What Is Predictive Analytics?

1
Module -5 Predictive analysis 9 hours

Meaning of predictive analysis, how good are models at predictive behavior, benefits of
predictive models and applications of predictive analysis, reaping the benefits, avoiding the
pitfalls, importance of predictive model, process of predictive analytics. Predictive
Analytics, Data Mining and Big Data_ Myths, Misconceptions and Methods by Steven
Finlay.
__________________________________________________________________________
What Is Predictive Analytics?
Predictive analytics helps businesses look into the future and peer around corners with a
reasonable degree of accuracy. This capability has always been important – but it has never
been as critical as it is right now. Companies have had to navigate major trade and supply
chain disruptions, sudden spikes (or nosedives) in demand, brand new risks and challenges,
and overall unchartered waters. That’s why predictive analytics has shot to the top of priority
lists for organizations around the world.
Predictive analytics definition
Predictive analytics is a branch of advanced analytics that makes predictions about future
events, behaviors, and outcomes. It uses statistical techniques – including machine
learning algorithms and sophisticated predictive modeling – to analyze current and historical
data and assess the likelihood that something will take place, even if that something isn’t on a
business’ radar.
Predictive analytics is relevant to most industries and has myriad uses, including:
 Reducing employee and customer churn
 Identifying customers who are most likely to default on payments
 Supporting data-based sales forecasting
 Setting optimal pricing
Prof. Manjunatha S, Assistant Professor, Department of Management Studies, Sai Vidya Institute of Technology, Bengaluru-64.
2
 Tracking when machines will need maintenance or replacement
Actionable, accurate predictions are essential in helping decision-makers navigate a world

where rapid change and market volatility are constants. And while that was true before
COVID-19, the ability to pivot and forecast and plan for multiple possible scenarios is more
critical than ever during the pandemic.
Predictive analytics is also being used in the fight against COVID-19. Hospitals and health
systems use predictive models to gauge risk, predict disease outcomes, and manage supply
chains for medical equipment and PPE. In turn, researchers are using models to map the
spread of the virus, predict case numbers, and manage contact tracing, all with the goal of
reducing infection numbers and deaths.
Predictive analytics, as shown above, can help businesses anticipate cash flow.
Predictive analytics today

According to a study from Allied Market Research, the global predictive analytics market is
projected to reach US$35.45 billion by 2027, growing at a compound annual growth rate
(CAGR) of 21.9%. Predictive analytics has truly come into its own in today’s world, where
massive amounts of data are being generated, computers have exponentially faster processing
power, and software has become more interactive and easier to use.
3
Companies not only collect huge volumes of data, they collect many different types – from
traditional structured data to unstructured data like Internet of Things (IoT), text, video, and
dark data. The ability for predictive analytics to combine and analyze Big Data from different
sources produces more accurate forecasts and surfaces insights that are deeper and more
powerful. The cloud is key for connecting all these different data sources – plus, storing data
in cloud-based data warehouses and lakes is more cost-effective and more scalable than
storing it on premise.
Today’s predictive analytics is also “augmented” with artificial intelligence (AI) technologies
like machine learning, deep learning, and neural networks. These augmented analytics can
analyze large volumes of data quickly, reveal insights that humans might miss, and make
predicting the likelihood of future events more nuanced and more accurate. They also
automate complicated steps in the predictive analytics process, such as building and testing
predictive models. And natural language processing (NLP), a type of AI that lets users ask
questions and get answers in conversational language, makes interpreting and understanding
these answers easier than ever.
Historically, the tools and techniques behind predictive analytics have been so sophisticated –
and so complicated – that only data scientists and professional analysts have been able to use
them effectively. But with augmented analytics, business users with minimal training are now
able to generate accurate predictions and make smart, forward-looking decisions without help
from IT – an advantage that can’t be ignored in a fiercely competitive market.
Examples of predictive analytics
Predictive analytics is applicable and valuable to nearly every industry – from financial
services to aerospace. Predictive models are used for forecasting inventory, managing
resources, setting ticket prices, managing equipment maintenance, developing credit risk
models, and much more. They help companies reduce risks, optimize operations, and increase
revenue.
Predictive analytics in HR
HR is a field that naturally tracks a large quantity of people data. With predictive analytics,
that data can be analyzed to determine if a potential employee is likely to be a cultural fit,
4
which employees are at risk of leaving an organization (shown below), whether a company
needs to upskill an employee or hire to fill skills gaps, and if employees are productively
contributing to business outcomes. These abilities mean that HR can contribute to overall
business outcomes rather than act as an isolated function.
Predictive analytics in HR can be used to predict employee churn.

Predictive analytics in healthcare
In today’s world, hospitals and healthcare organizations are under immense pressure to
maximize resources – and predictive analytics makes that possible. Using predictive
analytics, healthcare officials can improve financial and operational decision-making,
optimize inventory and staffing levels, manage their supply chains more efficiently, and
predict maintenance needs for medical equipment. Predictive analytics also makes it possible
to improve clinical outcomes by detecting early signs of patient deterioration, identifying
patients at risk for readmission, and improving the accuracy of patient diagnosis and
treatment.
5
Predictive analytics in retail
Retailers gather vast amounts of customer information both online, such as tracking online
activity via cookies, and in the real world, such as monitoring how customers navigate their
way through a store. Other information tracked includes customers’ contact details at the
point of sale, their social media activity, what they’ve purchased, and how often they
purchase specific items or visit a store. Using predictive analytics, retailers can leverage that
data for everything, from inventory optimization and revenue forecasting to behavior
analytics, shopper targeting, and fraud detection.
Predictive analytics in marketing
The models generated by predictive analytics are extremely valuable for marketers in making
their campaigns more targeted and effective in a world where customers can order what they
want, when they want, from almost anywhere online. Predictive marketing analytics drives
data-driven customer and audience segmentation, new customer acquisition, lead scoring,
content and ad recommendations, and hyper-personalization. Marketers can use a customer’s
data to feed them promotions, ad campaigns, and suggestions for other products they may
like at just the right time, improving customer experience and retention.
Predictive analytics in supply chain
Predictive analytics has become essential for running an agile, resilient supply chain and
avoiding disruption. It analyzes massive data sets from many different sources to generate
accurate supply and demand forecasts, determine optimal inventory levels, improve logistics
and on-time deliveries, predict equipment maintenance issues, detect and adapt to unexpected
conditions – and much more.
Companies using predictive analytics

