Data Analytics The Ultimate Guide to Big Data Analytics for

Business, Data Mini - 0011 - Chapter 11

Chapter 8, Behavioural Analytics, using big data analytics to find hidden customer
behaviour patterns. In history, companies have only been able to track some amount of
information about their customers, like phone numbers, names and previous
transactions. We are however now able to track just about every key movement of our
lives and turn this info into data.

Businesses can now access a whole slew of fresh info about their customer base across
numerous decision points, which include their reactions to various stimuli in
advertisements and social media. Businesses have a great deal of opportunities to
understand better how their customers arrive at various decisions, as well as predict
customers' future behaviours, responses to different stimuli. This means that a customer
may not necessarily know why he bought the jacket, the business from which he bought
it will.

New effective techniques enable businesses to actually analyse different data points
over a long period of time of interactions so that they disaggregate how customers make
their decisions. You may know this influx of data is indeed very exciting, but it is also
hard to interpret. Data is meaningless if there is no capacity to interpret it properly, and
this is where behavioural analytics comes in.

So much like using root cause analysis to the patterns in weather, human behaviour can
also be predicted if the right information is available. Behavioural analysis targets to
understand why and how consumers behave the way they do, so that it becomes
possible to predict what they'll probably do when they are offered new situations. As a
behavioural analyst, when you compile and analyse raw data points like tweets, clicks,
as well as online purchases, you can be able to uncover all manner of hidden inner
drivers.

Such important insights will help businesses to understand the approaches to follow in
marketing to various customers in different times and places. Behavioural analytics at
work. Currently, many businesses are utilising these analytics.

As an example, there is a company known as Brillio which worked with a US retailer quite
recently. The company used behavioural analytics to be able to pinpoint the most loyal
sections within their customer base. They tracked millions of people through the years of
transactions and were able to make various useful discoveries.

For instance, they learned that customers who engaged with this particular retailer for
their wedding planning stayed very devoted to the brand over a significant period of
time and became one of the most valuable types of customers they had. The company
indeed utilised this insight. They started investing in and marketing towards customers
who shared the same underlying traits, particularly during the special occasions, and it
brought about a significant revenue uptick and long-standing value.

In another retailer, the company took advantage of a deep learning engine to predict the
customer buying decisions. When customers purchase a pair of shoes or shirt, their brain
sort of makes millions of decisions all at once. The company created what's referred to
as a neural network algorithm for the retailer to help in modelling the many decisions
that take place in the brain when exposed to various forms of stimuli.

This algorithm helped the company map the different tastes of customers, and they were
able to use the map to be able to predict different future behaviour. To give you an
illustration, if a customer preferred a specific pair of red shoes, the algorithm would
automatically predict their affection for a particular shirt. The company's client used this
information to make more effective marketing suggestions that were more targeted.

So where can you start? As you already know, sifting through big data is a big
undertaking, and even though it does seem like it is too much to process, the good thing
is, it can offer a competitive edge to your business. If your business has not yet begun
harnessing big data analytics, to be able to understand customer behaviour, it is never
too late to begin. Just start small, but start now.

Find a good partner. I admit that it can be costly and time-consuming to develop this
capability in-house. The quickest and safest way to plunge into this space is working with
a partner who's got a good grasp of this space and can guide you through it.

Just find an established expert who knows your business well and can advise you on how
to mine data and interpret it to generate the kind of insight and impact you require. Even
if you choose to work with a partner, you'll eventually need to own this capability. You
therefore need to be sure that the journey is strategically building upon your own
abilities.

This is determined during the phase of generating insights after data analysis. Answer
the right questions. Before you dive in to explore your data, you need to determine the
four or three big questions which need to be answered first.

This is important because it is usually easy to let new, seemingly exciting paths to
distract you when you begin to look at the data. The sample questions can be, who are
my favourite customers? How do these customers shop? When you narrow this down,
you will get a clear focus which will assist you to guide the data ecosystem you want to
define. Make it sustainable.

When you begin using data to find the answers to these questions, you might want to
consider how the answers can change over time. As you create your data ecosystem,
you can try to describe a process that can be refreshed and recreated. This way, you'll
be able to respond to the same critical questions in future and also be able to answer
new questions coming up along the way as well.

For instance, when you embark in a major data processing before a project, you need to
ask your teams to record these steps and also keep all the links and sources live. This
allows your data engineering team or partner to execute refreshes in future when there
is a generation of new data. This agility prepares you to manage changes as they occur.

Do not wait until the data is faultless. Last but not least, this is the worst time to be a
perfectionist. There are analyses that could take several years to get right.

Executives usually delay on making decisions that are based on data until the analysis is
completely correct. This can waste valuable resources and time. Instead of just waiting
for an entirely clear result, you need to pay attention to what signals your data is having
and know you are making decisions with a particular level of confidence, which in most
cases is enough to make a no-go, go decision.

In this case, don't forget that it is often riskier not to take risks. If you are ready and
willing to fail quickly, small and cheaply, your company will learn and as a result become
stronger and wiser. For forward thinking folks, big data can bring about larger profits.

You thus need to partner with the best tools and people you can get and dive in. One
real example. Capital One.

This company has for a long time been relying on behavioural data in order to shape its
offerings to various customers. For example, their deal optimisation engine helps to
analyse demographics as well as spending patterns of customers to determine when,
where and how to place offers before people. This leads to increased revenue for the
company and a better experience with the brand for the customers.

They have also created a tech-driven think tank called Capital One Labs in which
employees sort through prospective opportunities such as new modes of banking using
big data. As you ruminate on behavioural analytics, you need to consider something else
that is equally important to note. The need for a more advanced approach to big data to
help with pattern discovery, among other things.

In the recent years, companies have increasingly become engrossed on how to store and
manage big data. How should they architect their enterprise stack best to gain value
from big data in terms of Hadoop, NOSQL, traditional data warehousing and complex
event processing? Should they host our data on the cloud or on-premise? Inasmuch as
these are fair questions to ask, they don't actually get to the centre of why big data is so
much of a big deal as we tend to put it. It is only by using advanced analytics i.e. using
machine learning to be more specific, can organisations be able to tap in to their
abundant experience vein truly and mine it to discover insights automatically and
generate pretty predictive models so as to utilise the entire data that they are capturing.

With this technology, instead of looking into the past for producing reports, your
business can be able to predict what will happen in future by analysing the existing data.
Let's talk more about that in the next chapter.

Transcribed by TurboScribe.ai. Go Unlimited to remove this message.