ChangeOS Pilot working name

Where is my agent in the change cycle, and what do I do next?

Derek's one-sentence version of the product, Sep 24. A daily view for the manager of who is falling out of adoption, a flag when they do, and the playbook behind the flag: the questions to ask, the conversation to have. First for Intradiem adoption. Proven first on our own AI tools.

Phase 1 · by Dec 31Find who is falling out. Tell the manager what to do next.Triggers and flags on the change cycle, the playbook behind each one. Internal pilot on our AI tools with data behind it; the Intradiem interface as a mockup.Open Phase 1 › Phase 2 · 2027, with EngineeringInside the product. Hub feeds it, the playbook becomes plays.Coaching modules teed up as actions the manager just hits. Nudges through the rules engine. Rules the customer writes.Open Phase 2 › Phase 3 · Derek's pie in the skyConnect everything else they use.Their other tools through connectors. Risk predicted before the drop. A conversation layer over the data. Stickiness.Open Phase 3 ›

Arrow keys move. C opens contents. Every product screen carries a sample-data ribbon: a fictional customer, four teams, 144 people. Nothing here is an Intradiem or customer result.

The session, Sep 24 · Derek's answers to the six questions

It is all about adoption of Intradiem. Period.

What it isA daily or intraday view that tells the manager where an agent is falling out of the change cycle and what to go do next. Triggers and flags, then a playbook. Not a report. Most of the time the next step is a conversation, and the product tees it up.
The dataThe two adoption metrics we already measure with every customer: accept rate on dynamic sessions and self-cure rate on alerts. Front office, back office, any product. Every customer tenant already has it. Nobody does anything with it.
The differentiatorThe playbook behind the flag. Why would they use this over something that tells them the same thing? Because it comes with the questions to ask and the conversation to have. That is where Kathryn's expertise and Joey's tribal knowledge go.
How it is soldA paid add-on to any Intradiem product. Independent of Engagement Hub and of Pro. Hub feeds it once Hub ships; Pro makes it better; neither is required. Not given away.
Not thisNot a replacement for the customer's own change management program, and not Engagement Hub Pro. Pro is about the burnout of the agent, after the fact. This is about the adoption of the product, before it becomes burnout or coaching.
Other toolsPie in the sky, his words. Connectors to the customer's other tools make it sticky and are the long-term vision. Not the first priority, and not the only input either: a tool that only shows Intradiem adoption while they check five other places will not get used.

The north star · Jen Lee's doc

Change management is not a service. It is a loop.

01SenseAccept rate, self-cure, use vs avoidance, time to adoption. Behavioral data on a daily grain, not a survey after the rollout.

02DecideWhere in the cycle did this person fall out: never bought in, never learned it, tried and stopped? Each one is a different conversation.

03ActTell the manager, with the questions. Later: nudge the agent, tee up a coaching play, adjust the workflow.

04LearnFourteen days later, did use come back? Keep what works, drop what does not, never repeat a failed play on the same person.

"The value of the product, which you don't know today, is how do I proactively identify someone has fallen out in the adoption of the tool before it's too late, and I'm just seeing it in my adoption rates."Derek, Sep 24

Derek's read on the north star: the first half of Jen's doc (readiness, telemetry, triggers) is largely covered by what is already built. The second half, the product screens and the "now what do I do with this", is where the work is.

Derek's brief, May 1 · the five deliverables

Five components. Each has a home in a phase.

Component	What it is	Phase 1	Phase 2	Phase 3
Change readiness assessment	Scores a team's readiness: workflow complexity, role exposure, manager capability, history. Kathryn's Current State Assessment, sections A to G, already a page that scores itself.	Runs first		Simulation
Adoption telemetry	Accept rate, self-cure, use vs avoidance, time to adoption. Real signals in place of surveys.	Intradiem + our AI tools	Hub feeds it	Any tool
Intervention playbook	If adoption drops below X, do Y. The questions and the conversation behind each flag. Kathryn's content.	Three flags, questions	Plays, rule builder	Automated
Manager enablement	Team view, what to do next, 1:1 tee-ups, a daily briefing. Managers lead change without becoming experts in it.	Team today	Copilot	Before the drop
ROI and adoption dashboard	Adoption percent, risk flags, and later productivity lift and savings realized against expected.	Score, cycle, heatmap	Across products	ROI tracker

Same five as Jen's doc and the charter. Internal pilot is customer zero; the playbook plus the rule builder is the intervention framework; the engine, dashboard and data contract are the telemetry; the executive presentation closes with Phase 2 as the ask and Phase 3 as the horizon.