AG Real Estate: AG Real Estate, Belgium’s largest real estate group, uses predictive
analytics to predict shopper behavior more accurately. By combining data from sources such
as mall cameras, weather forecasts, and their financials, they are able to improve their
customers’ experiences and keep their shopping centers competitive in the e-commerce era.
6
Rainforest Connection: Up to 90% of logging in the rainforest is illegal. Rainforest

Connection is working to stop illegal deforestation by using old cellphones positioned in trees
to capture audio data – and accurately predicting when illegal logging will start just by
analyzing the sounds the animals make.
Basic steps in the predictive analytics process

The predictive analytics process involves defining a goal or objective, collecting and cleaning
massive amounts of data, and then building predictive models using sophisticated predictive
algorithms and techniques. This traditionally complex process is becoming more automated
and more accessible to the average business user thanks to new AI technologies, but
companies may still need IT to help in certain steps or to build certain models.
In very simple terms, the steps in the predictive analytics process are as follows:
7
The steps in the predictive analytics process.

1. Define your project’s objectives.
What is the desired outcome? What problem are you trying to solve? The first step is to
define your project’s objectives, deliverables, scope, and data required.
2. Collect your data.
Gather all the data you need in one place. Include different types of current and historical
data from a variety of sources – from transactional systems and sensors to call center logs –
for more in-depth results.
3. Clean and prepare your data.
Clean, prepare, and integrate your data to get it ready for analysis. Remove outliers and
identifying missing information to improve the quality of your predictive data set.
4. Build and test your model.
Build your predictive model, train it on your data set, and test it to ensure its accuracy. It may
take multiple iterations to generate an error-free model.
5. Deploy your model.
Deploy your predictive model and put it to work on new data. Get results and reports – and
automate decision-making based on the output.
6. Monitor and refine your model.
Regularly monitor your model to review its performance and ensure it’s providing the
expected results. Refine and optimize your model as needed.
7. Predictive vs. prescriptive analytics
After building and deploying predictive models that generate accurate, timely predictions –
what’s next? Many businesses see prescriptive analytics as the next logical step.
Predictive analytics helps you determine what’s likely to happen next, whereas prescriptive
analytics can tell you what to do about it – or how you could achieve a better result if you did
X, Y, or Z. This type of advanced analytics builds on predictive analytics and takes many,
many different factors into account to prescribe the best possible course of action or decision.
Prescriptive analytics is often described as the “last phase of business analytics.” It’s also the
most complex and relatively new – currently sitting at the peak of Gartner’s Hype Cycle for
Analytics and Business Intelligence 2020.
Predictive Modeling: Types, Benefits, and Algorithms
Predictive modeling is a method of predicting future outcomes by using data modeling. It’s
one of the premier ways a business can see its path forward and make plans accordingly.
8
While not foolproof, this method tends to have high accuracy rates, which is why it is so
commonly used.
What Is Predictive Modeling?
In short, predictive modeling is a statistical technique using machine learning and data
mining to predict and forecast likely future outcomes with the aid of historical and existing
data. It works by analyzing current and historical data and projecting what it learns on a
model generated to forecast likely outcomes. Predictive modeling can be used to predict just
about anything, from TV ratings and a customer’s next purchase to credit risks and corporate
earnings.
A predictive model is not fixed; it is validated or revised regularly to incorporate changes in

the underlying data. In other words, it’s not a one-and-done prediction. Predictive models
make assumptions based on what has happened in the past and what is happening now. If
incoming, new data shows changes in what is happening now, the impact on the likely future
outcome must be recalculated, too. For example, a software company could model historical
sales data against marketing expenditures across multiple regions to create a model for future
revenue based on the impact of the marketing spend.
Most predictive models work fast and often complete their calculations in real time. That’s
why banks and retailers can, for example, calculate the risk of an online mortgage or credit
card application and accept or decline the request almost instantly based on that prediction.
Some predictive models are more complex, such as those used in computational
biology and quantum computing; the resulting outputs take longer to compute than a credit
card application but are done much more quickly than was possible in the past thanks to
advances in technological capabilities, including computing power.
Top 5 Types of Predictive Models
Fortunately, predictive models don’t have to be created from scratch for every application.
Predictive analytics tools use a variety of vetted models and algorithms that can be applied to
a wide spread of use cases.
9
Predictive modeling techniques have been perfected over time. As we add more data, more
muscular computing, AI and machine learning and see overall advancements in analytics,
we’re able to do more with these models.
The top five predictive analytics models are:
1. Classification model: Considered the simplest model, it categorizes data for simple
and direct query response. An example use case would be to answer the question “Is
this a fraudulent transaction?”
2. Clustering model: This model nests data together by common attributes. It works by
grouping things or people with shared characteristics or behaviors and plans strategies
for each group at a larger scale. An example is in determining credit risk for a loan
applicant based on what other people in the same or a similar situation did in the past.
3. Forecast model: This is a very popular model, and it works on anything with a
numerical value based on learning from historical data. For example, in answering
how much lettuce a restaurant should order next week or how many calls a customer
support agent should be able to handle per day or week, the system looks back to
historical data.
4. Outliers model: This model works by analyzing abnormal or outlying data points.
For example, a bank might use an outlier model to identify fraud by asking whether a
transaction is outside of the customer’s normal buying habits or whether an expense
in a given category is normal or not. For example, a $1,000 credit card charge for a
washer and dryer in the cardholder’s preferred big box store would not be alarming,
but $1,000 spent on designer clothing in a location where the customer has never
charged other items might be indicative of a breached account.
5. Time series model: This model evaluates a sequence of data points based on time.
For example, the number of stroke patients admitted to the hospital in the last four
months is used to predict how many patients the hospital might expect to admit next
week, next month or the rest of the year. A single metric measured and compared over
time is thus more meaningful than a simple average.
10
Common Predictive Algorithms
Predictive algorithms use one of two things: machine learning or deep learning. Both are
subsets of artificial intelligence (AI). Machine learning (ML) involves structured data, such
as spreadsheet or machine data. Deep learning (DL) deals with unstructured data such as
video, audio, text, social media posts and images—essentially the stuff that humans
communicate with that are not numbers or metric reads.
Some of the more common predictive algorithms are:
1. Random Forest: This algorithm is derived from a combination of decision trees,

