Get meeting outcomes

What was committed

Mia will revise the first-run screen so a difficulty toggle appears before the listening sample, and redesign the progress ring on screen 2 so it is obvious what it is counting.

Why

In the design review nobody could say what the progress ring measured, and the difficulty toggle was decided as part of the listening-first onboarding flow that ships this month.

Done when

• Difficulty toggle shows before the first listening sample
• Progress ring has a label and counts lesson steps
• Updated flow is in the design file and signed off by design

Context

Learners who open the podcast in their first week churn at twice the rate of those who reach it later, because the episodes assume vocabulary they have not learned yet.

How we decided

The question

The new onboarding opens with a 2-minute listening sample followed by a comprehension check instead of a placement quiz. How should this flow handle learners who start with literally zero vocabulary? Should they even receive a listening sample?

Why it matters

Absolute beginners are the largest signup group. If the first thing they hear is a sample they cannot follow, we risk making them feel lost in the first minute instead of showing them the value of listening-first learning.

Options raised

• Skip the sample for self-declared beginners and go straight to first words
• Play a slower, ultra-short sample built only from cognates
• Keep the sample but frame it as a preview, not a test

Description

Priya Rao, the district admin who championed Saga at Northwind Academy, is leaving in two months with no named successor, and the renewal lands 90 days out.

Risk to mitigation

What's blocked

The streak repair release is ready, but Apple rejected the build over the new in-app purchase flow, so learners keep losing streaks to the timezone bug it fixes.

Unblock path

Shared screen: what learners face after a weekend away

Verbatim

"I skipped two days on a trip and came back to 140 reviews. It felt like punishment, so I just didn't open the app again for a week."

What's behind it

The review scheduler keeps stacking due cards with no daily cap. A learner who pauses for even a couple of days returns to a wall of reviews, and the session estimate on the home screen makes it look worse than it is.

What was said

Priya Rao, a district admin at Northwind Academy, said their pilot committee is also evaluating LingoPeak. Teachers there liked its dashboard showing which students practiced this week and where they plateau. She asked directly whether Saga can show per-student practice at the class level.

Where we stand

• Saga's school reporting shows streaks per class, not per student
• LingoPeak has no listening or podcast practice, our strongest differentiator
• Priya called the dashboard a must-have for the spring rollout

Suggested response

Lead with listening practice and podcast lessons, then confirm the per-student progress view planned for the schools tier. Flag this deal to product so the dashboard timeline reflects a live requirement.

Job statement

When my workweek fills up, I want to keep making visible progress in Spanish in the small gaps I have, so my streak and confidence do not collapse.

From ask to job

What we learned

Across the last eight onboarding interviews, every learner who read a personalized story in their first week was still active a month later. Learners described the stories built from their own vocabulary as the moment Saga 'clicked', while drill-only learners struggled to say why they should come back.

Supporting evidence

• Retention cohort: week-one story readers hold streaks 2.1x longer
• 6 of 8 interviewees named stories as their favorite feature, unprompted
• Drop-off concentrates on day 3 for learners who never opened a story

Why it matters

Stories look like our strongest retention lever, but today they unlock after lesson five. Surfacing the first story earlier in onboarding is the obvious experiment to run next.

Shared screen: where 4 of 6 gave up

Task attempted

Recover a lost streak after skipping a day of lessons.

Observed friction

The streak-repair offer sits behind the flame icon, but participants read the greyed-out flame as "streak gone for good" and never tapped it. Four of six gave up on the home screen without finding the repair flow.

Think-aloud quote

"Oh no, it's grey. I guess my streak is just dead. I'm not going to tap a dead flame."

Recommendation

Surface the repair offer proactively: show a one-tap "Repair your streak" prompt on the first session after a missed day instead of hiding it behind the flame icon.

Shared screen: what Amara saw after one missed day

Verbatim

I missed one day because I was sick, and the first thing Saga showed me was a price. It felt like the app was charging me for being human.

