What shipped
Changelog
What has shipped in Facets, newest first, each entry dated by the day it shipped (UTC).
November 2026
- A project's manager can describe a rule in a sentence and have AI draft it in the rule editor, where they read it and save it, turned off. A rule can also ask AI to suggest labels, set the priority, choose a section, draft a checklist or subtasks, or estimate how long the task it runs on will take, from that task's title, description and checklist. AI never picks a person, and teammates are told what a rule sends to AI. What AI changes can set off the manager's other rules, which do what the manager wrote. How Facets compares with other project trackers. What a rule sends to AI.
- A project's manager can build automation rules of their own: when something happens to a task, or on a schedule, and the conditions hold, up to five things happen, such as moving or assigning the task, telling people on the project, posting to the project's channel, or creating a task in another of their projects. Each change says which rule made it, teammates see the rules that are on, and ten recipes start the common ones. How Facets compares with other project trackers.
- A project now shows its critical path: which tasks set the end date, and how many working days every other task can slip before that date moves, with a filter for the critical tasks. The manager can keep up to five named baselines and compare the plan with any of them, task by task and for the finish date. Teammates see both and change neither. How Facets compares with other project trackers.
- Where a workspace's owner or admin allows it, a manager can let an AI assistant they connect read the names of the people they manage and when their 1:1s are, and, under a tick of its own, what the open follow-ups and the manager's points for next time say. It can add a point for next time, or a follow-up of the manager's own that waits until the manager keeps or removes it, and each is marked as added through an AI assistant. Each permission is unticked until the manager ticks it. It never reads reviews, goals, check-ins, assessments, 1:1 notes, highlights or anything from a survey, and no email or message is sent for what it adds. How to connect an assistant, and what it reads. What a connected assistant can see.
- A comment on the team survey, the engagement survey or the Leadership 360 is saved on the person's link as they type it, and is no part of their answers until they submit. The team report and the program report can be printed with chosen measures left out, and the printed copy says how many it leaves out. Who sees survey answers, and when.
- A teammate can look at an idea board by status and see which ideas changed since they last looked. On a board opened on or after 19 November 2026 (UTC), the themes an AI model suggests come with a short summary of what the ideas say together, shown to the manager alone. After a board closes, each person on it can answer "Was this worth your time?" for seven days, and the manager sees only how many chose each answer, once those days have passed and only where at least three people answered and not everyone did. Who sees a board's answers and counts.
- In a project's quick add, a due day typed in plain words at the end of a title is shown as a due date before the task is saved. The task list has keys to move between tasks and open a task's status, assignee, due date, labels or priority, and the search box finds tasks in a person's running projects. A project's owner can undo, from a task's activity trail, an AI assistant's change to its status, owner, dates, title or section. What a connected assistant may change.
- Before each 1:1 the manager sees a brief of what is still open and what changed since the last one, and keeps a list of points for next time. After it, an AI model can propose a summary, follow-ups and points for next time from the manager's notes or a pasted transcript, each kept only if the manager adds it; the transcript is sent once and not kept. Highlights are the manager's own notes of what a person did well, and are never sent to an AI model. Who sees a manager's 1:1 records.
- A manager can rehearse a hard conversation on the Conversations page, from a list of 15 scenarios or one they describe, with no person's record involved. They choose how hard the other person is and, on a listed scenario, how that person responds; they type or speak each turn; and the feedback at the end is on their own words. Nothing typed or said is kept, and each reply and the feedback use the manager's AI credits. Who can see a practice.
- A manager can speak a practice turn instead of typing it, in a browser that can recognize speech on their own device, such as Chrome on a computer. The sound stays on that device: it is not recorded, kept or sent. They see how the turn sounded as plain numbers: pace, pauses, filler phrases, pitch range and loudness range. Who can see those numbers.
- A team of nine or more can run a lighter team survey: each person answers the three extended dimensions about five teammates, picked by a draw, and everything else about everyone. It saves each person about 2 to 4 minutes of 35 to 40. The lead chooses it when opening a run, where the extended dimensions are on; a full run stays the default, and a run in the research program is always full. Minutes per person, by team size.