none of which are related, and can use both classification and regression to classify
vast amounts of data.
2. Generalized Linear Model (GLM) for Two Values: This algorithm narrows down
the list of variables to find “best fit.” It can work out tipping points and change data
capture and other influences, such as categorical predictors, to determine the “best fit”
outcome, thereby overcoming drawbacks in other models, such as a regular linear
regression.
3. Gradient Boosted Model: This algorithm also uses several combined decision trees,
but unlike Random Forest, the trees are related. It builds out one tree at a time, thus
enabling the next tree to correct flaws in the previous tree. It’s often used in rankings,
such as on search engine outputs.
4. K-Means: A popular and fast algorithm, K-Means groups data points by similarities
and so is often used for the clustering model. It can quickly render things like
personalized retail offers to individuals within a huge group, such as a million or more
customers with a similar liking of lined red wool coats.
5. Prophet: This algorithm is used in time-series or forecast models for capacity

planning, such as for inventory needs, sales quotas and resource allocations. It is
highly flexible and can easily accommodate heuristics and an array of useful
assumptions.
11
Predictive Modeling and Data Analytics
Predictive modeling is also known as predictive analytics. Generally, the term “predictive
modeling” is favored in academic settings, while “predictive analytics” is the preferred term
for commercial applications of predictive modeling.
Successful use of predictive analytics depends heavily on unfettered access to sufficient

volumes of accurate, clean and relevant data. While predictive models can be extraordinarily
complex, such as those using decision trees and k-means clustering, the most complex part is
always the neural network; that is, the model by which computers are trained to predict
outcomes. Machine learning uses a neural network to find correlations in exceptionally large
data sets and “to learn” and identify patterns within the data.
Benefits of Predictive Modeling
In a nutshell, predictive analytics reduce time, effort and costs in forecasting business
outcomes. Variables such as environmental factors, competitive intelligence, regulation
changes and market conditions can be factored into the mathematical calculation to render
more complete views at relatively low costs.
Examples of specific types of forecasting that can benefit businesses include demand
forecasting, headcount planning, churn analysis, external factors, competitive analysis, fleet
and IT hardware maintenance and financial risks.
Challenges of Predictive Modeling
It’s essential to keep predictive analytics focused on producing useful business insights
because not everything this technology digs up is useful. Some mined information is of value
only in satisfying a curious mind and has few or no business implications. Getting side-
tracked is a distraction few businesses can afford.
Also, being able to use more data in predictive modeling is an advantage only to a point. Too
much data can skew the calculation and lead to a meaningless or an erroneous outcome. For
12
example, more coats are sold as the outside temperature drops. But only to a point. People do
not buy more coats when it’s -20 degrees Fahrenheit outside than they do when it’s -5
degrees below freezing. At a certain point, cold is cold enough to spur the purchase of coats
and more frigid temps no longer appreciably change that pattern.
And with the massive volumes of data involved in predictive modeling, maintaining security
and privacy will also be a challenge. Further challenges rest in machine learning’s
limitations.
Limitations of Predictive Modeling
According to a McKinsey report, common limitations and their “best fixes” include:
1. Errors in data labeling: These can be overcome with reinforcement

learning or generative adversarial networks (GANs).
2. Shortage of massive data sets needed to train machine learning: Apossible fix is
“one-shot learning,” wherein a machine learns from a small number of demonstrations
rather than on a massive data set.
3. The machine’s inability to explain what and why it did what it did: Machines do
not “think” or “learn” like humans. Likewise, their computations can be so
exceptionally complex that humans have trouble finding, let alone following, the
logic. All this makes it difficult for a machine to explain its work, or for humans to do
so. Yet model transparency is necessary for a number of reasons, with human safety
chief among them. Promising potential fixes: local-interpretable-model-agnostic
explanations (LIME) and attention techniques.
4. Generalizability of learning, or rather lack thereof: Unlike humans, machines have

difficulty carrying what they’ve learned forward. In other words, they have trouble
applying what they’ve learned to a new set of circumstances. Whatever it has learned
is applicable to one use case only. This is largely why we need not worry about the
rise of AI overlords anytime soon. For predictive modeling using machine learning to
be reusable—that is, useful in more than one use case—a possible fix is transfer
learning.
13
5. Bias in data and algorithms: Non-representation can skew outcomes and lead to
mistreatment of large groups of humans. Further, baked-in biases are difficult to find
and purge later. In other words, biases tend to self-perpetuate. This is a moving target,
and no clear fix has yet been identified.
Plan & Forecast