What we heard

Amara, a 9-month learner with a 140-day streak, lost it after one missed day. The streak repair offer appeared before any encouragement, which she read as the app monetizing her slip rather than helping her recover. She said it made her want to open the app less, not more.

Signals to watch

• She hesitated when asked if she would recommend Saga to a friend
• Second interview this month describing streak repair as 'punishing'
• She suggested one free repair per month would feel fair

Shared screen: the audio-only card Ravi muted in public

Quote

"On the bus I keep it on silent. I just guess what the word sounds like and move on."

Observation

Audio-dependent exercises silently fail for public-transit learners. Ravi never saw an error, he just lost the listening practice. A muted-mode fallback, like captions or a tap-to-reveal phonetic hint, could keep these sessions useful.

What's broken

Diacritics fade out on intro cards the first time a learner meets a word. Intro cards should always show full tashkeel, so beginners can read the word correctly before it starts fading in review.

Steps to reproduce

• Start a session with at least one new word
• Open the intro card for that word
• See the diacritics already partly faded

Expected

Intro cards render full tashkeel. Fading only begins once the word enters review.

Prompt

Schedule the daily streak reminder using each learner's device timezone instead of server time. Store the timezone on the profile at app open, compute the send time locally per learner, and skip the reminder if the lesson is already done that day.

Context

All reminders currently fire at 18:00 UTC, so learners in Asia get them after midnight and lose streaks they thought were safe. Support sees a steady stream of streak-loss complaints from these regions.

Constraints

• Reuse the existing push pipeline, no new delivery service
• Fall back to the account's signup timezone when the device value is missing
• One reminder per learner per day, even across timezone changes

Acceptance criteria

• A learner in Tokyo and one in Lisbon both get the reminder at 18:00 local time
• No reminder is sent once the day's lesson is complete
• Travelling across timezones never produces a duplicate reminder

Condition

Learners west of UTC got their streak reminder in the early afternoon, so many missed the evening window and lost the streak.

Root cause trace

Lesson

Any critical generation dependency needs capacity planning, quota thresholds, and a fallback path. Treat vendor quotas like production capacity, not a billing detail.

Why

A single exhausted text-to-speech quota stopped podcast generation for every affected learner, and the team first heard about it from support tickets, not monitoring.

What we changed

• Quota usage alerts fire at 80 percent for every critical vendor
• A fallback speech provider takes over when the primary is out
• New dependencies get a capacity review before launch

Shared screen: day-7 retention as streak repair rolled out

Since last

Shipped streak repair to 10 percent of learners and reviewed the first results in the growth sync. Early numbers look good: about a third of lapsed learners use the repair, and most of them finish a lesson the same day.

Next

• Widen the rollout to 50 percent by Thursday
• Add the repair prompt to the reminder email
• Draft the day-7 retention readout for the next growth sync

Blockers

The reminder email still renders the old streak count for repaired streaks. Waiting on Theo for the template fix before widening the rollout.

Shared screen: relocating-staff conversational rate against the Q3 target

Supporting quote

"Our goal is 70% of relocating staff conversational by Q3. If we hit that, the renewal is easy." Their L&D lead, Priya Rao, said this in the QBR.

Context

Meridian is moving 140 staff to its new Lisbon hub and rolled out Saga for Portuguese. Weekly active usage has slipped to about 45% from 70% at launch, so hitting the target needs a usage recovery before the 90-day renewal check-in.

What they asked for

A way to export each class's progress as a simple report: lessons completed, streaks, and time spent per student. Today admins click into every student one by one, which does not work for a school with 400 learners.

Why it matters

• Northwind needs the report for their quarterly board review
• Two other pilot schools asked for the same thing this month
• Blocks the renewal conversation scheduled for August

Suggested shape

A per-class CSV export from the admin dashboard covering a selectable date range. A shareable summary view could come later, the CSV alone unblocks them.

What was committed

Mia will revise the first-run screen so a difficulty toggle appears before the listening sample, and redesign the progress ring on screen 2 so it is obvious what it is counting.