- The team survey and the Leadership 360 can be answered in French as well as English and Spanish, so every survey Facets runs is now offered in all three: one language a run, chosen when it is set up. A language model made the French, which was back-translated and reconciled; no professional translator or native speaker has reviewed it. Results stay in English, and runs in different languages are not compared. How each translation was made and checked.
- A manager can practice a hard conversation before having it. A stand-in plays the other side from the conversation guide the manager prepared, and the feedback at the end is on the manager's own words. Nothing of a practice is kept, and nobody else can read it, an owner or admin included. It needs AI switched on for the workspace, and each reply and the feedback uses the manager's AI credits. Who can see a practice.
- Version 1.1 of the iPhone and iPad app is on the App Store. You can sign in with a passkey, see who has answered a team survey and nudge each person once, follow up a Leadership 360 and compare it with your last one, read the topics a person added before a 1:1, and open a team survey or a 360 in Spanish where your workspace offers it. Facets.Team on the App Store.
- Facets is on the App Store for iPhone and iPad, for anyone with a Facets account: the day's list, projects and tasks, idea boards, your people's records and survey runs with their reports, with records about people and surveys kept behind Face ID. People who answer a survey or work on a project through a link need no app. Plans and payment stay on the website. Facets.Team on the App Store.
- A member can now let an AI assistant they connect add tasks to the projects they run and change them: a task's title, description, status, priority, assignee, section, dates and estimate. They allow it by ticking it when they connect; an assistant connected earlier changes nothing until it is connected again with the box ticked. It can never delete or archive anything, add or remove people, or change a project's settings, and no email or message is sent for a change it makes; the people on the project see it in Facets and in their regular digest. Every change is recorded and counted, and marked in the task's history as made through an AI assistant. How to connect an assistant, and what it reads. What a connected assistant can see.
- A member can connect an AI assistant of their own, such as Claude or ChatGPT, to read the projects they run and the idea boards that said before they opened that an assistant may read them. It read only at first (the entry above says what it may now change), it is never told who wrote an idea, and every read is counted and shown to the member, who can disconnect it. It read nothing of the people side at first (the 18 November entry on People records says what it may read there now), and still no check-in, assessment, survey answer or 360. How to connect an assistant, and what it reads. What a connected assistant can see.
- Engagement cycles, their pulses and the onboarding and exit surveys can be answered in French as well as English and Spanish, and the team survey and the Leadership 360 in Spanish as well as English: one language a run, chosen when it is set up. A language model made each translation, which was back-translated and reconciled; none has been reviewed by a professional translator. Results stay in English, and runs in different languages are not compared. How each translation was made and checked.
- Owners and admins on the Organization plan can open a calibration session: each manager's people by name with their check-in band and their assessment's band, a draft's band marked as a draft, and never the words. Every read is logged and shown to the manager, a manager's distribution is drawn where they track at least three people, and the manager alone changes a band; a flag asks the manager to look again. Calibration for an organization. Who sees a calibration session.
- An organization can turn on an onboarding survey, sent about 30 and 90 days after someone starts, and an exit survey, sent when someone leaves. Answers are kept without the person's name, unit or manager; owners and admins read the results for the whole organization only, by calendar quarter, after the quarter has ended and once at least 5 people have answered, and a manager never sees them. The surveys on the engagement page. Who sees an onboarding or exit survey.
- Between engagement cycles an organization can run a pulse: 2 to 4 of the last cycle's measures, asked of the roster as it stands, under the same floors and the same hiding rules, and compared with that cycle where the comparison is reliable. A pulse has no open questions and no summary; 2 come with each cycle, and a further one is priced by band. The pulse on the engagement page. Who sees a pulse.
