Privacy guide for Microsoft MyAnalytics
MyAnalytics helps users better understand how they collaborate with colleagues and how they spend their time at work. It provides useful information that can help users prioritize and spend their time effectively. The goal of this guide is to provide a high level overview of the types of data collected and used by MyAnalytics, and how privacy, choice, and controls were incorporated into its design.
MyAnalytics provides insights using three types of data:
- Email and calendar activity that is available in the user’s Office 365 mailbox, such as time spent in meetings, emails sent to a specific person, or statements that a user makes in emails.
- Data about the actions of people that the user emails with, but in an aggregated form designed to protect individual privacy.
- Office 365 data generated by the actions of people across the user’s organization, also in aggregated form.
Data from external users is not used in MyAnalytics.
The first type of data is referred to as “Mailbox Data.” The second and third data types are referred to as “Incremental Data.”
MyAnalytics surfaces Mailbox Data in entirely new ways. For example, MyAnalytics provides views which allow users to quickly understand how much time they spend in meetings, how much time they spend in email every day, who they collaborate with the most, who they are losing touch with, and to whom they have made commitments and requests. All of this is surfaced entirely from information that is already available to the user in their mailbox.
Using Incremental Data, MyAnalytics provides organization-wide benchmarks so users can effectively compare their work habits with others across their organization. Incremental Data is also used by the MyAnalytics Outlook add-in, which provides statistics on the activity surrounding the user’s email. When using Incremental Data, MyAnalytics does not identify users by name in an effort to protect individual privacy.
... into insights
To understand how Mailbox Data serves as the basis for many MyAnalytics insights, consider the following scenarios. If a user wanted to discover who sent them the most email over the last week, they could manually count the total number of emails received from all of the different people they work with, and then rank them from the person who has sent the most to the person who has sent the least. Similarly, using simple math, a user could determine the average response time to their emails using timestamp information readily available in their mailbox. Furthermore, a user could review their emails to create a list of each person to whom the user has made a commitment. There are a wide variety of metrics and quick views that can be developed from a user’s own Mailbox Data:
- The amount of time spent on meetings and emails.
- The time of the day a user is most active on email.
- The amount of meeting-free blocks of time on a user’s calendar that can be spent focused on uninterrupted work.
- Colleagues with whom a user emails and meets with most frequently.
- Average response time to emails received and sent.
- An aggregated view of things that a user has agreed to do for others.
For example, the image below shows a user’s MyAnalytics personal dashboard. It shows how much time the user spent on emails and meetings, the time they have to themselves (Focus hours) and time they spend outside of work (After hours).
All of the statistics below are computed based on information that is already available to the user from their mailbox, and is part of Mailbox Data.
See MyAnalytics Dashboard for information about how these and other metrics are computed.
Another example of how MyAnalytics uses Mailbox Data is through commitment reminders. MyAnalytics uses automated machine learning to identify instances in emails where a user has promised to do something for someone else. This information is readily available to the user in the user’s emails. MyAnalytics simply presents this information in a way that is easier for the user to keep track of and view.
In addition to insights based on Mailbox Data, MyAnalytics surfaces statistics based on Incremental Data, such as the average time people across an organization spend in meetings and email, and when people are most likely to read email. A key point about these statistics is that they are presented in aggregate form using "de-identified" data in an effort to protect individual privacy. MyAnalytics does not use Incremental Data for the purpose of identifying any individual’s activity.
“De-identified” means when data may relate to a specific person, but there is a high level of confidence that the actual human subject of the data cannot be identified, directly or indirectly, by the data alone or in combination with other retained or available data.
For example, MyAnalytics lets the user view what the organizational averages are for time spent in meetings as part of their dashboard view. No one is identifiable from this view.
Another example of Incremental Data that MyAnalytics surfaces to a user are insights into what percentage of their emails are being read. The image below shows what percentage of emails are read when the recipient is directly addressed, and what percentage is read when the recipient is part of a group.
This information is not available to the user from their mailbox. Therefore, it is Incremental Data, but it is presented in a way that is intended to avoid identifying any individual or their activity.
Privacy from the ground up
MyAnalytics makes all of these insights available through a dashboard that is personalized for each user. MyAnalytics has no mechanism or option that allows anyone but the user to access their dashboard. Specifically, MyAnalytics does not give any user’s manager or organization access to the user’s dashboard.
MyAnalytics is designed to limit what users see in their dashboards about specific people to what is already in their mailbox.
When it comes to using Incremental Data for statistics, MyAnalytics strives to protect individual privacy by reporting these types of insights only when a minimum group size is met . In addition, statistical results of “0%” or “100%” are not shown because that would identify individual activity, that is, the user would know that none or all email recipients have read the user’s email. Rather, the user will see “low read activity” or “high read activity,” or similar phrasing. Using personalized dashboards and aggregated and de-identified data, MyAnalytics is designed to safeguard user privacy.
The current version of MyAnalytics enforces a minimum group size of five.
When users contribute to Incremental Data, they empower themselves and their colleagues with valuable context and comparisons. However, users who are not comfortable with contributing their data to Incremental Data statistics can opt-out of MyAnalytics at any time. They can also change their minds and opt back in whenever they want.
Choice and control when enabling MyAnalytics
Along with these privacy safeguards, MyAnalytics provides flexible and configurable controls which are designed to enable organizations and their users to address varying legal and policy needs regarding privacy and use of employee data.
When enabling MyAnalytics for the organization, admins can make the following choices:
- Which users have access to, or can use, MyAnalytics.
- Admins can determine which users will access and use MyAnalytics by issuing licenses to only those individual users who should have access (“licensed users”).
What the licensed user’s initial default setting is for contributing to Incremental Data.
- Admins can configure MyAnalytics to be default off, meaning that licensed users must individually opt-in to MyAnalytics in order to contribute to Incremental Data and have access to their dashboard and Outlook add-in. Alternatively, MyAnalytics can be configured to be default on, meaning that licensed users will automatically contribute to Incremental Data and have access to their dashboard and Outlook add-in, but can subsequently opt out through the Settings menu.
Licensed users need to be in the opted-in state in order to contribute to Incremental Data and access their MyAnalytics experience. If a user opts out, they lose access to their MyAnalytics experience.
Which users in sensitive roles who should not ever contribute to Incremental Data.
- Some organizations may have users in roles that may be deemed unsuitable for contributing to Incremental Data. To support this, MyAnalytics provides admins the ability to mark such users as “excluded.” Excluded users will not be able to opt-in to contribute to Incremental Data. However, the MyAnalytics experience will still be available to such users provided they are licensed.