Why

In the design review nobody could say what the progress ring measured, and the difficulty toggle was decided as part of the listening-first onboarding flow that ships this month.

Done when

• Difficulty toggle shows before the first listening sample
• Progress ring has a label and counts lesson steps
• Updated flow is in the design file and signed off by design

Context

Learners who open the podcast in their first week churn at twice the rate of those who reach it later, because the episodes assume vocabulary they have not learned yet.

How we decided

The question

The new onboarding opens with a 2-minute listening sample followed by a comprehension check instead of a placement quiz. How should this flow handle learners who start with literally zero vocabulary? Should they even receive a listening sample?

Why it matters

Absolute beginners are the largest signup group. If the first thing they hear is a sample they cannot follow, we risk making them feel lost in the first minute instead of showing them the value of listening-first learning.

Options raised

• Skip the sample for self-declared beginners and go straight to first words
• Play a slower, ultra-short sample built only from cognates
• Keep the sample but frame it as a preview, not a test

Description

Priya Rao, the district admin who championed Saga at Northwind Academy, is leaving in two months with no named successor, and the renewal lands 90 days out.

Risk to mitigation

What's blocked

The streak repair release is ready, but Apple rejected the build over the new in-app purchase flow, so learners keep losing streaks to the timezone bug it fixes.

Unblock path

Shared screen: what learners face after a weekend away

Verbatim

"I skipped two days on a trip and came back to 140 reviews. It felt like punishment, so I just didn't open the app again for a week."

What's behind it

The review scheduler keeps stacking due cards with no daily cap. A learner who pauses for even a couple of days returns to a wall of reviews, and the session estimate on the home screen makes it look worse than it is.

Summary

Northwind Academy wants Saga for its Spanish program, roughly 180 students across three levels. Priya Rao, the department head, is the primary contact and has expressed strong purchase intent after a two-week informal trial.

Budget and timeline

• Approved language-tools budget of about $24,000 per year
• Decision needed before the fall semester starts in 6 weeks
• Pilot scope: 30 students, starting next month

Decision process

• Dr. Whitfield, the academic director, must sign off on the purchase
• IT review of SSO and student-data handling, estimated 2 weeks
• Priya Rao is the internal champion driving the evaluation

Risks and blockers

• SSO integration and student-privacy compliance are mandatory
• Teachers need a class progress dashboard, currently in beta
• A competing tool is already licensed, so we must show clear gains in the pilot

The objection

Elena Marsh, curriculum director at Brightpath Schools, pushed back on per-seat pricing. Their students rotate through language electives each term, so only about a third of licensed seats are active at any time. Paying for 3,000 seats to serve 1,000 concurrent learners feels like paying for shelf space.

Why it matters

• Brightpath is the largest district deal in the Q3 pipeline
• Elena said the board will not approve a per-seat contract at this ratio
• A competitor has offered her an active-learner pricing model

Response given on the call

Noah acknowledged the rotation pattern and floated a concurrent-seat tier as a possible fit, but did not commit to numbers. Elena agreed to share last year's enrollment data so we can model both options.

Follow-up

• Get concurrent-seat tier approved or rejected by pricing council this week
• Send Elena a side-by-side model once her enrollment data arrives

What was said

Priya Rao, a district admin at Northwind Academy, said their pilot committee is also evaluating LingoPeak. Teachers there liked its dashboard showing which students practiced this week and where they plateau. She asked directly whether Saga can show per-student practice at the class level.

Where we stand

• Saga's school reporting shows streaks per class, not per student
• LingoPeak has no listening or podcast practice, our strongest differentiator
• Priya called the dashboard a must-have for the spring rollout

Suggested response

Lead with listening practice and podcast lessons, then confirm the per-student progress view planned for the schools tier. Flag this deal to product so the dashboard timeline reflects a live requirement.

What was said

Priya Rao said "We need Saga live for our Spanish classes before the fall term starts" and then asked "Could we run a pilot with 40 students in one class next month?" She also asked who signs the agreement on our side.

