|
As
differentiation becomes increasingly difficult in a congested
marketplace, vendors are turning to innovative IT solutions
based on cutting-edge artificial intelligence technology
to execute their precision marketing strategies and capture
the minds and hearts of the discerning consumer. By Dr
Kaustubh Chokshi, CEO of Intelligent Business Systems
 India
is today witnessing a consumer boom that''s unprecedented
throughout its vast and glorious history. Steady and sustained
economic growth has slowly but surely roused the burgeoning
and upbeat middle class into uninhibited spending on a
wide range of goods and services. Disposable incomes have
increased, and, in a break from tradition, so has the
willingness of consumers to dispose of those incomes.
In
such a scenario, it is not surprising that vendors are
falling over each other in their eagerness to attract
and retain customers by any conceivable means. From banking
to telecom, organised retail to real estate and beyond,
everyone''s out to attract customers, keep them happy,
and ensure their steadfast loyalty.
Marketing
Myopia
With the fight to retain customers and expand markets,
terms like "customer-centric approach", are
turning out to be adopted endearingly, in the hope belief
that words alone would widen the revenue base and maintain
optimum profitability. One would assume that these intentions
could easily be actioned, given the expanding markets.
If
markets are growing, so is the competition. Moreover,
vendors operating in such a dynamic marketplace need to
constantly respond to the new opportunities and threats
that are continuously emerging. Alliances are being forged
and broken and these further affect the dynamics of the
market. The irony is that organisations in a similar domain
all have similar underlying goals as their competitors;
share a largely overlapping customer base (albeit expanding);
and, operate in the same business and economic environment.
With
the increased complexity that this brings about, business
agility and flexibility take a beating. Creation of new
customer value is difficult, as differentiation of the
organisation and its products is difficult. In their desperation
to stay in the market-share race, most organisations resort
to "brute force" marketing, wherein every existing
and potential customer is bombarded with all possible
cross-selling, up-selling and general promotional offers
under the sun through every conceivable marketing channel.
Not
only is this a costly proposition for the organisation,
it could result in an exodus of confused or irritated
customers, while at the same time diluting the message
intended for the actual target audience.
The
fact of the matter, in most cases, is that customer relationship
management is yet to reach any significant level of sophistication
that would ensure that the "customer-centric approach"
is anything beyond mere imaginary marketing hype.
But,
organisations can continue to operate in this mode only
to their own detriment, and ultimate demise. Unquestionably,
''precision marketing'' - in essence creating the right
product-customer fit — is the biggest challenge being
faced by marketers today. Traditional methodologies and
strategies fall short, because when it comes to the accurate
customer profiling required for precision marketing, they
can only deliver profiling that''s based on demographics;
or other static rules; or, at best, based on historical
spends and loyalty.
Where
AI Comes In
That''s why organisations today are turning to software
that''s based on artificial intelligence (AI) technology,
where a high level of decision automation in arriving
at the right product-customer fit is facilitated. A carefully
designed AI engine, customised to the precise needs of
a client - be it a banking major or a chain of retail
stores — can be trained to offer true decision automation,
for instance in the selection of a target market from
an existing or potential customer base, with razor-sharp
profiling based on a dynamically determinable parameter-set
and action-set.
Data
from the data warehouse is selected and cleaned and goes
through a first level of pre-processing transformation,
during which the data is normalised. Using data mining
techniques, patterns in the data are established. At this
stage the AI engine takes over in order to automate decisions
based on the knowledge generated from the raw data.
Fig 1: From Raw Data to Decision Automation
AI-based
solutions can be designed to perform context-sensitive
customer acquisition, behaviour analysis, cross-selling,
up-selling and retention programmes, thus enabling multi-product,
multi-channel organisations to drive more efficient, cost-effective
and profitable customer interactions.
With
AI software, decisions and predictions can be based on
actual historical data. First the software establishes,
or helps establish, the criteria to make predictions from
existing sets of data. Next, it generates scoring algorithms
for assigning weights to different data characteristics.
Finally,
it segments populations into sub-groups, enabling variable
treatment for each business decision. This is predictive
analytics in action — predicting likely future results
from the patterns found to be prevalent in historical
data. Most AI software has neural network technology at
its core, wherein the system is "trained" using
historical data and expert opinion to detect hidden patterns
and produce the desired output.
Automation
of a decision involves a derivation of the decision criteria
and the actions to be taken when these criteria are met
as well as when they are not met. After the formulation
of criteria and actions, they must be incorporated into
the system in a manner that makes them available to other
relevant modules and systems as well as to non-technical
users via a user-friendly interface.
The
heart of analytic modelling is a mechanism to transform
all variables and equations to code in such a manner that
new transactions provide inputs for further predictive
measurements. In addition, the sequence and timing of
rule execution is of prime importance, in accordance with
inputs and computed data values.
AI
in Action
In simple terms, Artificial Intelligence software facilitates
the process of transformation of the raw data that an
organisation collects from its various operations / sources
into usable information and then converts that information
into smart decisions.
Take
a supermarket, for instance. Several thousand transactions
are recorded at each checkout counter every day. In its
raw form this transactional data provides only basic levels
of information, such as what items were sold, when they
were sold, and the price at which they were sold.
However,
using AI-based software, that raw sales data can be transformed
into actionable information, enabling the supermarket
to gain deeper insights into marketing strategy and consumer
behaviour. How are discounts impacting sales trends? What
are the best-selling items in each department? How does
shelf-positioning affect sales? What trends and patterns
are emerging that might have an impact on future sales?
Which marketing channels are more effective? What product-mix
works best? Obtaining answers to these and other similar
questions could provide valuable insights for fine-tuning
the overall marketing operations.
Armed
with such predictive analysis and decision automation,
the supermarket management can better plan for the future.
By predicting buying trends and consumer behaviour, inventory
control can be totally streamlined. Moreover, insights
into frequently clubbed purchases in the past (such as
potato chips and salsa dips) may enable better layout
of the store shelves to increase revenues. Promotions
that do well in impacting sales in a test location can
be replicated across the chain in order to boost sales
and profitability.
From
the consumer viewpoint, if the entire system is linked
to a loyalty card scheme that is actually based on buying
patterns in the past, then special offers, cross-selling
and up-selling deals can be tailored to the needs of the
individual consumer (or groups of consumers), thus enhancing
customer satisfaction considerably and automatically increasing
yields.
Conclusion
As is evident from the above example, a well-designed
AI-based system can empower an organisation to offer the
right value proposition through the right product mix
to the right consumer at the | 
