Essay · Kamonchanok Prasert
Why tap paths break after the third screen
Most Bangkok product reviews still open with add-to-cart. It is a flattering number. It is also, on mobile, often the last honest step in the journey. By the time a buyer has crossed three distinct screens — say product, cart, and shipping — the path has accumulated load that no single conversion rate will confess.
In Path Signal Mapping we ask students to stop the recording at the third screen and list every demand the interface has already made: script switches, permission prompts, address fields that mix Thai and Latin, a coupon box that steals focus, a sticky bar covering the primary tap. Teams are routinely surprised by the length of that list. They had been calling the same sequence “a simple checkout.”
Add-to-cart is not a path
Add-to-cart is a gate. It tells you someone was willing to start. It does not tell you whether the next two frames were readable on a phone in traffic on Vibhavadi, whether the Thai labels wrapped under English placeholders, or whether the keyboard hopped from Kedmanee to QWERTY three times before the buyer gave the screen back to a messaging app.
When we overlay session order on warehouse events, the silent exits after screen three often never fire a “begin checkout” event at all. The product tool therefore reports a healthy cart and a mysterious paid gap. Leadership argues about payment partners. The path, if anyone watches it, argues about wrapping text and a province picker.
Keyboard switching is not localisation trivia
Thai buyers writing a Latin email, then a Thai street name, then a numeric postcode, pay a tax that Western funnel lore barely names. Each switch is a hesitation. Each hesitation is a chance to notice a better use of the next thirty seconds. Mobile Conversion Path Analytics treats that tax as instrumentation. If your events cannot see keyboard changes, your model is incomplete for this market.
We do not claim that removing a field always helps. Sometimes a longer Thai-script label that does not wrap saves more time than a shorter English one that forces a guess. The third screen is where those trade-offs become visible, because working memory is already busy.
What to do on Monday
Sit with ten recordings that died after cart. Do not average them yet. Mark the third screen only. If eight of ten share the same wrapping, picker, or keyboard tax, you have a path claim worth taking to Friday review. If they do not share anything, you do not yet have a story — and inventing one from the add-to-cart chart will waste a sprint.
This is slower than a dashboard tour. It is also how the studio keeps Mobile Conversion Path Analytics from becoming another synonym for “look at Mixpanel.” The third screen is usually where the honest work starts.