Deconstructing the Fitbit IPO and S-1
Fitbit is the latest iconic digital health company planning to IPO, with a drastically different operating history than another iconic company from last year’s group. Quote from Health Innovation Summit (2013).
Beginning early last year, whispers of Fitbit’s revenues started to circulate. Testing the numbers with investors, responses were mixed—some thought the figures were outlandish, and others confirmed them as absolutely true or undersold. This is the story of wearables, a polarizing category that has been misunderstood, underestimated, and perhaps wrongly maligned. And this narrative will shadow Fitbit’s public offering—with one journalist already calling it a “sucker IPO” and another praising the company’s “surprising profits.” Fitbit’s S-1 provides answers to questions many of us have debated for years, and leaves a number of others completely unanswered.
Objectively, Fitbit represents a tremendous growth story. In less than six years from the launch of its first device—and on the back of less than $70M in venture financing—the company has built a billion dollar plus revenue (run rate) business (at 50% gross margin), legitimately pioneering the category of “connected health and fitness” trackers. Regardless of any S-1 analysis, IPO price, or post-IPO price change, what James Park and his team have done at Fitbit should not be shortchanged. It is an incredible startup tale from a determined, visionary founder—a founder that no one wanted to fund. And yet here they are, on the cusp of an IPO.
Since launching its first device in late 2009 (a year later than promised), Fitbit has experienced explosive revenue growth, with a 246% CAGR over the 2010-2014 time period, culminating in nearly $750M in sales in 2014. Using the Q1 2015 actual results, demonstrated growth in the average selling price (ASP), and historical holiday quarter sales, we forecasted Fitbit’s full year 2015 revenue between $1.6-2.0B.
Growth in the last year—and likely extending into 2015—has been driven not solely by the increase in devices sold, but also by improvement in the ASP of Fitbit’s devices. ASP growth is likely through the combined effects of both a better unit mix towards Fitbit’s higher-priced offerings (its first non-recalled device sold above $99 MSRP debuted in Q4 2014) and a positive shift towards higher margin distribution channels. (Note: the calculated ASP is overstated by the average subscription revenue per device sold and the average accesories revenue per device sold, both of which have been assumed as negligible.)
In a surprise to some observers, the company’s growth has been profitable, with operating income of $158M in 2014. Gross margins (controlling for the quarter where the majority of the Fitbit Force recall effects were felt) averaged above 48% for the last eight quarters and operating margins averaged above 21% over the same time period. These are impressive figures that help explain how the company has avoided taking on additional venture capital, financing its operating activities from its own cash flows.
Users, engagement, and retention
In the S-1, Fitbit provides limited data on engagement. It’s unclear from the filing how much users actually engage with the Fitbit software platform based upon traditional mobile metrics (daily active users and monthly active users). Presently, this is the largest criticism of wearables—that the devices are abandoned after six months by the majority of people. Instead, Fitbit presents its own metric: Paid Active Asers (PAUs). A PAU has been defined as a user that in the last three months has done any of the following: had an active Fitbit Premium or FitStar subscription; paired a tracker or scale to their Fitbit account; logged 100 steps; or logged a weight measurement. A simple metric of PAUs versus registered users (RUs) was used to test the level of retention (higher is better) on the Fitbit platform—PAU/RU was 46% in 2014, and grew to 50% in the most recent quarter (2015 Q1)—this is likely a recency effect of measuring in a single quarter versus an entire year.
But this doesn’t seem like a complete picture of engagement. Based on this information alone, it can only be guessed that there are more than 10M inactive Fitbit devices, representing more than 50% of its units ever sold, or stated differently, more than the total number of devices sold from 2009-2013 combined. Based on the S-1 alone, it cannot be definitively stated that the great majority of Fitbit devices sold before 2014 are no longer used, but any other conclusion seems improbable.
Another way to look at engagement on the Fitbit platform is by looking at the churn or negative PAUs. This metric can be constructed by subtracting the change in PAUs over a given time period from the total devices sold in the same period (this assumes those device buyers become active and are the primary driver of new PAUs, versus reactivated users or new subscriptions).
