Now Reading
Forget Big Data. We need “Clean & Structured” Data

Forget Big Data. We need “Clean & Structured” Data


Much has been said in the last couple of years about the explosion of ‘Big Data’. The immense opportunities that ‘Big data’ can provide to a multitude of industries ranging from Telecom to Retail, CPG to Energy & Transportation. For once, we all agree that Data is the new common currency for identifying opportunities for incremental product / service growth. Failure to capitalize on such rich information of data can be threatening to organizations, not as a sum of parts but as a whole.

Having travelled around APAC extensively for a while, I have come to a rather starling (and not surprising to those of you into solution sales in the region) conclusion: The Key Challenge is not yet to capitalize on the ocean of User Generated Content but to possess a structured approach to capture data: not including textual information or any semi structured data. We are talking about “Structured” data: Data that Banks’ Billing systems possess and Core Banking operations, Data that Customers provide with relation to Demographics and Location, Data that Telco’s capture in their BTS & IN systems, Data that logistics firms capture on their CRM.



Imagine this Scenario: I am fortunate to get an opportunity to meet a C-Suite Marketing gentleman in one of the largest on-line Money remittance to scope their existing analytical capabilities. I had presented the challenges that arise from mass marketing and 1-1 Marketing is ‘Way to Go’. And after an hour of pretty good presentation, he says ‘Well! This looks good…But how do I identify unique customers from a plethora of customer ID’s?’ I was taken aback – I assumed that his question might be aligned with something related to ETL & Data Aggregation but what he really meant was:

I do not have a system to identify duplicates in my customer base and therefore: How can I be sure I am not targeting the same customer more than once.

This is not a peculiar problem. Consider this case: A large top 10 Asian Bank had planned to embark on a Analytics and Campaign Management solution. I was fortunate to design an integrated framework to address the bank’s Analytics and Marketing objectives and drive implementation. I had assumed that the bank relatively proactive to address customers’ requirements had a Analytics Datamart as part of the “Single Source of Truth”. In reality, the bank didn’t have the structure in place. Also there was no way to automate the workflow to generate reports on who was getting what campaign and the existing process was entirely manual driven ( script based). How was the marketing team supposed to leverage such a siloed structure to scale on shifts in customer behavior?

In a nutshell, there can be 2 major types of issues companies already face today and are taking steps to plug the gaps:

  1. Incomplete data arising from a semi-structural process and

    See Also
    Alexa

  2. Siloed data across multiple data sources

Although the thought of interacting with customers with an omni-channel perspective is a great strategy (and not new nonetheless), we must not under sell the structural gaps arising out of implementing such a strategy. The complexity is even more when understanding the design of data structures when considering machine learning and unstructured data from sources as diverse as RFID’s and Sensors. We must not only look at suggesting solutions but also guiding clients from the start on what data structure and data cleaning activities are imperative to scale and sustain their overarching objectives.  


Enjoyed this story? Join our Telegram group. And be part of an engaging community.

Provide your comments below

comments

What's Your Reaction?
Excited
0
Happy
0
In Love
0
Not Sure
0
Silly
0
View Comments (0)

Leave a Reply

Scroll To Top