May 12, 2017
VIA EDGAR
United States Securities and Exchange Commission
Division of Corporation Finance
100 F Street, NE
Washington, DC 20549
Attention: | Barbara C. Jacobs, Assistant Director |
Ivan Griswold, Staff Attorney | |
Stephen Krikorian, Accounting Branch Chief | |
Morgan Youngwood, Staff Accountant | |
Re: | Fitbit, Inc. |
Form 10-K for the Fiscal Year Ended December 31, 2016 | |
Filed March 1, 2017 | |
File No. 001-037444 |
Ladies and Gentlemen:
We are submitting this letter on behalf of Fitbit, Inc. (the “Company”) in response to comments from the staff (the “Staff”) of the Securities and Exchange Commission (the “Commission”) received by electronic mail on April 19, 2017 that relate to the Company’s Form 10-K for the fiscal year ended December 31, 2016 (File No. 001-37444) filed with the Commission on March 1, 2017 (the “Form 10-K”). The numbered paragraphs below correspond to the numbered comments in the Staff’s letter and the Staff’s comments are presented in bold italics.
Form 10-K for the Fiscal Year ended December 31, 2016
Non-GAAP free cash flows, page 38
1. | Please revise to include all three major categories of the statements of cash flows with your presentations of the non-GAAP liquidity measure of free cash flow. In addition, explain why you believe free cash flow is both a liquidity and profitability measure. We refer you to Questions 102.6 and 102.7 of the updated Non-GAAP Compliance and Disclosure Interpretations issued on May 17, 2016. |
The Company acknowledges the Staff’s comment and advises the Staff that it began including all three major categories of the statements of cash flows with the presentation of free cash flow in its earnings release and Form 10-Q for the quarter ended April 1, 2017 and that the
United States Securities and Exchange Commission
Division of Corporation Finance
May 12, 2017
Page 2
Company will continue to do so in future earnings releases and periodic filings. In addition, the Company has revised its disclosure in its Form 10-Q for the quarter ended April 1, 2017 to specify that the Company believes non-GAAP free cash flow is solely a liquidity measure and will continue to do so in future periodic filings.
Item 7. Management’s Discussion and Analysis of Financial Condition and Results of Operations
Key Business Metrics
Active Users, page 40
2. | Tell us whether the Active User metric includes devices that were acquired directly from prior users. To the extent that this measure includes users that have activated used devices, please clarify your disclosure. Also, clarify whether the prior owner and new owner of a single device could be counted as separate active users during the measured period and discuss any impact on the metric. |
The Company advises the Staff that as further described in the response to Comment 3 below, the active user metric can include users who acquired a device directly from another user. It is possible that the prior owner and new owner of a single device could each be counted as unique active users during a particular measurement period. This could happen, for example, when an existing user purchases a new Fitbit device and gives her old device to a friend, assuming both individuals meet the active user definition during the particular measurement period. The Company’s active user definition (the “AUD”) requires that an individual be a registered Fitbit user who, within the three months prior to the date of measurement, has (a) an active Fitbit Premium or FitStar subscription, (b) paired a health and fitness tracker or Aria scale with his or her Fitbit account, or (c) logged at least 100 steps with a health and fitness tracker or a weight measurement using an Aria scale.
However, the prior owner and new owner of a single device will not necessarily each be counted as active users in future measurement periods. If, in the previous example, either the existing user or the friend ceases to meet the AUD in a future measurement period, she would not be included in the reported active user count for that period. The existing user, for example, may cease using any Fitbit device or subscription service in a future measurement period while the friend continues to use the device she acquired from the existing user. In this case, the existing user would not count as an active user but the friend would count.
While, as described above, it is possible to have multiple active users associated with a single device at different points in time, the Company believes that the number of active users associated with returned devices, devices directly acquired from another user, or devices re-paired to the same Fitbit account for other reasons, has been approximately 10% or less of the total
United States Securities and Exchange Commission
Division of Corporation Finance
May 12, 2017
Page 3
number of the Company’s active users in any particular historical measurement period. As described further in the response to Comment 3 below, the Company does not believe that it is possible for multiple active users to be concurrently attributed to a single device except in a few unique situations that would not have a material impact on this metric.
