0001628280-20-013491.txt : 20201209 0001628280-20-013491.hdr.sgml : 20201209 20200911152449 ACCESSION NUMBER: 0001628280-20-013491 CONFORMED SUBMISSION TYPE: CORRESP PUBLIC DOCUMENT COUNT: 2 FILED AS OF DATE: 20200911 FILER: COMPANY DATA: COMPANY CONFORMED NAME: Cloudera, Inc. CENTRAL INDEX KEY: 0001535379 STANDARD INDUSTRIAL CLASSIFICATION: SERVICES-PREPACKAGED SOFTWARE [7372] IRS NUMBER: 262922329 STATE OF INCORPORATION: DE FISCAL YEAR END: 0131 FILING VALUES: FORM TYPE: CORRESP BUSINESS ADDRESS: STREET 1: 395 PAGE MILL ROAD CITY: PALO ALTO STATE: CA ZIP: 94306 BUSINESS PHONE: 650 644 3950 MAIL ADDRESS: STREET 1: 395 PAGE MILL ROAD CITY: PALO ALTO STATE: CA ZIP: 94306 CORRESP 1 filename1.htm Document
image_01.jpg



September 11, 2020

VIA EDGAR TRANSMISSION

United States Securities and Exchange Commission
Division of Corporation Finance
Mail Stop 3030
Washington, D.C. 20549

ATTN:Brittany Ebbertt, Senior Staff Accountant
Melissa Kindelan, Senior Staff AccountantFOIA Confidential Treatment Requested Under 17 C.F.R. § 200.83

Re:Cloudera, Inc.
Form 10-K for the fiscal year ended January 31, 2020
Filed March 27, 2020
File No. 001-38069

         
Dear Mss. Ebbertt and Kindelan:

On behalf of Cloudera, Inc., a Delaware corporation (the “Company”), we hereby submit this letter (this “Letter”) in response to comments made by the staff (the “Staff”) of the Securities and Exchange Commission (the “Commission”) in a letter from the Staff dated August 27, 2020 (the “Staff Letter”) regarding the above-referenced Annual Report on Form 10-K for the fiscal year ended January 31, 2020. This response also references the letter from the Staff dated July 31, 2020 (the “Initial Staff Letter”), as well as Cloudera’s response to the Initial Staff Letter dated August 20, 2020 (the “Initial Response)”.

We are respectfully requesting confidential treatment for certain portions of this Letter pursuant to Rule 83 promulgated by the Commission, 17 C.F.R. § 200.83. This Letter is accompanied by such request for confidential treatment because of the commercially sensitive nature of the information discussed in this Letter. A redacted letter will be filed on EDGAR, omitting the confidential information contained in this Letter.

The following responses are numbered in accordance with the comments in the Staff Letter. For your convenience, we restate the comment from the Staff Letter in bold before each response.

Form 10-K for the fiscal year ended January 31, 2020

CONFIDENTIAL TREATMENT REQUESTED BY
Cloudera, Inc.
CLDR-1001



Page 2

Item 7. Management's Discussion and Analysis of Financial Condition and Results Operations, page 48

1. You indicate in your response to prior comment 1 that renewal rate is an evolving metric. You further state that management evaluates renewals using different measures that are not calculated on a consistent basis. Please clarify what makes this an evolving metric. Also, explain further the different measures that management uses to evaluate renewals and how the calculations differ. In this regard, in your April 30, 2020 Form 10-Q, you disclose that revenue increased from the three months ended April 30, 2019 to the three months ended April 30, 2020 due, in part, to a higher renewal rate. Please tell us the renewal rates for these two periods and, if not calculated consistently, explain why such reference is an appropriate explanation for the increase in your revenue.

Alternatively, if the renewal rates are calculated consistently, please further explain why you believe disclosing such rates is not useful information for investors.

Response: We have separated our response to this comment into several subsections below, corresponding to the different questions within the comment.

(a) Please clarify what makes [renewal rate] an evolving metric.

Our renewal rate is an evolving metric given we have changing pricing and usage models and we may measure renewal rates in multiple ways. In addition, prior to our merger with Hortonworks in January 2019, Hortonworks did not consistently calculate and disclose renewal rates, and therefore we have limited history with a significant portion of our business, which makes renewal rate comparisons of year-over-year periods difficult. For these reasons, we have not disclosed a renewal rate metric historically. We recently introduced Annualized Recurring Revenue (ARR) as our primary metric to monitor customer retention and growth. Our focus of analysis has been on ARR, and we have not devoted as much time to analyzing separately renewal rates or other metrics for understanding customer retention and growth. We are continuing to develop our thoughts on renewal rate metrics, including how best to systematize and present such a metric to investors.

We have recently introduced a consumption-based cloud pricing model, which enables our customers to purchase and use our software in a consumption-based manner (for example, hourly) rather than on a subscription basis. This shift in pricing and usage adds complexity to calculating and determining renewal rates on a consistent basis, as noted above.