What ships by December · agreed on the call

Two faces of one prototype: data behind one, a mockup of the other.

Face 1 · the internal pilot, interactive, data behind itOur own AI tools: the internal assistant and Copilot.Real usage exports, the three flags, managers responding, a lift window. This is the part people can play with, and the part that produces a number. Kathryn's suggestion, Derek's agreement.

Face 2 · the Intradiem adoption interface, mockup onlyWhat it looks like inside the product, fed by accept rate and self-cure.Looks like Engagement Hub, another tab on the same screen, a teaser for Hub. No data behind it: the adoption rows live in every customer's tenant and Engineering has to be in the room before anything reads them. Side quests build outside the product, then get approved in.

Why twoChange management is change management. Whether the product is AI or automation or intraday, the loop is the same once the usage data is in. Proving it on our AI tools is the honest test; the Intradiem face shows the room what it becomes.
StickinessHandle Time will be a solution of the past in a couple of years. This is another reason to keep using Intradiem no matter what else they run. It fits Matt McConnell's longer vision for the product.

Read this before you click

What is real today, and what is a drawing.

Real, running, testedThe measurement layer for the internal pilot.Adoption engine (every framework signal, four triggers, 90 checks), telemetry dashboard with a live threshold panel, manager cards with a response log that closes the loop, Kathryn's assessment as a page that scores itself, two source adapters (the internal assistant's shape and Microsoft's Copilot report shape), a delivery contract for interventions. Sample data. Live at adoption-telemetry.pages.dev.

A drawing, on purposeEverything on the Intradiem face and everything past Phase 1.Accept rate and self-cure rows, the plays that go back into the product, the seven outside connectors and their routes, the predictive weights, the ROI tracker, the simulation weights. These screens show the shape so the room can say what is wrong.

Sample dataA fictional customer, four teams, 144 people. Coached people recover because the generator says so. No lift figure here is evidence.

Placeholders that need an ownerStage labels, thresholds, the questions behind each flag and the minutes per play are stand-ins so the pages could move. Kathryn is writing the real content; Joey holds the customer friction.

HostingSample data sits on a personal Cloudflare account. Real employee or customer data never does.

To December 31

Calibrate on November 3. Present in December.

NowPlaybook and exportKathryn writes the content behind each flag (any format, JSON is fine). Joey with Jason Dowden on the assistant's export; the Copilot report request goes to the tenant admin now.

OctReal rowsExports meet the adapters. Sample flag flips off. Managers in the pilot briefed with the questions, not a course.

Nov 3Derek calibration, 9:00Face 1 running on real internal data, Face 2 as the Intradiem mockup, Phase 2 and 3 as drawings. One hour before he leaves for Nashville.

NovLift windows closeThe first real number on customer zero. Joey brings the friction the Success Managers hold.

DecExecutive presentationFramework, pilot result, the Intradiem interface, Phase 2 as the ask with Engineering, Phase 3 as the horizon.

Derek travels every week after his return until Thanksgiving. November 3 is the one calibration; nothing else waits for his calendar.

Still open after Sep 24

Four decisions the call did not close.

01 · Level of effortHow much the customer administers and configures. Derek's own warning: heavy to set up and they will never use it. Phase 1 answer here: three flags on by default, thresholds set once, nothing to build.
02 · The Intradiem rowsWho lets the prototype read accept rate and self-cure from one tenant. Anonymized rows from Matt McConnell's team would turn Face 2 from a mockup into a demo. Needs Product and Engineering in the room.
03 · Stage vocabularyDerek framed the cycle in ADKAR terms; Kathryn's rule is no ADKAR or Prosci wording in the product. Her plain-word labels replace the placeholders on the change-cycle screen.
04 · Back officeFew UPT customers live, and business readiness for Back Office Optimizer and Queue Optimizer has not happened. The Intradiem face is front-office accept and self-cure until Product says what back-office adoption is.