E-Commerce and market intelligence
Chen, Chiang and Storey (2012:1165) consider business
intelligence and analytics as an important area of study
and research to solve data-related problems in companies.
This is vividly illustrated by what web and e-commerce
vendors have developed and implemented over the last
few years (see introduction of section 3 and McAfee and
Brynjolfsson 2012:61). Amazon, amongst many similar
internet-based companies, is an excellent example of this,
where they have developed analytics tools (algorithms)
to analyse every ‘click’ on their website by thousands,
possibly millions, of customers browsing for and buying
products. This has enabled Amazon to, for example, not
only propose customer preferences for certain products
but also to manage inventories at their various distribution
centres by tracking sales of different types of products and
how that impacts inventories at the distribution centres. Just
recently Amazon revealed its plans for predictive shipping,
whereby consumer-generated information is scanned and a
sophisticated analytics algorithm aims to optimise fulfilment
strategies.
Therefore, instead of simply analysing sales, predictive
analytics provide retailers with a view into the future and
an opportunity to identify patterns that lead to effective and
highly personalised customer engagement strategies. Where
descriptive analytics measures what has already happened,
predictive analytics applies statistical modelling and data
mining to study recent and historical data, thus allowing for
more accurate forecasting.
Loyalty cards are another example of the use of big data and
analytics that has major applications for product ranges,
inventories, etc., as well as market intelligence. For each
item bought at a retail store using loyalty cards, the store can
through proper analysis of the point-of-sale data determine
exactly what items are bought, what the preferences of loyalty
card holders are, the size of purchases, etc., and what the
implications are for different products, product inventory for
different times of the month/year (introducing seasonality),
order quantities and when to order, which stores are more
profitable, and which stores are solicited more often, leading
to better location decisions, etc. Marketing initiatives can be
guided significantly by analysis of the big data obtained in
this manner.