Why this matters

• A firm deadline (fall term) creates natural urgency on their side
• A concrete pilot size shows they have already scoped internal buy-in
• Asking about the agreement suggests procurement is warming up

Suggested next step

Send the pilot proposal this week: 40 seats, one Spanish class, a 4-week evaluation window, and a check-in call at week two. Confirm who at Northwind approves the budget.

Who they are

Priya oversees the language program for about 900 students across Spanish, French, and Mandarin. She is the one who signs off on new tools, but the IT office runs a security review before anything reaches classrooms.

What they care about

• Students plateau outside class, wants practice that carries over to homework
• Per-student progress visibility for teachers, not just aggregate stats
• Keeping total cost inside the existing Mango Languages budget

Constraints to respect

• Solution must support SSO and pass the district's privacy review
• Wary of app stability after a rough trial with another vendor last spring
• Wants a decision before the semester ends in six weeks

How to engage

Lead with the teacher dashboard and a 30-student pilot offer. Keep emails short and concrete, she forwards them straight to her department heads.

What's going on

Priya Rao reports weekly active seats down to 38 percent from 72 at launch, and she can no longer export data for her board report.

Signal to save play

Job statement

When my workweek fills up, I want to keep making visible progress in Spanish in the small gaps I have, so my streak and confidence do not collapse.

From ask to job

What we learned

Across the last eight onboarding interviews, every learner who read a personalized story in their first week was still active a month later. Learners described the stories built from their own vocabulary as the moment Saga 'clicked', while drill-only learners struggled to say why they should come back.

Supporting evidence

• Retention cohort: week-one story readers hold streaks 2.1x longer
• 6 of 8 interviewees named stories as their favorite feature, unprompted
• Drop-off concentrates on day 3 for learners who never opened a story

Why it matters

Stories look like our strongest retention lever, but today they unlock after lesson five. Surfacing the first story earlier in onboarding is the obvious experiment to run next.

Shared screen: where 4 of 6 gave up

Task attempted

Recover a lost streak after skipping a day of lessons.

Observed friction

The streak-repair offer sits behind the flame icon, but participants read the greyed-out flame as "streak gone for good" and never tapped it. Four of six gave up on the home screen without finding the repair flow.

Think-aloud quote

"Oh no, it's grey. I guess my streak is just dead. I'm not going to tap a dead flame."

Recommendation

Surface the repair offer proactively: show a one-tap "Repair your streak" prompt on the first session after a missed day instead of hiding it behind the flame icon.

Shared screen: the audio-only card Ravi muted in public

Quote

"On the bus I keep it on silent. I just guess what the word sounds like and move on."

Observation

Audio-dependent exercises silently fail for public-transit learners. Ravi never saw an error, he just lost the listening practice. A muted-mode fallback, like captions or a tap-to-reveal phonetic hint, could keep these sessions useful.

Prompt

Schedule the daily streak reminder using each learner's device timezone instead of server time. Store the timezone on the profile at app open, compute the send time locally per learner, and skip the reminder if the lesson is already done that day.

Context

All reminders currently fire at 18:00 UTC, so learners in Asia get them after midnight and lose streaks they thought were safe. Support sees a steady stream of streak-loss complaints from these regions.

Constraints

• Reuse the existing push pipeline, no new delivery service
• Fall back to the account's signup timezone when the device value is missing
• One reminder per learner per day, even across timezone changes

Acceptance criteria

• A learner in Tokyo and one in Lisbon both get the reminder at 18:00 local time
• No reminder is sent once the day's lesson is complete
• Travelling across timezones never produces a duplicate reminder

Condition

Learners west of UTC got their streak reminder in the early afternoon, so many missed the evening window and lost the streak.

Root cause trace

Lesson

Any critical generation dependency needs capacity planning, quota thresholds, and a fallback path. Treat vendor quotas like production capacity, not a billing detail.

Why

A single exhausted text-to-speech quota stopped podcast generation for every affected learner, and the team first heard about it from support tickets, not monitoring.