The Company further advises the Staff that the active user metric is not intended to be an indicator of device sales in any period, as the Company reports device sales as a separate metric. The Company does not currently believe that the active user metric has a direct effect on its revenue and operating results since substantially all of the Company’s revenue to date has been derived from sales of its connected health and fitness devices. Instead, the active user metric is intended to show the potential size and growth of the Company’s engaged user community, as further discussed below. The Company will note in future filings that the active user metric includes users who acquired a device directly from another user.
3. | Describe whether returned devices are excluded from the Active User metric, and whether in the case of returned devices that are resold, both the new and prior owners are counted as active users. |
The Company advises the Staff that the active user metric is not intended to be a reflection of devices sold by the Company but rather a reflection of the potential size and growth of the Company’s engaged user community. The active user metric is not associated with the particular device(s) owned by a user. Instead, the active user metric is based on activity associated with each Fitbit user account. It is possible in the case of returned devices for new and prior owners to be counted as active users if they meet the AUD. At the same time, a user with multiple devices is counted as only one active user if she meets the AUD regardless of the number of devices associated with her account.
Accordingly, in a particular measurement period, a user who buys but then returns a device may be counted as an active user if that user meets the AUD during that measurement period (e.g., pairs the device to her account). If that returned device is refurbished in that same measurement period, the new owner would also be counted as an active user in that period if she meets the AUD. Therefore, it is possible that in the case of returned devices that are refurbished during the same measurement period, that both the new and prior owners are counted as active users during the same period. However, because of the time it takes to refurbish a device once it has been returned, the Company does not believe that this sort of activity has historically had a material effect on the Company’s active user metric.
The Company believes it is appropriate to include both new and prior owners of a particular device in the active user metric because the metric is intended to provide a snapshot of the potential size and growth of the Company’s engaged user community during any three-month period. Since both the new and prior owners meet the AUD, the Company believes both users would be appropriately included in the active user metric as both users independently have demonstrated
United States Securities and Exchange Commission
Division of Corporation Finance
May 12, 2017
Page 4
a level of engagement with the Company’s devices and platform. The active user metric is not a measure of the levels of continuous engagement of individual users over longer periods of time.
The Company does not believe that the active user metric currently has a direct effect on its revenue and operating results since substantially all of the Company’s revenue to date has been derived from sales of its connected health and fitness devices. However, the Company believes that a large active user population can be indicative of future opportunities for monetization through the additional development of services and content.
The Company will clarify its disclosure in future filings to note that users who return a device may be included in the active user metric during a particular measurement period.
4. | Tell us your consideration of disclosing known material trends related to the Active User metric, including whether sales of devices during each fiscal year presented impacted growth in the metric. Also, explain the reason for differences in the growth rate between the number of registered device users and active users in fiscal 2016. Refer to Section III.B. of SEC Release 33-8350. |
The Company respectfully acknowledges the Staff’s comments and advises that it will include a key trends section related to the active user metric in future filings. The below reflects a proposed example of disclosure with regards to key trends:
• | As a percentage of the active user metric, users that were first time registered device users has declined in recent periods, from 38.4% as of December 31, 2015 to 26.1% as of December 31, 2016. |
• | As a percentage of the active user metric, users who logged at least 100 steps with a health and fitness tracker or a weight measurement using an Aria scale but registered in previous quarters increased from 61.4% as of December 31, 2015 to 73.4% as of December 31, 2016. |
The Company advises the Staff that the growth rate for registered device users on a year-over-year basis from 2014 to 2015 and 2015 to 2016 was 162% and 73%, respectively, while the growth rate for active users as of the year ended December 31, 2015 and December 31, 2016 was 152% and 37%, respectively. The growth rates for these two metrics will differ because they each measure different aspects of the Company’s platform. The registered device user metric reflects the overall number of devices in the market with users who registered a Fitbit account. A user is only counted as a registered device user once, upon establishing a Fitbit user account (provided that such user does not use multiple emails to register more than one Fitbit account). Further, the registered device metric is presented on a cumulative basis, meaning that the number of registered device users includes account users who have become inactive.
