2024 — now04/04
growth manager
leading a digital analytics & growth practice. data lifecycle: instrument → read → write down what ships next. mostly the writing.
observed//growth manager · digital analyst//jakarta · GMT+7
data is noise. signal is the work.
n =0,000samples
↓ densitymagnitude →
move disturbs the pile·click rain
9years
20+enterprise accounts·50+dashboards·2,400in this pile
scroll for receipts↓
— 02 / thesis
growth work is observation before execution.
The chart is a sketch of the question. The headline is the conclusion drawn beneath. Nine years means I’ve watched a lot of curves refuse to bend until exactly the right thing was changed — and most of the time, the thing was small.
03 / receipts
the numbers, plain.
01years
9
on the practice
from 2016 →
02accounts
20+
enterprise relationships
retail · F&B · telco · healthcare
03dashboards
50+
built / archived / rebuilt
mostly rebuilt
04 / history
four positions. one direction.
2021 — 202403/04
lead digital analyst
ran enterprise analytics across various industries. built the practice's experimentation rigor — p < 0.01 became the floor, not the goal.
2018 — 202102/04
digital analyst
in the weeds: tracking implementations, custom dashboards, the early lifecycle work that became the loyalty thesis. 50+ dashboards started here.
2016.08 — 201801/04
digital analyst
first real fluency. retail e-commerce, indonesia. spent a year watching where attention dies. still the most useful year.
before00/04
—
whatever came before was practice for sitting still long enough to see the shape.
01 / 05
← drag horizontal · or scroll →
05 / speaking & advisory
what i’ll talk about, who i’ll sit with.
talk+45 min
cohort retention without the magic.
how to read a retention curve honestly. why most "uplift" disappears in the second cohort, and what to ask before believing the first.
▓▓▓░░░ ▓▓░░░░ ▓░░░░░
workshop2-day
building the dashboard you'd actually use.
a working session for analyst teams. by the end you have one dashboard you'd defend, and a list of four you'd kill.
┌──┐ ┌─┐ │▓▓│ │▓│ └──┘ └─┘
advisory3-month engagements
growth stack & experimentation discipline.
for series A/B teams who keep declaring wins and missing the next quarter. we'll instrument, read, and write down what to ship.
· ·· · ··* ·· ·· ·
fireside30 min
nine years of getting the same lesson again.
the personal version. what the loyalty work actually taught me, what i wish i'd known in year two, and the one chart i still haven't beaten.
/\ /\ \_\/_/ ··
agenticdeep dive
agentic workflows, from someone down the rabbit hole.
i go deep on AI — agents, CLI-first tooling, Claude Code, the whole stack. how to wire an analyst loop that ships work without you in the room. live builds, not slides.
▶ ▶ ▶ ▷▶▷ ▶ ▷ ▶
open15 min
just say hi.
if you're a junior analyst in jakarta or south-east asia, the calendar is open. no pitch, no agenda. ask the question.
→ → → → → →
06 / about
who’s writing this.
Faiszal reads data for a living. mostly looks at the same kind of page for nine years.
Growth Manager at Antikode in Jakarta — across 20+ enterprise accounts, 50+ dashboards, and about ~15% retention uplift on the work that started this career.
Work loop is small: instrument the data, sit with what it actually says, write down the next thing a team should ship. The instrumenting is craft; the sitting is the job; the writing is the deliverable.
Currently most interested in: AI agents that ship analyst work, CLI-first workflows over chat windows, and the smallest loop a team can run on Claude Code without me in the room.
GA4LookerMetabaseSQLPythonR · sometimesClaude Coden8n