Someone recently asked me - "what would you considered an effective personalization?", to which I promptly asked "how would you define what Effective is?"... of course rhetorically.
To me, in it's most abstract sense; Personalization needs to be both
- Relevant to the customer
- And help us achieve a specific business outcome
Again, "relevant" can mean different things and unfortunately is pretty generic. So this is how I visualize the various maturity levels for Personalization:
Using Retail Bank as an example, the above chart categorizes how "relevant" can we attempt to be, when we take personalization to it's extreme mature end.
To me, this demands our data and downstream personalization systems to take into account:
- Who customers are? (profile data)
- How they consume our services? (transactional data)
- What behavior and intent have we captured from their recent interactions? (behavioral data)
This is what I label as "Hyper-Personalization". i.e. the approach where we leverage everything we know about our customers, even their near real-time behavior and use that to optimize our personalization strategy.
Achieving such an end state is no cinch though. We need our data, decisioning and delivery systems to be:
- Centralized
- Support Identity Management (else we'll be personalizing for devices and not people)
- Built to function in real-time (if we hope to power in-moment personalization)
- Should have a channel agnostic perspective, where we solve for all channels (as against approach solution design for a single channel).
- Be respectful of the privacy, consent and local regulation on how data can (or cannot be) collected and used.
So a complete version of the chart maps the datasets on the y-axis, where as we go right on the personalization scale, we have to go up the data food chain (and figure a way to combine, process and use all these datasets):
In, subsequent blogs; I'll cover a framework I use to apply Hyper-Personalization for my customers.
Comments