In contrast, the active user metric reflects the potential size and growth of the Company’s engaged user community during a particular measurement period. A user is only counted as an
United States Securities and Exchange Commission
Division of Corporation Finance
May 12, 2017
Page 5
active user in any period in which such user meets the AUD, but that same user will remain a registered device user regardless of activity. For example, if a user initially purchased a Fitbit device and established a Fitbit account in December 2015, the user would be counted both as a registered device user and active user for 2015. However, if that user does not meet the AUD in the fourth quarter of 2016, that user would still be counted as a registered device user, but not as an active user for 2016.
As such, the growth rates for registered device users will not necessarily correspond to the growth rate in active users. Even if the Company did not have any new registered device users in a given period, the active user metric could still grow during the period if previously inactive users became active again.
Liquidity and Capital Resources
Cash Flows from Operating Activities, page 50
5. | The increase from 56 to 85 days in your DSO during the last fiscal year is attributed to slower collections in the four quarter of 2016 as compared to the fourth quarter of 2015. Disclose the factors that led to slower collections during this period. |
The Company advises the Staff that its days sales outstanding (“DSO”) fluctuates between periods and is dependent on a variety of factors including linearity of sales, customer and geographic mix, buying patterns, and new product introductions. During the first quarter of 2017, DSO decreased to 68 days as of April 1, 2017.
The Company advises the Staff that the increase in DSO and slower collections during the fourth quarter of 2016 were attributable to specific reasons as follows:
• | The Company experienced longer payment cycles for certain customers it believes was due to their management of working capital. However, these customers have exhibited shorter payment cycles during the first quarter of 2017. |
• | Generally, collection cycles are longer for customers outside of the United States. There was an increase in sales to international customers in the fourth quarter of 2016 as compared to the fourth quarter of 2015 and the Company experienced turnover in collections staffing in the fourth quarter of 2016, particularly in EMEA. International revenue was 34% of revenue in the fourth quarter of 2016 as compared to 25% of revenue in the fourth quarter of 2015. |
• | An increase in sales and invoice billings later in the fourth quarter of 2016 as compared to the fourth quarter of 2015. Due to the fact that the Company’s customers’ payment terms are generally 30 to 90 days, the Company does not expect to receive payment of invoices issued later in the quarter until the subsequent quarter. |
United States Securities and Exchange Commission
Division of Corporation Finance
May 12, 2017
Page 6
The Company further advises the Staff that it will provide enhanced disclosures on material changes in DSO in future filings to discuss any material trends that have affected DSO.
Critical Accounting Policies and Estimates, page 52
6. | Your disclosures of the critical accounting policies for rights of return, stock rotation rights, price protection, and product warranties are just a restatement of the policies in the notes to the financial statements. Please tell us your consideration of providing enhanced disclosures in your critical accounting policies to discuss how the estimates and related assumptions were derived, how accurate estimates/assumptions have been in the past, and whether the estimates/assumptions are reasonably likely to change in the future. Refer to Section V of SEC Release 33-8350. |
The Company acknowledges the Staff’s comment and advises that it will ensure its critical accounting policies and estimates disclosures in the MD&A section do not simply duplicate the description of accounting policies in the notes to the financial statements for future filings. The Company will also revise its critical accounting policies in future filings to provide enhanced disclosures on how the estimates and related assumptions are derived, how accurate estimates/assumptions have been in the past, and whether the estimates/assumptions are reasonably likely to change in the future. The below reflects proposed revised disclosures with additional disclosures in bold:
“Rights of Return, Stock Rotation Rights, and Price Protection: We offer limited rights of return, stock rotation rights, and price protection under various policies and programs with our retailer and distributor customers and end-users. Below is a summary of the general provisions of such policies and programs:
• | Retailers and distributors are generally allowed to return products that were originally sold through to an end-user under provisions of their contracts, called “open box” returns, and such returns may be made at any time after original sale. |
• | All purchases through Fitbit.com are covered by a 45-day right of return. |
