Southeast Asian consumers are turning to generative AI to discover products, with one in four (24%) now using it as part of their shopping journey. That adds another layer to a journey that may already span creators, social commerce, retail media and marketplaces before a purchase is made.
A consumer might discover a product through an AI assistant, sees it again in a creator's video, compares it on a marketplace, then buys it through an app days later. By the time the transaction happens, several channels may have already shaped the decision.
For marketers, the challenge is not simply keeping up with another discovery channel. It is understanding how these interactions work together, and whether the customers they bring in actually create value. Traditional attribution wasn't built to answer that.
The measurement gap is getting wider
Last-click attribution tells marketers where a conversion happened but not necessarily what shaped the decision to buy. In today's consumer shopping journey, the app purchase is easy to measure. The AI recommendation, the creator review and the price comparison that built the intent to buy aren't.
The problem isn't that the last click is irrelevant. It's that treating it as the whole story can overvalue the channel closest to conversion while undervaluing the ones that built the demand in the first place.
The stakes are higher in a region where mobile commerce keeps expanding. Singapore's eCommerce app installs grew 67% year on year in the first half of 2026, which means more installs mean more acquisition opportunities, and more pressure to understand what happens to those users after they download.
A campaign can look efficient when judged by install volume alone. That efficiency can evaporate once marketers see one group of users coming back to buy again while another disappears after a single transaction.
The question isn't simply where a customer converted, but what made that customer worth acquiring in the first place.That means looking at acquisition, engagement and retention together rather than as three separate reports.
Building measurement for a fragmented journey
The answer is not to give every touchpoint equal credit. It is to build enough visibility across the journey to understand which interactions consistently contribute to valuable customers.
That starts with the data marketers already collect. Acquisition numbers mean something different once they're read against engagement, retention and repeat purchase. A channel delivering cheap installs can look far less attractive once those users are tracked past day one.
It also means building a measurement approach that can absorb new discovery sources without a rebuild every time consumer behaviour shifts. AI recommendations won't behave like a conventional click or referral, and neither will whatever replaces them next. A separate measurement model for each new channel adds complexity without adding insight. What holds up instead: identify where the customer came from, connect that interaction to conversion, then follow the customer past the first purchase.
That discipline matters more as marketers cut the number of channels they rely on. Singapore, Vietnam, Thailand and Indonesia all reduced their number of ad partners in the first half of 2026. Fewer channels means each spending decision carries more weight, which makes the data behind it harder to get wrong.
The same logic applies to AI and creator-led discovery. Marketers don't need to prove that every touchpoint caused a sale. They need enough evidence to tell which channels are bringing in customers who stay from which ones are just bringing in installs.
That is where attribution becomes more than a reporting function. It becomes part of how marketers decide where to invest next.
Channels will keep multiplying and consumer behaviour will keep moving between them. Whether marketers can still see that movement, across every touchpoint that shaped it, will determine how well they can distinguish customer acquisition from simply acquiring installs.