You are on page 1of 8

CUSTOMER ANALYTICS CHAPTER 1 ➢ Using mathematical models to detect patterns:

There are many number crunching, statistical


The purpose of business is to create and
analysis, and advanced modeling techniques that
keep a customer. This statement was made by
help turn raw data into more meaningful chunks.
Peter Drucker, the acclaimed 20th-century
management consultant. A simple statement that ➢ Finding the insight: From the patterns of the
reveals in just a few words that the long-term data come insights into causes of customer
viability of a company is not just about maximizing behavior.
revenue and minimizing costs. Long-term viability is
about understanding what it takes to attract ➢ Supporting decisions: Understanding past
customers by continuing to meet and exceed their behavior helps predict future customer behavior
physical and psychological needs. from data instead of relying on intuition.

Good customer management comes from ➢ Optimizing the customer experience: Detect
good customer measurement. Metrics are numbers problems with features, purchases, and the
assigned to everything from website visitors, same product or service experience.
store sales, and profit margin to call-center wait
➢ Mapping the customer journey: From
times. Analytics are metrics plus the methods that
considering, purchasing, and engaging with
drive meaningful decisions. You can think of
products and services, mapping the touchpoints
metrics as “informational,” while analytics are
and pain points helps identify opportunities for
“strategic. ” So while not all metrics are analytics,
improvement.
all analytics come from metrics. You need metrics
to get analytics. FOUR CRITICAL INGREDIENTS OF CUSTOMER
ANALYTICS
Increasingly, decisions are made on
numbers. If you know the numbers better, and can ➢ Customer focused: The first word in customer
articulate what those numbers mean and how you analytics is customer. This means that the metrics
can differentiate a product, organization, or brand, collected need to come from customer actions or
you can distinguish yourself from your competitors. attitudes, or are derived in some way that’s
connected to customers.
WHAT IS CUSTOMER ANALYTICS?
➢ At the individual customer level: You need
Although it might not be called customer
access to the lowest level of customer transaction
analytics, chances are, you’re already familiar with
data, not data rolled up at the product or company
some form of customer analytics. The efforts and
level.
activities of product development, marketing,
sales, and services are driven to anticipate and ➢ Longitudinal: Customer analytics involves
fulfill customer needs. That is, you can’t sell a looking at customer behavior over time.
product unless someone has a need for it.
➢ Behavioral and attitudinal: You need a mix of
ACTIONS & ACTIVITIES IN CUSTOMER what customers do and what customers think.
ANALYTICS? Although customer actions (purchasing,
➢ Gathering data: Pull together customer recommending) are ultimately what you care
purchase records, transactional data, surveys, and about, attitudes affect actions — so measuring and
observational data at all phases of a customer’s understanding customer attitudes helps to predict
journey. future behavior.
➢ Sales: Front-line contact with customers,
knowledge of pricing, revenue, transactions, and
BENEFITS OF CUSTOMER ANALYTICS
reasons for lost customers are included here.
The benefit of customer analytics is that
better decisions are made with data, which leads to ➢ Product development: This includes product
a number of tangible benefits: features, functions, and usability.
Multimetric
➢ Streamlined campaigns: You can target your
marketing efforts, thus reduce costs. No single metric can define customer
analytics. It requires a combination of both
➢ Competitive pricing: You can price your
behavioral and attitudinal data. Some common
products according to demand and by what
ones include:
customers expect
➢ Revenue: Simple enough, this is your top line
➢ Customization: Customers can select from a
and you’re probably tracking this for your
combination of features or service that meets their
accountant already.
needs.
➢ Transactions: How many transactions are you
➢ Reduced waste: Manage your inventory better
completing in a given time frame? Digging deeper
by anticipating customer demands.
into the data, transactions become important for
➢ Faster delivery: Knowing what products will sell finding patterns.
when and where allows manufacturing efforts to
➢ Customer Lifetime Revenue: The total top line
anticipate demand and prevent a loss of sales.
revenue a customer generates over some
➢ Higher profitability: More competitive prices, “lifetime,” which can be days, months or years.
reduced costs, and higher sales are results of
➢ Future intent: Will your existing customers buy
targeted marketing efforts.
from you again?
➢ Loyal customers: Delivering the right features at
➢ Likelihood to recommend: How likely will
the right price increases customer satisfaction and
customers recommend your company and
leads to loyal customers, which are essential for
products?
long-term growth.
➢ Product usage: Which features are your
DATA COLLECTED WITH CUSTOMER ANALYTICS
customers actually using?
Multidisciplinary
➢ Website visits: Are potential customers finding
The realm of customer analytics crosses your website and doing what you expect — finding
departments, skills, and traditional roles. It’s information or buying a product.
multidisciplinary and typically involves input from
and output to: ➢ Return rates: How many products are being
returned due to dissatisfaction?
➢ Marketing: This encompasses the messaging,
advertising, and the customer demographics and ➢ Abandonment rates: Did a customer start a
segments. transaction and then quit before completing?