• | Distributors are allowed stock rotation rights which are limited rights of return of products purchased during a prior period, generally one quarter. |
• | Distributors and retailers are allowed return rights for defective products. |
• | Certain distributors are offered price protection that allows for the right to a partial credit for unsold inventory held by the distributor if we reduce the selling price of a product. |
We estimate reserves for these policies and programs based on historical experience and record the reserves as a reduction of revenue and accounts receivable. Through December 31, 2016, actual returns have primarily been open-box returns. In addition, through December 31,
United States Securities and Exchange Commission
Division of Corporation Finance
May 12, 2017
Page 7
2016, we have had minimal price protection claims. On a quarterly basis, the amount of revenue that is reserved for future returns is calculated based on historical trends and data specific to each reporting period. The historical trends consider product life cycles, new product introductions, market acceptance of products, product sell-through, the type of customer, seasonality, and other factors. Return rates can fluctuate over time, but have been sufficiently predictable to allow us to estimate expected future product returns. We review the actual returns evidenced in prior quarters as a percent of related revenue to determine the historical returns rate. We then apply the historical rate of returns to the current period revenue as a basis for estimating future returns. When necessary, we also provide a specific reserve for products in the distribution channel in excess of estimated requirements. This estimate can be affected by the amount of a particular product in the channel, the rate of sell-through, product plans, and other factors. We also consider whether there are circumstances which may result in anticipated returns higher than the historical return rate from direct customers and record an additional specific reserve as necessary. The estimates and assumptions used to reserve for rights of return, stock rotation rights, and price protection have been accurate in all material respects and have not materially changed in the past.
Product Warranty: We offer a standard product warranty that our products will operate under normal use for a period of one-year from the date of original purchase, except in the European Union where we provide a two-year warranty. We have the obligation, at our option, to either repair or replace the defective product. At the time revenue is recognized, an estimate of future warranty costs is recorded as a component of cost of revenues. The estimate of future warranty costs is based on historical and projected warranty claim rates, historical and projected cost-per-claim and knowledge of specific product failures, if any, that are outside of our typical experience. We regularly review these estimates to assess the appropriateness of our recorded warranty liabilities and adjust the amounts as necessary. Factors that affect the warranty obligation include product failure rates, service delivery costs incurred in correcting the product failures, and warranty policies. The warranty obligation does not consider historical experience of the Fitbit Force product as a separate reserve has been established for the Fitbit Force recall. Our products are manufactured by contract manufacturers, and in certain cases, we may have recourse to such contract manufacturers. Should actual product failure rates, use of materials or other costs differ from our estimates, additional warranty liabilities could be incurred, which could materially affect our results of operations. The estimates and assumptions used to reserve for product warranty have been accurate in all material respects and have not materially changed in the past.”
The Company submits to the Staff that estimates and assumptions used to reserve for rights of return, stock rotation rights, and price protection have been accurate in all material respects and have not materially changed in the past. For example, during fiscal 2016, actual returns were 91% accurate, and the variance between estimated reserves and actual returns was approximately 1% of net revenue. Price protection claims have historically been immaterial.
United States Securities and Exchange Commission
Division of Corporation Finance
May 12, 2017
Page 8
The Company advises the Staff that estimates and assumptions used to reserve for product warranty have been accurate in all material respects and have not materially changed in the past. The Company discloses changes in estimate related to pre-existing warranties within the product warranty reserve activities in note 5 to the notes to the Company’s consolidated financial statements in the Form 10-K. For example, these changes in estimate related to pre-existing warranties were $4.1 million in fiscal 2016.
The Company also advises the Staff that it does not believe there have been any material changes from the disclosures in the Form 10-K to disclose in its Form 10-Q for the quarter ended April 1, 2017. However, the Company will include the enhanced disclosures noted above in its Form 10-Q for the quarter ending July 1, 2017.