Renewal rates can be measured in multiple ways. For example, renewal rates can be measured on a cohort basis (which refers to the subset of contracts up for renewal in a given quarter) or on a when-renewed basis (based on when the contract was signed). Renewal rates can also include or exclude the benefit of multi-year transactions. In addition, renewal rates can be measured on a quarterly basis or averaged over a year and
CONFIDENTIAL TREATMENT REQUESTED BY
Cloudera, Inc.
CLDR-1002



Page 3

be measured based on ARR or bookings. Further, a lost renewal can be measured either when a customer decides to cancel its contract with us or when the contract is actually terminated. Depending on the internal use, we may evaluate renewals using any or all of these approaches as further detailed in response (b), below.

For the reasons described above, we currently believe that a renewal rate is an evolving metric for us, which is a factor in our decision not to report a discrete renewal rate to our investors.

(b) Explain further the different measures that management uses to evaluate renewals and how the calculations differ.

Cloudera uses several measurements to evaluate our renewals:

ARR growth by cohort
(ARRCurrent Date / ARRCurrent Date - One Year) - 1
We exclude new customers from the cohort
We focus on net retention (renewal + expansion - churn) of our existing customers
Renewal rate on a cohort basis including the effects of multi-year transactions
(ARRCurrent Date - One Year - ChurnMeasured at the end of the Current Qtr ) / ARRCurrent Date - One Year
This as an ARR-based renewal rate
Renewal rate on a cohort basis excluding the effects of multi-year transactions
(Renewal OpportunityCurrnet Qtr - ChurnCurrent Qtr) / Renewal OpportunityCurrent Qtr
This is a bookings-based renewal rate where:
Renewal rates are measured at the end of the current cohort quarter
The calculated renewal rate looks at both contracts that have been signed as of the end of the current cohort quarter and those projected to be signed
We typically continue to re-measure this renewal rate for approximately two quarters after the cohort quarter end to determine the final renewal rate
Renewal rate on a cohort basis including the effects of multi-year transactions
(Renewal OpportunityCurrent Qtr - ChurnCurrent Qtr+ Value of multi-year transactions as if they renewed on their contract anniversary date) / (Renewal OpportunityCurrent Qtr + Value of multi-year transactions as if they renewed on their contract anniversary date)
This is a bookings-based renewal rate measured similarly to the previous measure
Renewals vs plan
(Actual RenewalsCurrent Qtr - Plan RenewalsCurrent Qtr) / Plan RenewalsCurrent Qtr
We look at the value of renewals and compare to plan
This metric does not normalize for early or late renewals
Renewals vs previous period
CONFIDENTIAL TREATMENT REQUESTED BY
Cloudera, Inc.
CLDR-1003



Page 4

(RenewalsCurrent Qtr - RenewalsPrevious Period) / RenewalsPrevious Period
We look at the value of renewals and compare to previous periods such as the same quarter in the previous year, or the prior sequential quarter
This metric does not normalize for early or late renewals
We will occasionally modify these calculations to normalize for large transactions or transactions where the termination decision was made well before the beginning of the quarter. As such, there are several additional permutations.

(c) In your April 30, 2020 Form 10-Q, you disclose that revenue increased from the three months ended April 30, 2019 to the three months ended April 30, 2020 due, in part, to a higher renewal rate. Please tell us the renewal rates for these two periods and, if not calculated consistently, explain why such reference is an appropriate explanation for the increase in your revenue. Alternatively, if the renewal rates are calculated consistently, please further explain why you believe disclosing such rates is not useful information for investors.

The renewal rate, as measured using ARR and including the impact of multi-year transactions, improved from [***] in fiscal Q1 2020 to [***] in fiscal Q1 2021. Renewal rates, as measured using bookings, also improved year over year.

The renewal rates used in the foregoing calculations were calculated consistently. The renewal rate disclosure provided a directional understanding of renewal behavior as a driver of revenue in the quarter. As such, we believe the statement about renewal rates was an appropriate explanation. However, we believe a specific renewal rate metric would not add meaningful incremental value to our investors for the reasons described in our response to subquestion (a) above.

Note 11. Income Taxes, page 98

2. In response to prior comment 5, you indicate that the impact from the one-time tax provision of approximately $72 million was fully offset by a valuation allowance. However, it appears that the valuation allowance impacted the provision by only $8.6 million. Please explain.

Response: In response to the Staff’s comment, the $72 million change in valuation allowance related to the one-time tax provision is one of several components of the total net change in valuation allowance of $8.6 million presented within the tax rate reconciliation table. Other components within the net change in valuation allowance include changes in other deferred tax assets that are not more-likely-than-not to be realized, such as the net operating loss carryforwards due to current year activity and research tax credits generated during the year.




CONFIDENTIAL TREATMENT REQUESTED BY
Cloudera, Inc.
CLDR-1004



Page 5

* * *


We hope the foregoing is responsive to your comments. Please feel free to contact me if you have any questions regarding this correspondence.


Very truly yours,
/s/ Jim Frankola
Jim Frankola
Chief Financial Officer
Cloudera, Inc.
cc:David Howard, Cloudera
Tad J. Freese, Latham & Watkins LLP
Philip S. Stoup, Latham & Watkins LLP
CONFIDENTIAL TREATMENT REQUESTED BY
Cloudera, Inc.
CLDR-1005


GRAPHIC 2 image_01.jpg begin 644 image_01.jpg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end