More Accurately
Free Product Tour
The Future of Predictive Modeling
Predictive modeling, also known as predictive analytics, and machine learning are still young
and developing technologies, meaning there is much more to come. As techniques, methods,
tools and technologies improve, so will the benefits to businesses and societies.
However, these are not technologies that businesses can afford to adopt later, after the tech
reaches maturity and all the kinks are worked out. The near-term advantages are simply too
strong for a late adopter to overcome and remain competitive.
Our advice: Understand and deploy the technology now and then grow the business benefits
alongside subsequent advances in the technologies.
Predictive Modeling in Platforms
For all but the largest companies, reaping the benefits of predictive analytics is most easily
achieved by using ERP systems that have the technologies built-in and contain pretrained
machine learning. For example, planning, forecasting and budgeting features may provide a
statistical model engine to rapidly model multiple scenarios that deal with changing market
conditions.
As another example, a supply planning or supply capacity function can similarly predict
potentially late deliveries, purchase or sales orders and other risks or impacts. Alternate
suppliers can also be represented on the dashboard to enable companies to pivot to meet
manufacturing or distribution requirements.
14
Financial modeling and planning and budgeting are key areas to reap the many benefits of
using these advanced technologies without overwhelming your team.
Benefits of Predictive Modeling
The benefits of predictive modeling include cost-cutting in forecasting business outcomes, all
environmental factors, market conditions, and competitive intelligence. Below are some of
the ways through which you can
harness the real power of predictive modeling,
 Very useful in contemplating demand forecasts.
 Planning workforce and customer churn analysis.
 In-depth analysis of the competitors.
 Forecasting external factors that can affect your workflow.
 Fleet maintenance.
 Identifying financial risks and modeling credit.
PREDICTIVE MODELING AND ITS APPLICATIONS
Customer Relationship Management (CRM) solutions often require the creation of customer
level models that accurately predict buying patterns of customers, based on historical and
current data. Or, you may want to determine the probability of a customer purchasing a
product based on the price points. Analyzing such historical and current data and generating a
model to predict the future outcomes of a product/service is termed as Predictive Modeling.
Types of Predictive Modeling
Predictive Modeling is a statistical technique that is widely used in research and analysis.
There are many types of predictive models to choose from, a few of them being:
 K-nearest Neighbor algorithm
 Majority Classifier
 Naïve Bayes
 Uplift Modeling
 Group Method Data Handling

15
 Logistic Regression
Popular Applications of Predictive Modeling
Predictive analytics is widely used in underwriting, risk management, collection analytics,

direct marketing and CRM, to name a few. Here are some examples of how predictive
modeling can be used:
 Auto insurance - Predictive modeling can be used to determine the risk of accidents to
policy holders
 Fraud detection systems - Predictive modeling can be used to identify high-risk

transactions/customers Pro-active customer retention - Predictive modeling can be used
to predict the probability of a customer terminating his/her services. Predictive
modeling can also be successfully carried out in live transactions.
The Use of Predictive Modeling in Social Media
Predictive Modeling is also extremely useful in Big Data scenarios where the data is large,
unstructured, and complex, and cannot be managed by using a normal database management
system. For example, social networks and web logs are sources of Big Data, which if studied
and analyzed carefully, can provide you with significant insights into user behavioral
patterns.
Such data available from social media is useful for companies to study and understand
customer behavior. Social media data types include those from blogs (Blogger, WordPress)
social networks (Facebook, LinkedIn) micro blogs (Twitter) and social news websites (Digg).
The data from social media is characterized by rich user interaction and can be extracted,
analyzed, and used in a variety of ways. Companies can benefit from understanding customer
behavior on social networks. For example:
 A predictive model can be used to understand whether a user will buy a product that is
advertised online
 Predictive modeling can help you evaluate the decision influencers for customers (like
the opinions of fellow buyers, and so on)
 Predictive modeling can be used to understand the correlation between Facebook 'likes'
and sales. This can help companies allocate an accurate marketing budget
16
Choose the Right Predictive Model with Outsource2india
While there are many predictive models, choosing the apt one is always a challenge.
Outsource2india has a well-defined process that can help you choose the right predictive
model for your business. Our expert analysts will first use a data audit to sanitize the data.
This will be followed by data analysis, which will establish correlations and build linear/non-
linear mathematical models. We will then use simulation models to help you understand the
outcomes of the data analysis.
How companies can reap the rewards of predictive analytics
For most companies the idea of predictive analytics fuelled by artificial intelligence to deliver
better website performance and an improved customer experience has been more of a
pipedream than a day-to-day business reality.
Research published by Econsultancy in partnership with RedEye has shown that

companies have been hindered by legacy technology platforms, disparate data sources and
the inability to turn insights into actions.
At the Festival of Marketing held in London earlier this month Matthew Kelleher, chief
commercial officer at marketing automation platform RedEye, described how clients
including Hotel Chocolat and Travis Perkins are reaping the benefits of AI-driven predictive
analytics.
According to RedEye, predictive analytics can be defined as a set of algorithms that assess
the likelihood that something will happen in the future, for example the probability that
someone will make a purchase.
In the case of Hotel Chocolat, RedEye used predictive engagement modelling to help retain
new and disengaged customers to lock in their lifetime value while increasing top-line
17
revenue. Revenue increased by 25% even though 40% fewer emails were sent, according to
Kelleher.
Among other examples given, building supplies retailer Travis Perkins increased retention by
almost 9%, while increasing transactions nine-fold.
Kelleher stressed the importance of understanding the nuances of the customer journey,
explaining: “If you don’t know where the customer is on the customer journey, you are going
to be less targeted in what you put in front of them.”
Effective predictive marketing involves real-time access to a range of customer-related data

