Bear (Boon Hui Chen) Hi, I’m Bear · Kuala Lumpur

Lab: what I’m testing right now.

This is my open notebook. Each experiment says what I’m trying, what stage it’s at and what happened on which day. I update it when something changes, not on a schedule.

Nothing here counts as a result until an entry shows the numbers. If a test fails, that stays on the page too.

the log · newest first1 experiment running · last entry 13 Oct 2026

Running14-day runStage 1 of 4Started 13 Oct 2026

One landing page per ad angle

Brands often run many ads with different angles, then send every click to the same page. I’m testing whether a page built for one ad, with the same headline, offer and image and a single next step, lowers the cost per lead.

Diagram. Now: three ads with different angles, price, before and after, and gift idea, all lead to the same page. The test: each ad leads to its own page that repeats its angle, and the old page and the matched page are compared in a 50/50 split test.
The idea in one picture. Now: how a lot of brands do it. The test: what I’m trying.

The plan 4 stages

  1. 1

    Build a list of 30 brands from the public Meta Ad Library.

    Now
  2. 2

    Build one page template and 5 sample pages.

    Next
  3. 3

    Show the sample pages to the brands.

    Later
  4. 4

    Run one paid 2-week split test and publish the numbers, good or bad.

    Later

Log newest first

  1. Started. Stage 1: I’m building the list of 30 brands from the public Meta Ad Library. No results yet.

Who it’s for
Malaysian brands with their own website and 30 or more live Meta ads.
How I’ll know
A 50/50 split test on the same traffic: the current page against the matched page.
Results so far
None. No client for this yet, and no numbers. When there are numbers they go here, good or bad.