Using negative PAUs or churn provides a better picture on Fitbit’s retention of customers on its platform, although it’s hardly comprehensive. Even if for 2014, an analysis attributed the 6.8M churned PAUs only to PAUs at the end of 2013 (a total of 2.6M), there would still be 4.2M churned users to account for—PAUs that could only originate from the first three quarters of 2014 (by definition, PAUs from the fourth quarter cannot churn before the end of 2014 since PAUs are measured in three month windows). This implies that more than 70% of Fitbit purchasers from the first three quarters of 2014 churned before the end of the year (using a revenue-based allocation of 2014 devices sold in Q2/Q3 2014) Note: The previous sentence was edited following the original posting to clarify the specific assumption window.
How much does this matter for a company that makes all of its money from the gross margins on hardware versus subscription revenues—after all, aren’t many iPhones collecting dust somewhere, with customers rushing out to buy the latest device every two years? For Fitbit, the answer is likely not the same. Considering its relatively minuscule sales two years ago, in 2012, it’s hard to assess whether customers are upgrading to newer devices as they abandon old devices. However, the average devices per customer can be estimated using a metric of cumulative devices sold versus total registered users; this number has been steadily declining as device volumes increase, approaching one (1.09) through the first quarter of 2015. This will be a metric to watch as the company moves out of a hyper-growth period.
A lack of long-term engagement and repeat purchasing has not mattered in the face of a growing market. Compress the time periods analyzed down to a single quarter, and it’s clear—at least currently—that there are more customers joining than leaving. In the short term, Fitbit is well-positioned to take advantage of a trend it helped create, selling millions of devices at 50% gross margins along the way. But these engagement and retention figures are easily the most alarming numbers in the Fitbit S-1. They reveal a device and software platform that has a serious problem with retention, and a limited ability to engage customers over the long-term. In healthcare, these questions matter—wellness is a function of eating well, being physically active, and getting a good night’s sleep more days than not, over a long period of time. Fitbit’s not there, yet, in terms of demonstrating its platform can deliver on improved health, but no one else is either.
Overcoming this challenge is what will separate Fitbit from being a good company that was able to ride a wave successfully, or a great company that creates an iconic consumer health brand. The question for investors is how long the market will continue to grow at this rate, and whether Fitbit can execute on growing engagement before it approaches peak wearable, i.e., the number of devices sold per year reaches saturation.
Distribution and marketing
Fitbit significantly ramped its efforts in both distribution and marketing in 2014, helping to drive the company’s 175% growth in 2014. Growing (or perhaps establishing—it’s unclear from the S-1) the relationship with Wynit Distribution was the most significant change in the year; by selling Fitbits everywhere from Costco to Sports Authority to the Army and Air Force Exchange Services, Wynit moved from <10% of revenue as late as Q1 2014 to 13% by the end of the year, and became Fitbit’s largest single distributor. Best Buy and Amazon were noted as the next two largest distributors, with the overall top five responsible for around half of the company’s revenues, helping push its products to over 45,000 retail stores across fifty countries.
Much has been written about the growing use of activity trackers within corporate wellness programs. And while Fitbit notes penetrating corporate wellness as one of five growth levers for the business, the S-1 provides no insight into its success in the market beyond a reference to its first 1,000 unit corporate wellness sale, which came all the way back in October 2010 (who was this pioneering company?). It is unclear how many units are being sold through this channel, but the relative depth of the offering (employee onboarding, leaderboards, reporting, analytics) suggests how seriously Fitbit takes competing in this market.
As anyone who has flown Virgin America recently or watched the NFL Playoffs noticed, Fitbit launched its first national advertising campaign in 2014. Advertising spend reached $71M, and was the primary reason overall sales and marketing expenses grew by more than 300% in 2014 versus 2013. As the company itself notes, it has dominant market share in the U.S., having established itself as the leading brand in the category. Based on its advertising spend alone, the company knows it needs to push hard on consumer awareness to not only maintain, but grow its position as the dominant activity tracker.
For companies entering the wearables space, there are important lessons in Fitbit’s growth story. First, navigating distribution channels and finding the right partners is absolutely the key to fast growth. And advertising will play an ever-increasing role in building a brand in this category. This is becoming a pure play consumer electronics space, and (expensive) marketing has traditionally been table stakes to compete successfully.
