Flagship programme

Telemetry for Support Leads

Six weeks of app analytics designed for people who still own a queue. English-language live sessions, late Bangkok afternoon, twelve seats.

Ask for the next cohort

Workshop participants around a table with laptops

What you will be able to do

By the close you can open last night’s tickets, name the in-app moments they imply, and write a one-page brief a product manager can argue with. You will also know when not to escalate — which is the rarer skill.

  • Maintain a living event dictionary your floor actually uses
  • Separate crash clusters from complaint clusters
  • Run a 15-minute morning trace before standup
  • Brief engineering without borrowing their slide template
  • Spot tagging schemes that inflate “P1” noise

Modules

  1. Event vocabulary on a live ticket queue

    We start from three redacted tickets you bring. You learn to map verbs, screens, and properties without inventing events that do not exist in your app.

  2. Session paths without stalking users

    When session replay helps a Khmer or Thai checkout fail, and when it is just watching people type. Consent and proportionality sit in this week on purpose.

  3. Crash clusters versus complaint clusters

    Agents describe failures in human language. Crash reporters describe stacks. This module is the translation layer, including what to do when they disagree.

  4. Escalation heat as a product map

    Hour-of-day heat is usually a staffing chart in disguise. You overlay events until the map points at a feature, not a shift.

  5. Briefing engineering without losing the floor

    A one-page format, a 12-minute readout, and a role-play with a sceptical product counterpart (we supply one if yours cannot attend).

  6. Desk rituals: the 15-minute morning trace

    You install the ritual on your own floor mid-programme. We critique what you dropped when Monday got ugly — that failure is part of the grade.

A real limitation

We do not teach you to replace your existing BI stack. If your events are untagged or named like internal jokes, week one will feel like unpaid data cleaning. Some alumni dislike that. We keep it anyway, because skipping it produces pretty briefs about the wrong screens.

From people who sat this syllabus

Module four (escalation heat) was the only analytics class I have taken that admitted most “peak hour” charts are just when Bangkok traffic ends. We rebuilt ours around pay_attempt instead of clock time.

Pimchanok R. · CS lead, payments

I wanted more SQL. They refused, on purpose. If you need warehouse work, this is the wrong room — and they say so in the first hour.

Anonymous client in logistics SaaS

Questions we actually get

Do I need access to Amplitude / Mixpanel / Firebase already?

Yes, some read access. We will not provision a vendor for you. If your company blocks CS from analytics tools, write to us first — a residency sometimes unlocks that faster than a public cohort.

Is this useful if I am an individual contributor, not a lead?

Senior agents do well if they already coach others. Brand-new agents usually find the briefing module too political. Desk Trace is the better first step.

What if my events are a mess?

That is the limitation named above. Week one will be slow. We still take you, provided you can rename or document at least a handful of events during the six weeks. If you cannot change anything, wait.

Can my product manager sit in?

One guest seat for the briefing week is included. They do not get the full workbook. It keeps the room a CS room.

Ask about this programme

Tell us your team size and which analytics tool you can actually open. We reply within two working days.