What we changed

• Quota usage alerts fire at 80 percent for every critical vendor
• A fallback speech provider takes over when the primary is out
• New dependencies get a capacity review before launch

Shared screen: day-7 retention as streak repair rolled out

Since last

Shipped streak repair to 10 percent of learners and reviewed the first results in the growth sync. Early numbers look good: about a third of lapsed learners use the repair, and most of them finish a lesson the same day.

Next

• Widen the rollout to 50 percent by Thursday
• Add the repair prompt to the reminder email
• Draft the day-7 retention readout for the next growth sync

Blockers

The reminder email still renders the old streak count for repaired streaks. Waiting on Theo for the template fix before widening the rollout.

What they asked for

A way to export each class's progress as a simple report: lessons completed, streaks, and time spent per student. Today admins click into every student one by one, which does not work for a school with 400 learners.

Why it matters

• Northwind needs the report for their quarterly board review
• Two other pilot schools asked for the same thing this month
• Blocks the renewal conversation scheduled for August

Suggested shape

A per-class CSV export from the admin dashboard covering a selectable date range. A shareable summary view could come later, the CSV alone unblocks them.

What was committed

Mia will revise the first-run screen so a difficulty toggle appears before the listening sample, and redesign the progress ring on screen 2 so it is obvious what it is counting.

Why

In the design review nobody could say what the progress ring measured, and the difficulty toggle was decided as part of the listening-first onboarding flow that ships this month.

Done when

• Difficulty toggle shows before the first listening sample
• Progress ring has a label and counts lesson steps
• Updated flow is in the design file and signed off by design

Context

Learners who open the podcast in their first week churn at twice the rate of those who reach it later, because the episodes assume vocabulary they have not learned yet.

How we decided

The question

The new onboarding opens with a 2-minute listening sample followed by a comprehension check instead of a placement quiz. How should this flow handle learners who start with literally zero vocabulary? Should they even receive a listening sample?

Why it matters

Absolute beginners are the largest signup group. If the first thing they hear is a sample they cannot follow, we risk making them feel lost in the first minute instead of showing them the value of listening-first learning.

Options raised

• Skip the sample for self-declared beginners and go straight to first words
• Play a slower, ultra-short sample built only from cognates
• Keep the sample but frame it as a preview, not a test

What's blocked

The streak repair release is ready, but Apple rejected the build over the new in-app purchase flow, so learners keep losing streaks to the timezone bug it fixes.

Unblock path

Job statement

When my workweek fills up, I want to keep making visible progress in Spanish in the small gaps I have, so my streak and confidence do not collapse.

From ask to job

What we learned

Across the last eight onboarding interviews, every learner who read a personalized story in their first week was still active a month later. Learners described the stories built from their own vocabulary as the moment Saga 'clicked', while drill-only learners struggled to say why they should come back.

Supporting evidence

• Retention cohort: week-one story readers hold streaks 2.1x longer
• 6 of 8 interviewees named stories as their favorite feature, unprompted
• Drop-off concentrates on day 3 for learners who never opened a story

Why it matters

Stories look like our strongest retention lever, but today they unlock after lesson five. Surfacing the first story earlier in onboarding is the obvious experiment to run next.

Shared screen: where 4 of 6 gave up

Task attempted

Recover a lost streak after skipping a day of lessons.

Observed friction

The streak-repair offer sits behind the flame icon, but participants read the greyed-out flame as "streak gone for good" and never tapped it. Four of six gave up on the home screen without finding the repair flow.

Think-aloud quote

"Oh no, it's grey. I guess my streak is just dead. I'm not going to tap a dead flame."

Recommendation

Surface the repair offer proactively: show a one-tap "Repair your streak" prompt on the first session after a missed day instead of hiding it behind the flame icon.

Shared screen: what Amara saw after one missed day

Verbatim

I missed one day because I was sick, and the first thing Saga showed me was a price. It felt like the app was charging me for being human.

What we heard

Amara, a 9-month learner with a 140-day streak, lost it after one missed day. The streak repair offer appeared before any encouragement, which she read as the app monetizing her slip rather than helping her recover. She said it made her want to open the app less, not more.

