What the devil is Hyper-Personalization? Read these first:
Building on the previous articles, how could one go about envisioning an automated approach to shift from?:
1. Business Objective -> Segmentation Strategy -> Proposition to Customer
To
2. Customer Interaction -> Automated Decision (based on context, predictive and historic data) -> Proposition to customer
In both cases, we have an "outcome" which is a "Proposition to Customer"; however, the latter is expected to be significantly more focused and relevant to each individual; and as a result fetch a higher success rate.
Those with keen eyes and sharp brains would argue that I'm comparing "batch" or "outbound" activities with "interaction" type of activities (1 vs 2). And that most mature Banks (& other mature enterprises) have some way to deliver in-moment personalization as well.
The case I'm making here is not either / or, but a case for both. It's case for a vision to apply our personalization decisions in a consistent fashion, no matter the channel, or who initiates the interaction (marketer or the customer).
Banks & Financial institutions have some of the most complex and mature technology & data eco-systems, considering the banking operational work loads. To further complicate; there is the aspect of data security, compliance, governance and consent to worry about.
Often customer engagement and personalization decisions start off for a specific channel, built around use cases. And several such initiatives mushroom over time.
What we end up with is several isolated approaches (and technologies) tied to channels. These unfortunately, also end up shaping internal processes and force silos.
How do we attempt to fix this?
For starters, we've got to fundamentally solve for the foundation, irrespective of whether we can have centralized decisioning and delivery platforms, or federated ones.
Right on top, I'd start with Data & Identity Management. Any attempt for meaningful personalization goes out of the window, if we cannot link data and identities (we'd be personalizing for devices and not people).
Then, there is the aspect of Latency - i.e. how quickly can we centralize profiles, events, interactions etc. to allow business teams and downstream systems to act on incoming information? Some questions to ponder on:
- Do we need to solve for real-time scenarios? Yes, but do we pre-compute outcomes or support truly dynamic responses?
- What's the shelf life of data? (yes, unlike profile data, intent data may not be relevant after a while for driving personalization - more on this topic soon.)
Finally the aspect of Data Classification, Access Control & Governance; to support workstreams without compromising on regulatory and moral obligations.
And so on...
On the process bit, I often encourage my customers to think of marketing to "intent" or "likelihood" than "customers". So in an automated eco-system, our credit card, loan, deposit etc. teams would publish the desire to market to specific intent (or categories of intent) along with say data science based likelihood factors. And our beautiful automated systems would then map and prioritize our marketing promotions, to end customers on the fly.
#technology #blueprint #architecture #data #enterprise #FSI #Banking #personalization #hyperpersonalization #fintech
Comments