DEV Community

Cover image for From a 7-day inactivity list to a tutor queue: triaging silent students in WhatsApp cohorts
9bot
9bot

Posted on Originally published at 9bot.com.br

From a 7-day inactivity list to a tutor queue: triaging silent students in WhatsApp cohorts

Dropout in a paid online course rarely announces itself. The student stops replying, stops reacting to polls, and a few weeks later the access period ends with a certificate nobody claimed. If you run cohorts on WhatsApp, the earliest machine readable symptom you have is silence in the group. This post is about turning that symptom into a queue a human can work through every week.

Context and the full retention playbook are in the 9bot guide to keeping a student community moving. Quick disclosure: 9bot is a paid tool (there is a seven day trial), built independently of Meta, and WhatsApp does not sponsor or vouch for it.

The signal you actually get

9bot Reports expose, per group, the members who posted nothing in the last seven days. That is the raw input. It is cheap, it is weekly, and it is noisy, because a quiet student may be watching every lesson on the course platform. So the first rule of the queue is: never act on group silence alone. Join it with platform progress before a tutor spends time on anyone.

Triage tiers

Tier Group signal Platform signal Action Owner
T0 Silent 7 days Progress moving No contact; keep in the weekly list Nobody
T1 Silent 7 days Progress stalled Post an easy milestone in the group (a ten minute exercise) plus a poll on what is blocking people Community manager
T2 Still silent after the group nudge Still stalled Private message from the tutor Tutor
T3 Any Access period close to ending Notice before removal, with extension or renewal option Time based Access Control, then commercial team

Tiers T1 and T3 can run on configuration. T2 is deliberately manual: 9bot does not write privately to inactive students. It handles messages inside the group, commands, moderation, reports and timed access. The private conversation is where a person is worth the cost, so the design keeps it human.

Engagement data from 9bot feeding cohort decisions

A queue record

Here is the shape we suggest for the weekly hand off between the report and the tutor. It is a planning schema you can keep in a spreadsheet or your own tooling, not an API response from 9bot.

{
  "week": "2027-W03",
  "cohort_group": "sales-ops-bootcamp-b",
  "source": "9bot Reports: inactive members (7 days)",
  "joined_with": "course platform progress export",
  "items": [
    {
      "member_ref": "student-0142",
      "tier": "T1",
      "group_nudge_sent": "2027-01-18T19:00",
      "platform_progress": "stalled since module 2"
    },
    {
      "member_ref": "student-0177",
      "tier": "T2",
      "assigned_tutor": "tutor-ana",
      "contact_channel": "private, manual",
      "due": "2027-01-20"
    },
    {
      "member_ref": "student-0203",
      "tier": "T3",
      "access_ends": "2027-02-01",
      "notice": "sent by Time based Access Control"
    }
  ]
}
Enter fullscreen mode Exit fullscreen mode

Notice that member_ref is a pseudonymous key. Phone numbers do not need to travel into a tutor sheet. A paper on LGPD and GDPR duties in the 9bot research series points out that the main exposure inside a group is lateral, because members already see each other's numbers. Do not add a second leak in your own tooling.

Why the group nudge comes before the private message

The 2024 Frontiers in Education study cited by the article (309 students near graduation, in groups their institution administered) found reciprocity, information quality and service quality associated with satisfaction, and satisfaction most associated with intention to continue. A group level nudge that invites people to share their result is a reciprocity mechanism. A private message is a service mechanism. Running the cheap, social one first and the expensive, personal one second is a sensible ordering, though the study measured intention and not completion, so treat it as a design choice you validate.

Running the loop weekly

  1. Pull the inactive list for each cohort group.
  2. Join it with platform progress; drop T0.
  3. Fire the T1 nudge in the group on the scheduled slot.
  4. Seven days later, promote anyone still silent and stalled to T2 and assign a tutor.
  5. Check access end dates; confirm the notice for T3.
  6. Record exits before the end of the cohort, also visible in Reports.

What to measure

Completion rate lives on the course platform: finishers over enrolled. Replies to a milestone are counted in the group and divided by the cohort size. Compare cohorts that ran the queue with cohorts that did not. That is the only evidence that will tell your team whether the effort pays back.

Prices for the plan are on the article page (listed in US dollars, dated 30 October 2026, with a coupon at checkout), so read them there instead of from a copy.

Top comments (0)