See whether the training landed
The trend after a launch, and what changed.
- For
- L&D
- Routine
- After each launch
- Needs
- Juno only
- You approve
- every draft
What you get
The prompt
Paste it into Claude or ChatGPT. It asks its questions in one message, then starts from what you already have in Juno.
Tell me whether the training landed: Check whether a training changed anything, using our learning data first and any other tool I have connected second. Suggest, do not assume; present numbers as evidence, not verdicts. ═══ ONE ROUND OF QUESTIONS ═══ Ask me in ONE message, then wait: 1. Which training, and when it went out. Find it by name (search_lms_content) and confirm. 2. What "landed" would look like: a behaviour, a number, or a result that should move if the training worked. Propose two or three candidates from the training's own content (read_unit_text) and let me pick or add. 3. The window to compare: the weeks before and the weeks after, the same length each side. 4. Which other tools, if any, hold that outside signal (support tickets, product usage, call recordings, a survey), or whether we stay with learning data only. If a tool is not connected, say so and stop at what Juno can show. 5. The level to report at: whole audience, by department or manager, or by individual. Individual-level results stay behind my explicit sign-off before anyone else sees them. In that same first message, also ask me my role and the team I support. ═══ WHAT JUNO CAN SAY ═══ From our learning data (list_insights, then get_insights): who was assigned, who completed and when, scores, and how completion moved over the window. Show the trend in the grain the data offers and stamp each number with its as-of time. Never mix a current-state count with a trend. If learning data is not enabled on our account, say so in one line and ask me for an export instead. ═══ THE OUTSIDE SIGNAL, IF ANY ═══ Pull the chosen signal from the connected tool for the same window and the same groups, then put the two side by side: completed versus not completed, before versus after. Say plainly what a comparison can and cannot claim; a change after the training is not proof the training caused it. ═══ THE READ-OUT, THEN STOP ═══ One page: what moved, what did not, and where a weak result points back at the content rather than the people, so I can fix the training instead of blaming the learners. Frame every suggestion as a hypothesis for me to act on through the update line of this sheet; change nothing here. Offer to rerun this on the next window when I ask. Rules throughout: real data only, never fabricated scores; the same window and groups on both sides; individual results behind my sign-off; numbers relayed, not recomputed; nothing changed or sent from here; talk in product terms, not tool names. If this assistant supports artifacts connected to Juno, build it as an interactive artifact that re-reads Juno each time I open it, and say so; otherwise build it static and show the date the data was read. Routine: after the first run, offer to repeat this two weeks after each launch. If this assistant can run scheduled tasks, set one up only after I approve it; if not, give me a one-line prompt I can re-run.
What your assistant will ask you
- Q1Which training, and when it went out. Find it by name (search_lms_content) and confirm.
- Q2What "landed" would look like: a behaviour, a number, or a result that should move if the training worked. Propose two or three candidates from the training's own content (read_unit_text) and let me pick or add.
- Q3The window to compare: the weeks before and the weeks after, the same length each side.
- Q4Which other tools, if any, hold that outside signal (support tickets, product usage, call recordings, a survey), or whether we stay with learning data only. If a tool is not connected, say so and stop at what Juno can show.
- Q5The level to report at: whole audience, by department or manager, or by individual. Individual-level results stay behind my explicit sign-off before anyone else sees them.
How it runs
- 01
Connect
Juno and the tools below
- 02
Answer
its questions in one message
- 03
Check
the draft or the report it builds
- 04
Approve it
Nothing is published, assigned, or sent before that.
Routine: After each launch. The prompt offers to set it up as a scheduled task in your assistant, only after you approve.
Live viewLive view: if your assistant supports artifacts connected to Juno, you get an interactive view that reads Juno again each time you open it.
What it connects
- JunoRequired
Your assistant reads these tools and does the work in Juno.
Teams that use it
Journey Plays that go with it
Signature play
Trace a launch from training to customer usage
Who learned it, who talked about it, and which accounts now use it, every week.
Live viewEvery weekSee where everyone stands right now
One snapshot of progress, by team.
Live viewWhen neededMap completion by team and topic
See the gaps before the deadline.
Live viewEvery week




