Signals to watch

• She hesitated when asked if she would recommend Saga to a friend
• Second interview this month describing streak repair as 'punishing'
• She suggested one free repair per month would feel fair

Lesson

Any critical generation dependency needs capacity planning, quota thresholds, and a fallback path. Treat vendor quotas like production capacity, not a billing detail.

Why

A single exhausted text-to-speech quota stopped podcast generation for every affected learner, and the team first heard about it from support tickets, not monitoring.

What we changed

• Quota usage alerts fire at 80 percent for every critical vendor
• A fallback speech provider takes over when the primary is out
• New dependencies get a capacity review before launch

Shared screen: day-7 retention as streak repair rolled out

Since last

Shipped streak repair to 10 percent of learners and reviewed the first results in the growth sync. Early numbers look good: about a third of lapsed learners use the repair, and most of them finish a lesson the same day.

Next

• Widen the rollout to 50 percent by Thursday
• Add the repair prompt to the reminder email
• Draft the day-7 retention readout for the next growth sync

Blockers

The reminder email still renders the old streak count for repaired streaks. Waiting on Theo for the template fix before widening the rollout.

Shared screen: relocating-staff conversational rate against the Q3 target

Supporting quote

"Our goal is 70% of relocating staff conversational by Q3. If we hit that, the renewal is easy." Their L&D lead, Priya Rao, said this in the QBR.

Context

Meridian is moving 140 staff to its new Lisbon hub and rolled out Saga for Portuguese. Weekly active usage has slipped to about 45% from 70% at launch, so hitting the target needs a usage recovery before the 90-day renewal check-in.

What was committed

Mia will revise the first-run screen so a difficulty toggle appears before the listening sample, and redesign the progress ring on screen 2 so it is obvious what it is counting.

Why

In the design review nobody could say what the progress ring measured, and the difficulty toggle was decided as part of the listening-first onboarding flow that ships this month.

Done when

• Difficulty toggle shows before the first listening sample
• Progress ring has a label and counts lesson steps
• Updated flow is in the design file and signed off by design

Description

Priya Rao, the district admin who championed Saga at Northwind Academy, is leaving in two months with no named successor, and the renewal lands 90 days out.

Risk to mitigation

Shared screen: what learners face after a weekend away

Verbatim

"I skipped two days on a trip and came back to 140 reviews. It felt like punishment, so I just didn't open the app again for a week."

What's behind it

The review scheduler keeps stacking due cards with no daily cap. A learner who pauses for even a couple of days returns to a wall of reviews, and the session estimate on the home screen makes it look worse than it is.

Summary

Northwind Academy wants Saga for its Spanish program, roughly 180 students across three levels. Priya Rao, the department head, is the primary contact and has expressed strong purchase intent after a two-week informal trial.

Budget and timeline

• Approved language-tools budget of about $24,000 per year
• Decision needed before the fall semester starts in 6 weeks
• Pilot scope: 30 students, starting next month

Decision process

• Dr. Whitfield, the academic director, must sign off on the purchase
• IT review of SSO and student-data handling, estimated 2 weeks
• Priya Rao is the internal champion driving the evaluation

Risks and blockers

• SSO integration and student-privacy compliance are mandatory
• Teachers need a class progress dashboard, currently in beta
• A competing tool is already licensed, so we must show clear gains in the pilot

The objection

Elena Marsh, curriculum director at Brightpath Schools, pushed back on per-seat pricing. Their students rotate through language electives each term, so only about a third of licensed seats are active at any time. Paying for 3,000 seats to serve 1,000 concurrent learners feels like paying for shelf space.

Why it matters

• Brightpath is the largest district deal in the Q3 pipeline
• Elena said the board will not approve a per-seat contract at this ratio
• A competitor has offered her an active-learner pricing model

Response given on the call

Noah acknowledged the rotation pattern and floated a concurrent-seat tier as a possible fit, but did not commit to numbers. Elena agreed to share last year's enrollment data so we can model both options.