sources relating to personal and demographic information, onsite behaviour, engagement,
transactions, lifestyle and devices.
Central to RedEye’s approach is its customer data platform (CDP) that allows it to capture
channel engagements with customers across different touchpoints and devices, building
customer profiles and driving greater insights that enable companies to build customer value.
According to Gartner, by 2020, CDPs will power 20% of current multichannel marketing hub
deployments.
As well as the rise of the CDP, artificial intelligence is another trend starting to have a
tangible impact. Developments in AI and machine learning (ML) technology mean that better
real-time decisions can be made as companies seek to provide the right blend of messaging
and content in their quest to drive sales and extract maximum customer lifetime value.
According to Kelleher, many companies have adopted a ‘wait and see’ approach when it
comes to the use of AI, and this is consistent with the findings of Econsultancy’s Digital
Trends 2018 report. According to that study, 15% of companies surveyed said they were
already using artificial intelligence, while a further 31% said that they were planning to use it.
It is clear that the brands that benefit the most from increasingly AI-driven predictive
analytics in the coming months and years will be those that are able to integrate different data
sources effectively, enabling them to generate the best possible insights and translate that
information into effective marketing actions. This in turn will help them meet important
commercial objectives such as increased customer lifetime value.
18
The last word goes to Paul Morris, the Global Ecommerce Director at Specsavers, quoted in
the Embracing Predictive Marketing report: “Whilst most companies think they want
advanced AI/ML, actually the most critical action all businesses need to undertake is to clean
up and organise their data so that they are then ready to take advantage of, and embrace, the
obvious tailwinds of machine learning and artificial intelligence. Without easy access to the
right data of the right quality it makes it extremely hard to apply the services and APIs now
available.”
How to avoid seven common business pitfalls
Manage the stress factor
Going it alone is a big step, especially if it’s your first time. You may be juggling a full
time job while working on your business dreams, working long hours and worried that
your business won’t make any money. Or that if your new venture isn’t a roaring success
straight away then your friends and family will think less of you.
We get it, it’s stressful. But handling the pressure becomes second nature when you know
how, and there’s plenty you can do to help ease the burden:
 Take time out when you need to - retaining a balanced lifestyle will help you focus
better when you need it.
 Be sure to make time for family and friends – they all want to help you succeed.
 Don’t let initial setbacks get you down – keep a long term view and don’t put
unrealistic pressure on yourself.
 There's also no need to do everything yourself. Delegating, using outside advisers and
building business networks are vital for helping you manage your business
effectively.
How to build a support network article
Lock in your revenue streams

Many small start-up businesses suffer severe peaks and troughs in their revenue streams.
One month they're flat out on a project, the next it’s over and the cash flow has dried up.
19
Often the business has expanded – taken on staff or moved to bigger premises – too
quickly. As a result you now have a bigger payroll and greater overheads. In severe cases
this extra load can cause the business to fail.
Here are some solutions to help you keep afloat:
 Keep your focus on continued cash flow, even if you have to delegate other
responsibilities.
 Continue marketing at all times – especially when you’re busy.
 Be wary of expanding too quickly – have sustainable long-term revenue streams

locked in first.
 Think about contracting out instead of taking on permanent staff and overheads.
 Create good systems to ensure your business runs efficiently.
Understanding your balance sheet

Understanding your Profit and Loss statement
Do your market research
Many start-ups fail simply because they haven't done their homework first – and that
means plenty of market research.
It may be fine to go with your gut instinct for a while, but proper research is the only way
to find out:
 Whether your product or service has a market
 Who your target market is
 The advantages your competition has – and how you can position your business to
beat them.
Target your market
Don’t try to be everything to everybody – you could waste a lot of time, effort and money
marketing to people who are unlikely to buy your product anyway. Instead, define your
target market and focus tightly on them.
20
Understanding where they live, what their interests are, and what they read or watch will
help you target your marketing to the right people. For example, if you’re a tradesman
targeting people in your area, simply putting flyers in local mailboxes can be very
effective. If you’re targeting a younger audience, social media advertising can be a great
way to reach people who share similar characteristics with your target customers.
For help on defining your target market, see our article:

The importance of market research
Case study: marketing choices

A start-up cartridge recycling company decided to spend $10,000 a year marketing
through various media including (in order of spend): newspaper, leaflet drops, shop
signage, vehicle signage and online.
The business then decided to take a 'How did you find us?' survey, which revealed:
 36% were existing customers
 22% were referral
 21% came in as a result of online ads (mainly AdWords)
 10% saw the shop front
 7% came from the leaflet drop
 4% responded to newspaper ads
In fact, almost the opposite to their priority of marketing spend.
They have now changed from newspaper advertising to a customer loyalty and referral
campaign, and a bigger focus on online ads.
Retain your customers
You will work hard to attract customers, so it makes sense to keep hold of them. Some
businesses fail because they spend all their time attracting new customers, while ignoring
those they already have.
To retain your valuable customers and to also gain referral business, put a customer
retention plan in place. Think about the lifetime value of your customers and how to get
them to spend more.
21
Make your website work for you
Online marketing is a vital tool for almost every business - and a key part of that is your
business website. It doesn’t have to cost the earth, you just need to be smart about the way
you use it.
From selling your products and services to enabling your business to be found by new
customers, being online is a must. Here’s why:
 Your website is ‘open’ 24/7 - so potential customers can browse, research, and if you
have the facility, to book and buy online from you whenever they want, not just during
your ‘physical’ open hours.
 Reach people worldwide - a website also means you can attract international
customers, giving you export opportunities.
 Better customer experience – people like buying products or services online, the
option of an online booking system, and even if you don’t sell your product or service
online, customers like researching before they buy.
 Reduce costs – this is especially true if you’re doing all your business online. For
example, if you’re selling a skin-care range, you’re saving on shelving, displays, rent
and rates.
 Develop a customer database – you can use your website to build up a database of
leads. Providing you have the right privacy statements and opt-in functionality, you
can use the contact details you collect to send e-mails and to profile and analyse your
customers – another valuable marketing tool.
Build a successful culture
Before you take on your first employee, think about how you can motivate them to help
you achieve your business goal.
Remember that your people - including you - are your biggest asset. Keep your workforce
happy and they’ll ensure you stay competitive – and attract more great people to your
business too. It’s a win-win.
Ensure your business has a vision that everyone can believe in, and offer their input. Give
them responsibility, recognise and reward them when they do well.
22
You might also consider providing key staff with either a profit split, or some attempt to
reward consistent effort. Even setting short-term goals and rewarding people with a half-
day off can work wonders.
How to Avoid the Most Common Small Business Pitfalls
Starting any business is a learning experience, and you will make mistakes along the way –
and that's okay. To succeed, here are common business mistakes to avoid.
There is no question that running a small business is both a challenging and rewarding
experience, one that involves commitment and hard work to succeed. But according to
an Industry Canada study, 30 percent of small businesses won’t make it to two years, and
only 50 percent will survive to see their fifth anniversary. 1
How can you beat the statistics and run a successful business?
Develop a strategic plan
Many businesses begin with the spark of an idea and a passion to make it happen. That’s
great, but to build that idea into a viable business you need a plan. Start with your why — the
fundamental reason you are in business, how you benefit your customers, and what you want
to accomplish.
“Why we do it is really what drives our overall strategy,” says Becca Perren, Co-Founder and
Co-CEO of Pehr, a line of home décor and children’s products. “Why we are here today, why
we’re going to be here in 50 years. The why leads the strategic decisions.”
Ask your these seven key questions to develop a strategic plan to grow your business.
Some small business owners may approach the business day-to-day without having a plan in
place. But according to Cam Heaps, Co-Founder of Steam Whistle Brewing, a good strategic
plan can make daily decisions easy. “The stronger the plan — a good plan, really — just
makes day-to-day decision-making simple, easy and efficient,” he says. “So you can see
everyone’s going the same direction.”
“You can spend — even waste — a lot of time on a day-to-day basis if you’re not all going
towards the same direction,” says Cam. When faced with a difficult situation, ask yourself,
“Strategically, will this advance my business plan,” and if the answer is no, don’t do it.
23
Manage your time
The life of a small business owner is hectic – you’re likely wearing numerous hats and have
many situations pulling you in different directions. You may spend your day putting out fires
or being tempted by a new opportunity, and the next thing you know the day is over and you
didn’t cross off a single item from your to do list.
The key to managing your time is to focus on the areas that will move your business forward.
To do this, refer to your strategic plan with weekly, monthly, quarterly, and annual goals.
When a situation arises, ask yourself if it will contribute to your bigger plan. If not, leave it
until you have more time to think strategically or be prepared to let it go.
If your business is large enough that you have a staff, consider delegating the project to an
employee to investigate — they can spend time doing the legwork and present the proposal to
you. To be able to delegate these types of projects, put your people first to ensure they have
the tools they need.
“A lot of my time is now spent with people, managing people, having meetings,” says Becca.
“[I’m] making sure they’re able to get done what they need to do and get my input, and that
I’m training them sufficiently.”
Ensure you can check off these 9 ways to prepare your business for new products and
services.
Control your growth
Trying to develop too many products too quickly is a mistake many small business owners
make. When seeing positive growth, rather than focusing on their core products or services,
many business owners try to be too many things to too many people, which can negatively
impact the quality of their core business.
If you want to expand your offerings, consider all aspects – production, people, marketing
and sales, target customers as well as timelines – and build these thoughtfully into your plan.
Take risks
“The greatest risk any of us probably took is starting our own business,” says Cam. “Once
you’re in it you’re in it, and I’d do it 100 times over.” But once your business is established,
you may think you have too much to lose and shy away from risk. But risk can lead to
reward.
24
When assessing a risk, refer back to your strategic plan and consider if it is in line with your
goals, is manageable, and could move your business forward. If the answer to all questions is
yes, it may be worth taking a chance. If it does work out, that opens up a whole new
opportunity for growth. In the event it doesn’t work out you can learn from the experience
and make another decision to continue on your path.
“The worst decision you can make is not making one,” says Becca. “Sure we’ve made
mistakes, but we move on.” Cam agrees. “Think big and go for it!” he says.
Create a positive work environment
Many entrepreneurs get so focused on building their businesses that they forget about work-
life balance. Working too many late nights and weekends can lead to burnout, not only for
you and your employees but it could also negatively impact your family, and that’s not good
for anyone.
The most successful small business owners create a positive workplace. Offer flexibility,
support your team, give them respect, and reward successes.
“My partner and I believe in work-life balance. We promote people enjoying work, but also
getting home and enjoying their families,” says Cam. “If you’re not having fun, your balance
is off.”
The better the work environment, the happier and more loyal your employees will be – and
that’s good for business.
Starting any business is a learning experience, and you will make mistakes along the way –
and that’s okay. Having a strong strategic plan, managing your time, and planning your
growth are all keys to small business success.
Important Applications Of Predictive Modelling In Business
True-lift Modeling: This is a predictive modelling technique, also known as uplift modelling
that directly models a direct marketing action on an individual’s behaviour.
Online Marketing: This technique uses the web surfer’s past data and makes it run through
the algorithms for determining the type of products the user is most likely click on.
25
Fraud Detection: This model is used to detect the fraudulent by identifying outliers in a
dataset that indicates any fake activity.
Churn Prevention: This technique uses predictive analytics to predict when and why a
customer is most likely to end the relationship with the company. This study was developed
to predict churn of customer’s account information on telecom.
Sale Forecasting: This can be called the most used technique using predictive modelling.
Examining the past records, market-moving events, keeping track of sales, etc. results in a
realistic prediction sale in a company.
Process of Predictive Analytics
If predictive analytics is placed at the heart of an automated decision-making process, then it