Consolidated Financial Statements
Note 2. Significant Accounting Policies
Rights of Return, Stock Rotation Rights, and Price Protection, page 70
7. | We note that you no longer disclose that you review actual returns evidenced in prior quarters as a percent of the related revenue to determine the historical return rate, and then apply the historical return rate to the current period as a basis for estimating future returns. Please tell us whether there have been any changes in how you estimate and record reserves for the limited rights of return, stock rotation rights, and price protection. Describe the terms and conditions associated with the limited rights of return, stock rotation rights, and price protection. Quantify the increase in your return reserve attributable to open box returns, stock rotation rights, and price protection in fiscal 2016. Explain how you considered each of the factors outlined in ASC 605-15-25-3 and SAB Topic 13.A.4.b.in determining whether you are still able to make a reasonable and reliable estimate of product returns. Further, tell us whether the slower collections in the fourth quarter of 2016, referred to on page 51 suggest that the payments are contingent on the sale by your customers to an end user. We refer you to ASC 605-15-25-1(b). |
The Company respectfully advises the Staff that it inadvertently excluded the disclosure related to the review of actual returns to determine the historical return rate and application of such return rate to current period as a basis for estimating future returns, and that there have been no changes in how the Company estimates and records reserves for rights of return. The Company continues to review actual returns evidenced in prior quarters as a percent of the related revenue to determine the historical return rate, and then applies the historical return rate to the current period as a basis for estimating future returns. Effective 2016, the Company estimates future stock rotation returns at the time of sale based on historical stock rotation trends. In prior periods, stock rotation reserves were recorded on a case by case basis or for specific transactions, and were not material. The Company submits that in response to Comment 6, the Company will revise its critical
United States Securities and Exchange Commission
Division of Corporation Finance
May 12, 2017
Page 9
accounting policies and estimates in the MD&A and significant accounting policies in the notes to the consolidated financial statements to provide enhanced disclosures on how these estimates and assumptions are derived in future filings. Refer to the Company’s response to Comment 6 for proposed revisions of disclosures.
The Company submits that in its response to the Staff’s letter dated March 30, 2015, it outlined its considerations of ASC 605-15-25-3 and SAB Topic 13.A.4.b. The Company advises the Staff that there have been no changes in its methodology to estimate future returns, and that it continues to determine that the Company is able to make a reasonable and reliable estimate of product returns.
The following is a description of the terms and conditions associated with the rights of return, stock rotation rights, and price protection, which will be included in revised disclosures in future filings:
• | Retailers and distributors are generally allowed to return products that were originally sold through to an end-user under product warranty provisions of their contracts, called “open box” returns, and such returns may be made at any time after original sale. |
• | All purchases through Fitbit.com are covered by a 45-day right of return. |
• | Distributors are allowed stock rotation rights which are limited rights of return of products purchased during a prior period, generally one quarter. |
• | Distributors and retailers are allowed return rights for defective products. |
• | Certain distributors are offered price protection that allows for the right to a partial credit for unsold inventory held by the distributor if the Company reduces the selling price of a product. |
The Company advises the Staff that the increase in the Company’s return reserve attributable to open box returns and stock rotation rights as of December 31, 2016 as compared to the reserve as of December 31, 2015 is outlined below, in thousands:
(in thousands) | As of 12/31/2016 | As of 12/31/2015 | Increase | Percentage Change |
Open Box Returns | $86,666 | $70,759 | $15,907 | 22% |
Stock Rotation Rights | 12,185 | 3,286 | 8,899 | 271% |
Total Sales Return Reserve | $98,851 | $74,045 | $24,806 | 34% |
The Company advises the Staff that as of December 31, 2016, the Company provided a reserve for price protection claims of $4.0 million. There was no reserve required as of December 31, 2015. The reserve for price protection claims as of December 31, 2016 was 0.2% of the Company’s revenue for fiscal 2016. The price protection reserve is recorded as a liability separately from the sales return reserve, which is recorded as a reduction to accounts receivable.