R&D and growth
In explaining the year over year change in first quarter revenues (2015 versus 2014), Fitbit notes that $233M of the first quarter’s revenue was from new products released in the fourth quarter of 2014, or nearly 70% of Q1 2015’s revenue. This is significant, since Fitbit did not discontinue older products, including the Zip, One, or Flex, and all of the new products launched at higher price points ($129+) than existing products. It speaks to the need for a steady stream of new products, particularly for the holidays (the fourth quarter has represented 44%, on average, of annual revenues the past three years). At this point in its lifecycle, another Force episode (the Fitbit Force was released in October 2013 and recalled in February 2014 due to skin irritation issues) would be a disaster for the company, and not just from the standpoint of its net income.
Fitbit itself notes new products, alongside additional features and services, as its first two growth levers, also identifying brand awareness, global distribution, and corporate wellness as the other primary growth opportunities. What types of biosensing capabilities might Fitbit be working on for future devices? There is not even a hint about new products in the S-1, even with the company spending about 7% of revenues on R&D (versus 10% on advertising alone). Because the hardware in this category can so quickly be commoditized by competitors, it will be important for Fitbit to identify relevant physiological parameters to track for their core markets (everyday, active, and performance in the company’s parlance, respectively targeting general wellness, physical activity, and sports) and successively launch new devices incorporating these sensing features.
Today, the company’s devices are able to track steps (accelerometer), distance traveled (GPS), heart rate (optical), floors climbed (pressure), sleep duration and quality (accelerometer), weight (scale), and body fat (scale/conductivity). It is important to remember that the company is comfortable with form factors outside of wrist-worn devices—all of its trackers before the Flex in mid-2013 were clipped on or placed in a pocket—and it also offers a scale. Moves like the hiring of Aaron Rowe shouldn’t be overlooked, given his background at both Scanadu and Integrated Plasmonics. Will we ever see a clinical device from Fitbit? No clue is given in the S-1.
The company has noted many potential risks to growth. Beyond the engagement and retention issues previously discussed, the biggest risk to Fitbit is clearly its competition (the S-1 cites Garmin, Jawbone, Misfit, Under Armour, adidas, Apple, Google, LG, Microsoft and Samsung as competitors). Fitbit is being attacked on multiple fronts—from the low-end to the high-end, and even software-based tracking. It can be argued that Fitbit has been extraordinarily successful to date in a highly competitive market, beating not only venture-backed companies such as Jawbone, but also fending off Nike prior to running a single major advertising campaign. But Jawbone and Nike aren’t Apple. The Apple Watch wholly encompasses most of what the family of Fitbit devices can do across its core segments, with a wide range of additional, non-health related functionality. The opportunity for Apple Watch in activity tracking is not only to pull from Fitbit’s existing customer base (which will be available to switch, given churn rates), but also the vast majority who could never be bothered to wear a single-purpose device. Fitbit, in turn, can work to incorporate the fundamentals of smart watches (the recently released Fitbit Charge has caller ID, for example).
As always, the initial filing leaves out the expected pricing for the IPO, creating a small window of speculation. The most appropriate comp would seem to be GoPro, the wearable, connected camera manufacturer. While GoPro boasted higher full-year revenues ($985M) at the time of its IPO, the preferred metric seems to be adjusted EBITDA, and Fitbit’s $191M is significantly higher than the $134M from GoPro. Ultimately, GoPro debuted around a $3B market capitalization and quickly doubled to more than $6B. It is easy to assume that in the current equity market, $3B will serve as the floor for the Fitbit IPO, with some valuation metrics suggesting $5B (2015 revenue multiple) or even $8B (price/earnings).
Bonuses from the S-1
Fitbit was first incorporated as Healthy Metrics Research, Inc., in March 2007
FitStar was acquired for $11.5M in cash, $13.6M in stock (valued at $19.25/share), and $7.7M in contingent considerations; in addition, $5.4M in cash and stock was allocated for retention bonuses
Flextronics is the sole manufacturer of Fitbit devices
One of Fitbit’s products is regulated by the FDA: the Aria (read the 510(k))
More than 20% of Fitbit’s revenues are generated internationally
Get our weekly curated take on everything happening in digital health. Sign up for the Rock Weekly.