must be approached in a controlled and systematic way. Building a decision-making
infrastructure based on predictive analytics is just like any other sort of project, like setting up
an IT data center, refitting a factory or restructuring an organization.
Let us understand the different types of analytics used in many organizations today.
Descriptive Analytics uses data aggregation and mining for providing insights into the past
and answers: “What has happened?” These are the most common reports that organizations
generate. These reports show patterns of what happened in the past and allow the analyst to
make their predictions.
Predictive Analytics makes use of statistical models and forecasting techniques to
understand the future and answers: “What could happen?”
Prescriptive Analytics makes use of optimization and simulation algorithms to advice on
possible outcomes and answers: “What should we do?”
Predictive analytics not only about what’s happening, but it can also predict what WILL
happen in the future, which is precious stuff. Rather than just explaining the who, what,
where, when; predictive analytics PREDICT the best course of action that will generate the
most optimal return based on an algorithm, such as a regression equation.
Predictive analytics prime objective is to analyze data and manipulating variables to extract
forecasting capabilities from existing data. Predictive analytics techniques rely on variables
that can be measured, manipulating metrics to predict future behavior or outcomes given
various quantifiable approaches. Predictive analytics models combine multiple predictors, or
26
quantifiable variables, into a predictive model. This approach allows for the collection of data
and subsequent formulation of a statistical model, to which additional data can be added as it
becomes available.
The prediction process involves the following steps:
1. Problem Identification – Have a Predictive that needs an answer. Who to target?

Which coupon? What medium of communication. Example: What is the probability
that the VIP Customers are likely to renew if the price was increased by $4?
2. Determining the Outcome and Predictors – Renewal patterns using Customer data,
Store and Product attributes like pricing and the past membership renewals.
3. Explore data and segregate data – This involves exploring the data and separating
the data, cleaning the data for outliers.
4. Test the models – Assuming that through exploration we find that there is a strong
relationship between Customer’s age and the income levels for renewals; we can plot
a scatter plot using historical data and then based on a probability of renewals; we can
determine the membership price change.
5. Applying models to an identified population to predict behavior and evaluate –
Sometimes we can use the historical data to test the models to an already known
result. For example, taking the 2014 and 2015 data to test for a model for 2016
outcome, for which we already know the outcome.
Organizations involved in the Predictive Process:
 The Problem needs to be identified by the Business users as they see a need.
 Once the problems are identified, then the IT Organization or the business needs to
where to look for the data and\or how to integrate the data for an integrated
exploration. Most companies have a data warehouse or big data integration through
Hadoop.
 Exploration Tools such as Tableau, QlikView, and Lumira can be used to explore and
segregate data by the Business users as well IT resources depending on the
complexity of the data
 For Models, SAP has a Predictive Analytics Library as well as the ability to integrate
with R library to utilize pre-existing models or even to tweak them a little to suit the
desired outcome.
27
 Some examples of how Predictive Analytics can be used are:

o Market Basket Analysis
o Customer Segmentation
o Stock-Outs Elimination
o Customer Profitability
o Response Modeling
o Loyalty Program / Churn Analysis
o Pricing / Promotion Planning
o Inventory Optimization
Predictive analytics allows organizations to become proactive, forward-looking,

anticipating outcomes and behaviors based on data and not on any assumptions. You
need to plan what you are going to do, undertake feasibility work, understand the costs and
benefits and risks and issues, and have someone overseeing the whole show to ensure that it
delivers what was promised on time and to budget.
Predictive analytics
Predictive analytics is the branch of the advanced analytics which is used to make
predictions about unknown future events. Predictive analytics uses many techniques from
data mining, statistics, modeling, machine learning, and artificial intelligence to analyze
current data to make predictions about future.
Data mining
Data mining is the process of finding anomalies, patterns and correlations within large
data sets to predict outcomes. Using a broad range of techniques, you can use this
information to increase revenues, cut costs, improve customer relationships, reduce risks and
more.
28
10 Big Myths About Big Data
Big data, data science and big data analytics are perhaps some of the hottest terms in today's
technology world. But, at the same time there is a lot of misunderstanding and confusion
about those terms, so people may start thinking in different directions, which may not be
correct.
What Do Big Data, Data Science and Analytics Really Mean?
Before looking at the myths of big data, you must understand big data, data science and
analytics. Nowadays, everything is connected to the internet. These things generate data
every day, which can be utilized by businesses or organizations to get useful insights about
the users, and this is called big data.
Data science and analytics can be defined as a process of managing and utilizing such data
for getting insights. There are lot of tools available for analysis of these huge stores of data.
So, all these terms are connected with each other and revolutionize the world of data.
Why Are There Myths About Big Data?
Big data is considered to be a deity of sorts. It is so important for businesses that