Follow-up

• Get concurrent-seat tier approved or rejected by pricing council this week
• Send Elena a side-by-side model once her enrollment data arrives

What was said

Priya Rao, a district admin at Northwind Academy, said their pilot committee is also evaluating LingoPeak. Teachers there liked its dashboard showing which students practiced this week and where they plateau. She asked directly whether Saga can show per-student practice at the class level.

Where we stand

• Saga's school reporting shows streaks per class, not per student
• LingoPeak has no listening or podcast practice, our strongest differentiator
• Priya called the dashboard a must-have for the spring rollout

Suggested response

Lead with listening practice and podcast lessons, then confirm the per-student progress view planned for the schools tier. Flag this deal to product so the dashboard timeline reflects a live requirement.

What was said

Priya Rao said "We need Saga live for our Spanish classes before the fall term starts" and then asked "Could we run a pilot with 40 students in one class next month?" She also asked who signs the agreement on our side.

Why this matters

• A firm deadline (fall term) creates natural urgency on their side
• A concrete pilot size shows they have already scoped internal buy-in
• Asking about the agreement suggests procurement is warming up

Suggested next step

Send the pilot proposal this week: 40 seats, one Spanish class, a 4-week evaluation window, and a check-in call at week two. Confirm who at Northwind approves the budget.

Who they are

Priya oversees the language program for about 900 students across Spanish, French, and Mandarin. She is the one who signs off on new tools, but the IT office runs a security review before anything reaches classrooms.

What they care about

• Students plateau outside class, wants practice that carries over to homework
• Per-student progress visibility for teachers, not just aggregate stats
• Keeping total cost inside the existing Mango Languages budget

Constraints to respect

• Solution must support SSO and pass the district's privacy review
• Wary of app stability after a rough trial with another vendor last spring
• Wants a decision before the semester ends in six weeks

How to engage

Lead with the teacher dashboard and a 30-student pilot offer. Keep emails short and concrete, she forwards them straight to her department heads.

What's going on

Priya Rao reports weekly active seats down to 38 percent from 72 at launch, and she can no longer export data for her board report.

Signal to save play

Shared screen: what Amara saw after one missed day

Verbatim

I missed one day because I was sick, and the first thing Saga showed me was a price. It felt like the app was charging me for being human.

What we heard

Amara, a 9-month learner with a 140-day streak, lost it after one missed day. The streak repair offer appeared before any encouragement, which she read as the app monetizing her slip rather than helping her recover. She said it made her want to open the app less, not more.

Signals to watch

• She hesitated when asked if she would recommend Saga to a friend
• Second interview this month describing streak repair as 'punishing'
• She suggested one free repair per month would feel fair

Shared screen: the audio-only card Ravi muted in public

Quote

"On the bus I keep it on silent. I just guess what the word sounds like and move on."

Observation

Audio-dependent exercises silently fail for public-transit learners. Ravi never saw an error, he just lost the listening practice. A muted-mode fallback, like captions or a tap-to-reveal phonetic hint, could keep these sessions useful.

Shared screen: relocating-staff conversational rate against the Q3 target

Supporting quote

"Our goal is 70% of relocating staff conversational by Q3. If we hit that, the renewal is easy." Their L&D lead, Priya Rao, said this in the QBR.

Context

Meridian is moving 140 staff to its new Lisbon hub and rolled out Saga for Portuguese. Weekly active usage has slipped to about 45% from 70% at launch, so hitting the target needs a usage recovery before the 90-day renewal check-in.

What they asked for

A way to export each class's progress as a simple report: lessons completed, streaks, and time spent per student. Today admins click into every student one by one, which does not work for a school with 400 learners.

Why it matters

• Northwind needs the report for their quarterly board review
• Two other pilot schools asked for the same thing this month
• Blocks the renewal conversation scheduled for August

Suggested shape

A per-class CSV export from the admin dashboard covering a selectable date range. A shareable summary view could come later, the CSV alone unblocks them.