Page 84 - Building Big Data Applications

Chapter 3 Building big data applications 79

Analytics needs to happen now and later. This requires models that can execute with
speed, an acceptable margin of error, acceptable ranges of recommendations for dis-
counts, and competitive offers. The analytics needed to be not geared to prescriptive
but rather an introspective.
What these changes to data and analytics drive us to implement is the change of
how to market, where the generation that needs to be sold to today wants offers at
the time of their shopping not before, which means you need to be beyond the
thought of agile marketing. Here the concept to delivery of a campaign is in the
duration of the browsing and is tailored to the customer preferences, which are
collected based on sentiments and user behavior analysis, the socio-economic models
of the zip codes where the customers reside. These factors require us to not execute
segments and outliers but rather create buckets based on specific criteria and then
validate which buckets align closest to the market at the time of execution. This
complexity is very tough to resolve and takes companies time and effort. But this is
trick to resolve to be successful.
The process of complexity does not end here; we need to have the optimized supply
chain process and inventory management process, which will ensure that when the
customer places an order, the entire order will be shipped and delivered as promised
to the customer on that date and not have delays. This means your store in the
ecommerce model of business is connected with distribution centers which will have
access to inventory and shippable information ready. The model further changes
when you see back orders for a popular item and this means you will need to have
creative marketing strategies and offers to induce the conversion to become an order.
There are several complexities in integrating the data to make this happen seam-
lessly. Today when you look at Amazon or Baidu they have invested time, energy,
and effort in creating these platforms. However, they also have a large workforce
that is available to manage all the processes, and this is where we will see the evolu-
tion of process automation and integration.
As we look at the governance mechanisms and strategies, the ecommerce world has
added more transparency within the enterprise that requires us to ensure that
All appropriate users can see data on demand
All appropriate users can execute analytics on demand
All data is secure
All customer information is secure and privacy is 100% enabled
All product details are disclosed as required for each sale
All complaints are handled by a 24 7 365 call center
All languages are supported in the call center
All data is digital
All processes are digital
All complexity is documented and formulated
All decisions are available to be replayed as needed to learn and train.