- Owners and admins on the Organization plan see the manager tools in counts: for each manager, check-ins due and done, assessments marked final and acknowledged, and bands where a manager tracks at least three people, with a CSV of finalized assessments and no words; every read is logged and shown to the manager. The Engagement page counts follow-through across the year's cycles: groups published and, where three or more were published, how many held a conversation and how many commitments were made and met, never which. The manager tools for an organization. Who sees the counts.
- A published engagement cycle shows owners and admins a heatmap of its groups' results on each measure. A cell is colored only where the group's own 95% range leaves out the organization's figure, above or below; the rest stay plain, as too close or too few to call. Every figure is the published results' own.
- Facets has opened a research program to test its team survey on real teams. The first 60 to 100 teams to join run their team surveys free, two runs a team, in exchange for three commitments, and their answers are pooled with other teams', without anyone's name or the team's. Nothing has been measured yet: the instrument specification names the analyses, and they run once 60 teams have records, and not before. The offer and its three commitments. The research program in the instrument specification.
- The person a manager meets for 1:1s has a page of their own, by the same personal link and with no account. Where the manager chooses, it shows their follow-ups to tick and their goals and asks for a reflection before each check-in, and it carries a final assessment the manager shares, which they mark as read and may answer once.
- The getting-started checklist has a step for owners and admins to connect Slack or Microsoft Teams, and the Settings guide explains what Slack and Teams do in Facets in one place.
- An idea board or a project can send its members their personal link by Slack or Microsoft Teams direct message, with a board's reminders, and a count from the daily summary beside the email. No message carries an idea, a comment or a vote.
- A manager can have their own reminders and notices by Slack or Microsoft Teams direct message beside the email, and send the people they track or ask a copy of their 1:1 invitation, agenda link and input request the same way, each under a switch of the manager's own.
October 2026
- The team survey, its pulse, the Leadership 360 and its follow-up pulse can go by Slack or Microsoft Teams direct message where the organization chooses it, with the email for anyone Slack cannot find. Each person's page says how their link came.
- Workload shows how full each person's coming weeks are against the hours a week noted for them, across projects and work tracked in other tools.
- Microsoft Power BI can read a manager's projects, goals and Workload through a read-only feed, under a key only that manager holds.
- When a workspace removes Slack or revokes its access, every owner and admin is told at once, by email and on the dashboard. Before a cycle opens, a reach check counts how many on its roster Slack can find, and keeps only the counts.
- A manager can send a project's tasks to a basic Microsoft Planner plan shared with a group, through their own Microsoft 365 work account. Completing a task on either side completes it on the other at the next check.
- Microsoft Entra ID or Okta can keep a survey roster through SCIM 2.0. Facets previews the changes and applies them when an owner or admin confirms, and offers a manager the people it says report to them.
- Every member of an engagement cycle, however large, can be invited and reminded by direct message, with the email for anyone not reached within six hours. Anyone can ask for email instead, from a link in any message.
- Engagement cycles can be answered in Spanish as well as English, one language a cycle. A language model translated the Spanish, which was then back-translated and reconciled.
- An engagement cycle's invitations and reminders can go as a direct message in the organization's own Slack or Teams, with the email for anyone not reached.
- The person a manager meets for 1:1s can add topics to the agenda through a personal link, without an account.
- A manager's own Slack or Teams channel can be told about survey results and checks due, and an engagement cycle can have a channel of its own.
- A Leadership 360 includes a follow-up pulse 60 to 90 days after the brief, and compares the brief with the leader's last review.
- A team keeps a working agreement after its debrief, and each person can choose which behavior to work on.
- Beneath an engagement cycle's written summary, the organization sees options from a cited library.
- A getting-started checklist and tour walk a new account through setting up.
- A manager can record growth and stay conversations, and email a calendar invitation for a repeating 1:1.
- You can sign in with a passkey.
- Projects can be imported from Microsoft Project and exported for it, as XML.
- Each of a practitioner's clients can have a workspace of its own, with the practice's name and logo beside ours. A debrief can be run in the product, and practitioners gained the program report and the facilitator kit.