syncstorage-rs/glean/data-review/request.md

Request for data collection review form

Syncstorage: Daily Active User Count

All questions are mandatory. You must receive review from a data steward peer on your responses to these questions before shipping new data collection.

1. What questions will you answer with this data?

The objective is to measure Daily Active Users for Sync (DAU). This is to be an internal metric, emitted from the server of the application. We want to understand the usage of the service per user interaction. This is important for understanding how many users have Sync enabled and at a more granular level, whether the usage is on Firefox Desktop, Fenix or iOS. Multiple actions across devices or on the same device will be reconciled to count as a single Active User.

2. Why does Mozilla need to answer these questions? Are there benefits for users? Do we need this information to address product or business requirements? Some example responses:

• We need to understand usage patterns of Sync from the server-side.
• Would provide the most granular measurement of service interaction from the server-side.
• This is necessary to establish a baseline measurement of Daily, Weekly, and Monthly usage of Sync. Essential KPI metric for Sync.
• Helps us understand usage trends related to feature changes.

3. What alternative methods did you consider to answer these questions? Why were they not sufficient?

• Previously, this metric was measured by FxA/Mozilla Accounts. Logins were associated with an OAuth Client ID. However, as Relay and other account recovery tools make use of Firefox accounts for their auth, this will alter the reliability of this metric. A user could also make use of the aforementioned services without Sync being enabled, resulting in further inaccuracies.
• A decision was made to move away from attributing active use through auth tokens towards a metric that counts each time Sync collections are accessed. Links to these decision briefs are cited at the bottom of the data review document.

4. Can current instrumentation answer these questions?

• It does at the current moment, but in a short time will no longer be able to. As described in #3, we are needing to move the metric measurement to the service itself and from the server.

5. List all proposed measurements and indicate the category of data collection for each measurement, using the Firefox data collection categories found on the Mozilla wiki.

Note that the data steward reviewing your request will characterize your data collection based on the highest (and most sensitive) category.

Measurement Name	Measurement Description	Data Collection Category	Tracking Bug #
sync_event	Event to record an instance of sync backend activity initiated by client.	Cat 2: Interaction Data	https://bugzilla.mozilla.org/show_bug.cgi?id=1923967
hashed_fxa_uid	Sync user identifier. Uses `hashed_fxa_uid` for accurate count of sync actions. This is the Firefox Accounts (FxA) User Identifier (UID) value passed through a SHA-256 hash to render a value that is unique, but ensures the privacy of the original UID. A single user could make numerous sync actions in a given time and this id is required to ensure only a single count of daily active use is made, given a number of actions.	Cat 2: Interaction Data	https://bugzilla.mozilla.org/show_bug.cgi?id=1923967
hashed_device_id	Hashed identifier of device. This is necessary to correlate postential users that do not want telemetry tracking of this sort. The "deletion-request" ping within clint-side sync is the most likely mechanism to ensure associative removal of data.	Cat 2: Interaction Data	https://bugzilla.mozilla.org/show_bug.cgi?id=1923967
platform	Platform from which sync action was initiated: Firefox Desktop, Fenix (Android), or Firefox iOS.	Cat 1: Technical Data	https://bugzilla.mozilla.org/show_bug.cgi?id=1923967
device_family	Type of device being used to make Sync action. Desktop PC, Tablet, Mobile.	Cat 1: Technical Data	https://bugzilla.mozilla.org/show_bug.cgi?id=1923967
submission_timestamp	Glean internal submission timestamp of metric ping.	Cat 1: Technical Data	https://bugzilla.mozilla.org/show_bug.cgi?id=1923967

6. Please provide a link to the documentation for this data collection which describes the ultimate data set in a public, complete, and accurate way.

• This schema (which matches the table above) is defined within the service repository in the /glean directory. Documentation for metrics collection of DAU will be within this directory and all related server-side code. Will add data to Glean Dictionary and Probe Dictionary as development process proceeds.

• Decision Brief - Server-Side Sync Usage Attribution from Mozilla Accounts

7. How long will this data be collected? Choose one of the following:

• We will retain the data at present for 1 year, as defaulted within Glean, however we can modify the Probe Scraper configuration should we deem the collection period to be shorter.
• We will want to at least retain the trends and counts of usage for at least one year.

8. What populations will you measure?

• We will measure all active use of Sync (attributed from the server-side). This includes all countries, locales and release channels which Sync is used.

9. If this data collection is default on, what is the opt-out mechanism for users?

• Our initial plan for opting out of collection would invovle the client side "deletion-request" ping. This uses the hashed_device_id value referenced above to remove entries that are liked to the user's Firefox Account.
• Alternatively, opting out would involve a user not being signed into Sync as a service. Since there is no active attribution that could identify a user, especially since we are using a hash value and generalized collection representation, there is not an immediate user concern.

10. Please provide a general description of how you will analyze this data.

• Measurement of Daily Active Use:

• Based on the individual unique user_id (hashed_fxa_uid), we will count a single active user when, during a 24-hour period, the user makes a request to any collection within their Sync account (including bookmarks, tabs, prefs, etc.). Note we are using a value derived from the user_id, not the user_id itself. This value is passed through a SHA-256 hash to render a value that is unique, but ensures the privacy of the original user_id.

• A user that initiates multiple requests to the same collection or other collections will only be counted once. This is why the unique hashed_fxa_uid is required to derive uniqueness. Active use can also be across several devices, so an accompanying platform key defines whether the user initiated the action on Desktop, iOS or Android. Device family, whether desktop, tablet, or mobile will also be of interest. Actions taken across multiple platforms for the same user should only count as an active user once, within a 24-hour period.

11. Where do you intend to share the results of your analysis?

• The Sync Ecosystem Team, related product partners, and the broader PXI organization will use this data.
• We will be using Glean for our telemetry emission, aggregation and analysis. Queries will be defined through Google BigQuery and visualized using Glean's provided tools.

12. Is there a third-party tool (i.e. not Glean or Telemetry) that you are proposing to use for this data collection? If so:

• No, since we are using Glean, this is internal.