The Company advises the Staff that the slower collections experienced during the fourth quarter of 2016 do not suggest that payments are contingent on the sale by its customers to an end
United States Securities and Exchange Commission
Division of Corporation Finance
May 12, 2017
Page 10
user as retail and distributor customers are contractually obligated to pay for products purchased under standard payment terms irrespective of sell through to end customers. As described in the response to Comment 5, slower collections experienced during the fourth quarter of 2016 were due to longer payment cycles experienced for certain customers, an increase in sales to international customers as compared to the fourth quarter of 2015, an increase in billings later in the quarter as compared to the fourth quarter of 2015, and due in part to collections staff turnover, particularly in EMEA.
Sales Incentives, page 70
8. | Please explain the terms of the $42.2 million increase in rebates and promotions offered to retailers and distributors. Tell us whether any of the rebates or price reductions were granted subsequent to the initial sale and whether there was any impact on your normal return policies or your ability to conclude that your sale price is fixed or determinable. |
The Company respectfully advises the Staff that the Company experienced softer-than-expected demand during the holiday season in the fourth quarter of 2016, as disclosed in the Form 10-K. In response to the lower than expected demand and higher inventory levels in the channel, the Company for the first time discretionarily provided additional rebates and promotions (“rebates” or “rebate program”) during the fourth quarter of 2016 to its retailers and distributors. These rebates were granted to retailers and distributors subsequent to the initial sale. The majority of inventory subject to these rebates was sold to retailers and distributors in September 2016 through December 2016, and the rebates were communicated to certain customers in December 2016 and January 2017. Other inventories subject to these rebates were sold to retailers and distributors subsequent to the communication of the rebate program.
The Company also advises the Staff that the Company has not revised its normal return policies and did not provide additional contractual rights of return due to these rebate programs, and as a result, the Company does not believe the rebates will materially impact its reserve for future returns.
The Company respectfully submits that it believes it is able to conclude that its sale price is fixed or determinable taking into consideration the following factors:
• | The rebate program offered is the first program of its kind, which was discretionarily offered by the Company to its retailer and distributor customers. |
• | The Company has no contractual obligations, either explicit or implicit, to provide these rebates to its customers, however discretionary rebates were offered to certain retailer and distributor customers with the objective of encouraging sales of select products during certain holiday periods in 2017. |
• | The rebates were offered on a product-specific basis. The holiday periods subject to rebates |
United States Securities and Exchange Commission
Division of Corporation Finance
May 12, 2017
Page 11
included Valentine’s Day, Mother’s Day, and Father’s Day, and the rebates are being offered for short periods of time such as two weeks prior to Mother’s Day.
• | Rebate programs are initiated by the Company and the terms and conditions of such programs are under the Company’s control. |
• | These discretionary rebates are provided to customers in the form of credits upon receipt of sell-through data and associated rebate accruals are estimated based on the per-unit-cost of the program and the number of units expected to be sold in the short periods of time described above. Material assumptions used in our estimate include channel sell-through data, sales price data, and estimates of the likely amount of claims resulting from available rebates. The Company does not believe these assumptions and estimates are subject to a significant degree of judgment or uncertainty given the availability of channel sell-through data and sales price data, and historical sell-through experience during similar holiday periods in prior years. |
• | The Company accrues for such rebates in accordance with ASC 605-50-25-3 at the later of the date the related revenue is recognized or the date at which the rebate is offered to the customer. |
* * * * * * *
United States Securities and Exchange Commission
Division of Corporation Finance
May 12, 2017
Page 12
Should the Staff have additional questions or comments regarding the foregoing, please do not hesitate to contact me at (650) 335-7631, or, in my absence, Ran D. Ben-Tzur at (650) 335-7613.
Sincerely, | ||||
/s/ Jeffrey R. Vetter | ||||
Jeffrey R. Vetter |
cc:
James Park, President, Chief Executive Officer, and Chairman
William Zerella, Chief Financial Officer
Andy Missan, Executive Vice President and General Counsel
Fitbit, Inc.
Cynthia C. Hess, Esq.
Ran D. Ben-Tzur, Esq.
Fenwick & West LLP