organizations believe that without the support of big data, other businesses will overtake them
and they will be last in the race. Thus, thousands of myths surrounding big data have popped
up. And, if you concentrate too much on these myths, then your overall business efficiency
can be hampered.
What Are the Big Data Myths?
There are many myths surrounding big data. Many people are either overly enthusiastic about
adopting it, or fear about its adoption. These myths can seriously hamper your decision
making skills. Some of the most well-known myths are discussed below.
1. Big Data is Just Hype
It is a very popular opinion of the masses that big data is overhyped. They believe that big
data is actually nothing but the “same old data,” just in humongous amounts. They believe
that there is nothing new in the concept, except that only data scientists can read the
29
information from the data. This and the additional costs included for technology makes it
even more expensive. Thus, there is the expectation that big data won’t be used by smaller
businesses for a few years.
2. All Problems Can Be Solved With Big Data
Businesses believe that any problem related to analytics is a big data problem. However, not
everything is a big data issue. For example, if you are trying to match some terabytes of
information to a couple of fields according to a few conditions, it really isn’t a big data
problem.
3. Big Data Can Predict the Future
This one is not completely a myth, but rather it is what some would call a half-truth. Correct
use of big data can really give you some insights for prediction of the future, but these
insights are based on historic data. This means that the insights will depend on the data which
was analyzed and the requirements or the questions of the user. Therefore, big data is not
100% reliable for future predictions.
4. Big Data Is Only for Big Organizations
Many companies believe that big data is only for big companies with big budgets. This is the
reason why mostly big companies use big data solutions. Big data requires a lot of capital for
technological setup and manpower. However, as the cost of these components decrease, the
power of these technologies will increase too, and more startups will be able to use such
technologies. At the same time, we must remember that cloud computing is also making these
technologies and platforms available to the smaller organizations at a lower cost. So, big data
is becoming affordable to all types of organizations. (For more on big data and cloud
computing, see Big Data in the Cloud - How Secure is Our Data?)
5. Big Data Is Better, But Messy
In big data, accuracy of the insights depends completely on the magnitude and reliability of
the data being analyzed. So, this would mean that if you analyze the wrong type of data, then
your insights will be wrong too.
30
Large amounts of wrong data may also lead to bad decisions. Another example of this is
messiness of the data, as analyzing big data isn’t very easy work. However, as analytical
solutions are becoming more and more user-friendly, it’ll be easier to analyze the data.
So, the challenge is to make this messy data (big data) clean and then analyze it to get proper
insights.
6. Big Data Technologies Are Matured
In actuality, big data technologies are simply a network of different types of software with
special features for computing large volumes of data, and it evolves with time. Thus, big data
technology isn’t completely matured, as there are many flaws in these network/ecosystem
components. It still is not completely mature enough to analyze the recent flood of diverse
data types. Big data will gradually mature, as more and more people start adopting it.
7. Big Data Will Replace Existing Data Warehouses
This is a really dangerous myth. Big data is still not evolved enough to serve the needs of
every type of data-related issue. And, we must also remember that big data
technologies/platforms are not a replacement for traditional data warehouses or RDBMS. Big
data is for specific requirements and should not be applied everywhere. So, big data is not
meant to replace current data warehouses, though it may meet some requirements of data
warehouses in the near future. (To learn about storing big data, see How Big Data Impacts
Data Centers.)
8. Big Data Strategy Is Only an IT Responsibility
Having an IT department in a company really helps, as it often sets up the various kinds of
software and hardware required for big data. However, a dedicated IT team alone isn’t
enough to deploy a big data strategy. The big data strategy helps in making better decisions,
so the department in charge of the decisions must carefully evaluate the solutions.
9. Hadoop Is the Ultimate Solution for Big Data
Hadoop is often considered to be the best big data solution. However, there are many other
alternatives to Hadoop. The best solution actually depends upon your own requirements.
10. Big Data Is New

31
The term “big data” is new, and the data available today is also very new. But the concept of
big data and its uses are actually very old. Many companies used big data before it was
officially called “big data,” so this myth is not entirely true.
Are These Myths Really Important?
These big data myths are very obstructive and can result in bad business decisions. These
myths can cause you to waste your precious resources, which would have otherwise been
used to increase your businesses flexibility. These myths can even cause you to miss
important opportunities for your business and lead to bad decisions. Thus, you should know
the full truth, as half-truths can be really dangerous for your business.
Summary
Big data is a relatively new concept, and its proper adoption can lead to better business
decisions, more sales and customer satisfaction. However, like all new concepts, this too
comes with a lot of untrue facts, which are mainly rumors spread by ignorant people.
Believing in these rumors can hamper progress and lead to many other problems. Thus, you
must know how to tackle these rumors and ensure that your business works properly.

What Is Predictive Analytics?

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

What Is Predictive Analytics?

Uploaded by

Copyright:

Available Formats

1

Module -5 Predictive analysis 9 hours

What Is Predictive Analytics?

Predictive analytics definition

 Reducing employee and customer churn

 Identifying customers who are most likely to default on payments

 Supporting data-based sales forecasting

 Setting optimal pricing

 Tracking when machines will need maintenance or replacement

Actionable, accurate predictions are essential in helping decision-makers navigate a world

Predictive analytics today

Examples of predictive analytics

Predictive analytics in HR can be used to predict employee churn.

Predictive analytics in retail

Predictive analytics in marketing

Predictive analytics in supply chain

Companies using predictive analytics

Rainforest Connection: Up to 90% of logging in the rainforest is illegal. Rainforest

Basic steps in the predictive analytics process

The steps in the predictive analytics process.

Predictive Modeling: Types, Benefits, and Algorithms

What Is Predictive Modeling?

A predictive model is not fixed; it is validated or revised regularly to incorporate changes in

Top 5 Types of Predictive Models

The top five predictive analytics models are:

Common Predictive Algorithms

Some of the more common predictive algorithms are:

1. Random Forest: This algorithm is derived from a combination of decision trees,

5. Prophet: This algorithm is used in time-series or forecast models for capacity

Predictive Modeling and Data Analytics

Successful use of predictive analytics depends heavily on unfettered access to sufficient

Benefits of Predictive Modeling

Challenges of Predictive Modeling

Limitations of Predictive Modeling

1. Errors in data labeling: These can be overcome with reinforcement

4. Generalizability of learning, or rather lack thereof: Unlike humans, machines have

Plan & Forecast

Free Product Tour

The Future of Predictive Modeling

Predictive Modeling in Platforms

Benefits of Predictive Modeling

 Very useful in contemplating demand forecasts.

 Planning workforce and customer churn analysis.

 In-depth analysis of the competitors.

 Forecasting external factors that can affect your workflow.

 Identifying financial risks and modeling credit.

PREDICTIVE MODELING AND ITS APPLICATIONS

Types of Predictive Modeling

 K-nearest Neighbor algorithm

 Group Method Data Handling

Popular Applications of Predictive Modeling

Predictive analytics is widely used in underwriting, risk management, collection analytics,

 Fraud detection systems - Predictive modeling can be used to identify high-risk

The Use of Predictive Modeling in Social Media

Choose the Right Predictive Model with Outsource2india

How companies can reap the rewards of predictive analytics

Research published by Econsultancy in partnership with RedEye has shown that

Effective predictive marketing involves real-time access to a range of customer-related data

How to avoid seven common business pitfalls

Manage the stress factor

How to build a support network article

Lock in your revenue streams

Here are some solutions to help you keep afloat: