Utility Gardenhub is a small review desk in Kuala Lumpur. The name is older than the work; the work is newer than most of the send programmes we are asked to read. We take on app analytics for push notification engagement as a craft: the messages, the counts the phones return, and the gap between those two.
Why this subject
Nur Aisyah Hassan spent years inside a Malaysian telco’s messaging unit, where a “good week” on a slide could hide a permission collapse on Android and a quiet, healthy iOS list. The argument in the room was never really about taste. It was about whether anyone had kept the columns apart. The desk exists so that argument can happen with the log on the table, not with a borrowed industry average.
How we work
We ask for exports you already have. We read them in the order the user would have received the alerts. We keep Bahasa and English sends in the same week when that is how the programme actually ran. We will travel to a Klang Valley office for a findings sitting; we will not pretend a remote call is a site visit.
We do not keep a laboratory of sample apps. We do not publish league tables. Credentials in this field are mostly time spent with ugly files, and we would rather show that in the client notes than on a wall of badges.
People
Nur Aisyah Hassan leads reviews and writes the findings notes. She still marks copy with a pencil. She will stop a meeting if someone says “opened” when they mean “delivered.”
Daniel Tan prepares the count tables, splits operating systems when the export allows it, and is the person who notices that a “daily digest” went out four times on a public holiday.
Place
The desk is at Office 12, 31 Example Road, Kuala Lumpur 00000. Findings sittings can be there. Most of the reading happens from the files you send. If your senders are in Johor Bahru, Penang, or entirely overseas, the work still proceeds; the Malaysian festive calendar simply becomes something we have to be told rather than assume.
Relationship with clients
You keep the logs. You keep the customer list. We keep the findings note and the working papers for the retention period described in Privacy. If we cannot read a language fairly, we leave it out. If the sample is too thin, we say wait. That is the whole ethic.