➢ Information Technology (IT): The IT department ➢ Conversion rates: How many potential
usually has access to the databases of customer customers do you convert into actual customers?
transactions and data.
➢ Satisfaction: Are customers satisfied with your ➢ Neural networks/machine learning: Advanced
product, company, and service? software programs can adapt to patterns learned
from data mining and better predict customer
➢ Usability: Do customers have problems using
needs. This is covered in Predictive Analytics.
your products?
CUSTOMER ANALYTICS USAGE
➢ Findability: Can customers find the features
they’re looking for in your products, or find what Customer analytics are used across
they’re looking for in your website? industries in both small and large organizations.
Examples of customer analytics at work include:
Multimethod
➢ Retail: Targeted promotions based on past
No single method defines customer
purchase for individual customers mean retailers
analytics. Some common methods, most of which
anticipate needs and send coupons for things like
are discussed throughout includes:
home improvement or diapers.
➢ Surveys analysis: This involves collecting,
➢ Finance: Credit card companies can understand
analyzing, and posing decision questions directly to
when customers are more likely to cancel their
your customers.
account based on non-usage, as well as detect
➢ Customer segmentation: Not all customers have fraud based on unusual purchases.
the same backgrounds, goals, or buying patterns;
➢ Online: An understanding of which designs,
grouping your customers into similar patterns helps
layouts, navigation structures, and even how the
identify opportunities for better marketing and
color of buttons affect customer purchases (called
product development.
conversion rates) is used extensively across most
➢ Customer journey mapping: Understanding the Internet retailers.
process customers go through as they engage with
➢ Software: Customers who need sales force
a service uncovers pain points and opportunities
automation software often also need accounting
for improvement.
software and human resources software.
➢ Transactional analysis: This examines the Professionals who specialize in customer analytics
purchase frequency, amount, and the type of
products purchased together for patterns and Professionals who specialize in customer
predictions. analytics typically have a background that includes
a mix of mathematical and software skills. These
➢ Factor analysis: This statistical technique helps individuals typically go by titles such as:
identify clusters of similar customers and similar
response patterns from survey results. ➢ Data Scientist

➢ Cluster analysis: Similar to factor analysis, this ➢ Statistician


statistical technique groups customers together
➢ Database Analyst
into clusters and identifies the best labels for
customers to find items in website navigation. Just as most organizations are already
measuring customer analytics, most
➢ Regression analysis: This statistical technique businesspersons also can use customer analytics.
identifies the key variables that have the biggest You don’t need a PhD in statistics or even a
impact on customer satisfaction and customer specialization in math. All you need is some desire
loyalty. to better understand a customer and a willingness
to answer business questions with data. Customer mothers, are more likely to purchase certain
analytics is therefore also done by: products and aren’t sensitive to changes in price.

➢ Business analysts ➢Product A model is a description of a customer


developers interaction, process, or behavior that can be used
to predict future outcomes. For example, the sales
➢ Project managers ➢ ➢ Designers price of a house can be estimated by its total
square footage. In general, bigger houses sell for
more than smaller homes in the same
COMPILING BIG AND SMALL DATA neighborhood.
Customer analytics is often associated with This subject focuses on the methods and
big data. Big data refers to extremely large metrics that help answer business questions and
datasets, often containing millions or billions of set the groundwork for predicting. There is a
customer transactions or records. These large special branch of customer analytics that deals
datasets are analyzed with sophisticated software exclusively with making predictions with software.
to reveal patterns, trends, and associations. Big
data allows you to detect very subtle trends and ➢ What-if scenarios: Customer analytics allows
patterns that may have a large impact on revenue. you to test “what-if” scenarios by looking at past
customer data and estimating how future data may
But customer analytics is also about small
change based on manipulating things like product
data. While not as trendy as big data, small data
features, prices, messaging, or some combination
refers to finding insights with datasets that often
of those elements.
contain less than 30 customers. With small data,
you’re limited to seeing larger patterns in ➢ Customer experience: While customer analytics
attitudes. often involves the hard numbers of transactional
data, sales, and profitability, it also involves
CUSTOMER ANALYTICS
understanding the customer’s experience with a
While the field of customer analytics is still product or service.
being defined and varies across organizations and
Measuring the customer experience
industries, it usually involves some combination of
involves collecting metrics for the entire journey a
the following:
customer has with a brand or organization —
➢ Past behavior: Customer analytics is about including awareness, purchasing, and long-term
using data from the past to predict future behavior. usage. This involves collecting a mix of metrics
This is both a definition and a warning. What about behaviors and attitudes.
customers did in the past is no guarantee of what
Customer analytics includes the metrics for
they will do in the future. If a certain type of
the customer experience. It’s as much about how
customer purchased one type of product in the
the customer uses a product as it is about what
past, he is probably more likely to do so again in
goes into the product.
the future; however, there’s no guarantee.

