bandwidth, no website analytics, and Wix branding on their website — strong enough constraints to push a person to upgrade if she wants a ‘Professional Looking’ online identity.
Trello, a company that made the Kanban board ubiquitous, struggled to convert free users to paid plan as there wasn’t enough value in the paid plan. All that was in the paid plan was the ability to customize the background of Kanban board board or use stickers/emojis or upload large size files. And not many wanted to pay $5/month extra just for these niceties when free product can get the same job done.
To make it even more difficult, their competitors copied the Kanban board format (like Instagram copied Stories format from Snapchat) stunting the growth of Trello. Result — what could have been a multi billion dollar company was bought by Atlassian for just north of $400 million.
Tinder, a dating app, launched Tinder Gold last year that had only one additional feature on top of existing paid product Tinder Plus. Tinder Gold lets a user see who had liked his/her profile without the need to swipe right on countless profiles to find a match. And it completely changed the trajectory of Tinder’s business adding a whooping 476,000 subscribers within just a quarter after the launch taking their total subscriber count from 2 million to 2.5 million.
Acquisition is just the start; Engagement is end game
Customer churn is inevitable part of any business. And free customers are any ways less incentivized than paid ones to stick around making churn even more difficult to manage.
Successful companies try to overcome this challenge by making onboarding super slick so that user can experience core use case of the product as soon as possible and over time help her derive more and more value out of the product making it difficult to switch or leave.
Famous ‘Smile Graph’ of Evernote, a note taking and sharing app, is an excellent illustration of this idea. As you can see from the graph below that the percentage of users returning to the app drops first as few users churn out, but in second year it starts trending upwards as more and more users make Evernote an integral part of their professional/personal workflows.
That’s broadly it! To summarize, I’d say that Internet has made it possible to enable freemium model at a scale that wasn’t possible before. But you need to have the right ingredients — a valuable free product, very large market with some degree of virality, low cost to serve free customers, strong reason for customers to upgrade, and retention of both free and paid users — in place to make it tick, otherwise you will end up writing the obituary of your business with your own hands.
Are you aware of any other products/companies that have interesting freemium models? Let me know in comments.