Sharing similar demographics and media behaviour, conventional media planning always led us to BabyKingdom, whether talking to mums-to-be, pregnant women or super-premium parents. It was almost like saying different things to a “triplet” in a particular “room”, but being unable to tell who’s who.

However, from consumer studies, CRM and past campaign learnings, we knew 3 segments show different interest level on different parental information and have different habits in their life. We leverage on this 1% uniqueness to identify them, not only to give the most relevant message but also to trace them even as they walked out of the “room”. In the long run, we wanted to adapt messages from different product as they travel along the parental journey, to achieve long term business goal.

Creatively turning audience understanding into relevant messaging adapted to their preferences, digital behavior & lifestyles, we identified a few steps to take:

1. Create an audience pool with data and segment by their needs.

2. Base on audiences’ need states to serve product messages that are relevant

3. Move consumers along the purchase funnel especially driving product trials

4. Synergize marketing investment with cross brand learnings and optimize plans for better conversions

Once we have an initial idea of who the audiences is, we started feeding messages which we believe to be the most relevant. By measuring the real time response of each creative, machine learnings enable us to further fine tune the way we personify audiences, eventually helping us to predict the nature of next audiences who “look alike”. With a more accurate audience profiling, individual brands are able to optimize their spending, focusing only on their core target audiences and formulate more consumer centric social media content strategy.

We approached Google’s giant database by singling out those visited Baby Kingdom or searched for relevant information, while applying specific filter for each type audiences. Audiences are identified by Cookies and UDID. As they visited any website falling into our programmatic network, different creatives of respective products will be served based on their stage of consumer journey (Awareness, Consideration and Goal).

Looking beyond the conventional media matrix, we can now expand our KPIs to consider “precision” of our targeting and taking away “media placement” as a variable to evaluate each creatives fairly. Since the launch the big data initiative in 2016 Q2:

Identified 198K pregnant mum & 272K mass mum respectively. Machine learnings allow us to expand our audience pool and engage even more unique quality prospect with “look alike” characteristics.

Going beyond short term campaign performance, we believe Big Data could build a foundation for long term business success. But most of all, by delivering information that is more relevant, it fulfils our mission to make parental journey more enjoyable!