➢ Predictive modeling: Software programs are


able to detect patterns in behavior, even subtle
ones that are difficult for humans to detect with
intuition or just inspecting data. For example,
software can determine quickly that a certain
segment of customers, such as higher-income
continuous dictates the method you use in your
analysis and reporting.
The Science & Art of Metrics
CHAPTER 2 DISCRETE AND CONTINOUS DATA

Customer analytics largely involves turning Discrete data has finite values, or buckets.
customer actions and attitudes into data. You can count them. For example, the number of
questions correct would be discrete: There are a
Not all data is the same, and knowing what finite and countable number of questions.
type of data you’re dealing with guides you
through what you can do with it. When you can Other examples of discrete data are:
interpret the data that’s available to you, you can ✓ Number of products in your catalog
start making better decisions about product
features and service experiences. ✓ Number of employees you have

In this chapter, you will understand the ✓ Number of customer reviews for a specific
different types of data, and how to work with product
quantitative and qualitative data. Including best
Continuous data technically has an infinite
practices for identifying the best metrics to
number of steps, which form a continuum. The
manage.
time to find a product on a website is continuous
You’ll have to put on your analytical because it could take 31.627543 seconds. Time
thinking cap (the science of metrics) and your forms an interval from 0 to infinity.
creative thinking cap (the art of metrics). If you
Other examples of continuous data are:
aren’t a numbers person and decide to hire
someone who is, that’s okay. But this is still a useful ✓ Dimensions of a specific product
chapter because you need to know how to talk to
your analytics person. And you still need to know ✓ Miles to your retail store from a customer’s
how to interpret the numbers so you can make location
sound decisions about how to improve your ✓ Time for a customer to find the information he
business. or she is looking for on your website
Quantitative data ✓ Days until a product ships to a customer
Quantitative data is information that’s You can usually tell the difference between
broken down by concrete numbers — for example, discrete and continuous data because discrete data
how many products a customer places in the can’t be broken into smaller meaningful units. You
shopping cart (3) or how much revenue you earn can’t have half a customer, but you can have half a
from a specific customer ($2,000). minute.
Quantitative data falls into two categories: LEVELS OF DATA
✓ Discrete (countable items) Another way customer analytics data gets
✓ Continuous (measurements) divided is by the four levels of measurement.
They’re levels because they start with data that’s
ADDING UP QUANTITATIVE DATA more limiting in the type of analysis you can
perform to the least limiting:
You encounter a lot of numbers when
quantifying customer experience with products and
services. Knowing whether the data is discrete or
✓ Nominal: This includes discrete data such as the
name of your company, type of car you drive, or
name of a product. Nominal means essentially “in
name only”; if you have a name, it belongs in this
category. Nominal data is qualitative data.

✓ Ordinal: This includes data that has a natural


ordering. The ranking of customers by oldest to
newest, the order of callers in a queue for a call
center, the order of runners finishing a race, or
more often, the choice on a rating scale, such as
from 1 to 5.
With ordinal data, you cannot know with
certainty whether the intervals between each value
are equal. In measuring customers’ attitudes
toward their experience with products and
services, you have to rely heavily on questionnaire VARIABLES
data that uses rating scales. For example, on an 11-
A variable is a characteristic of a product or
point rating scale, the difference between a 9 and a
service that varies, which can often be manipulated. For
10 is not necessarily the same as the difference
example, price, delivery time, and color are product
between a 6 and a 7. variables. Customer variables can include gender,
✓ Interval: This is data that has equally split intervals income, geography, new customer versus existing
between each value. The most common example is customer, and type of industry, to name a few. When
temperature in degrees Fahrenheit. The difference you look at product and customer variables, you can
between 29 degrees and 30 degrees is the same understand how different product attributes attract
magnitude as the difference between 78 degrees and 79 more or less sales and how different customers respond
degrees (I much prefer the latter of these two to different products and feature combinations.
examples). 2 TYPES OF VARIABLE
✓ Ratio: This is interval data with a natural zero point. ✓ Dependent variables are usually the things you care
For example, time to find a product on a website is about but can’t affect directly, such as profitability,
ratio, because 0 time is meaningful. Degrees Kelvin has customer satisfaction, and customer loyalty. You can
a 0 point (absolute zero). The steps in both these scales influence dependent variables by changing the
have the same degree of magnitude. independent variables. An example of this relationship
Whenever you can establish that data is ratio, is shown in Figure 2-2.
you can make reasonable deductions, such as ✓ Independent variables can be directly controlled or
“customers are twice as satisfied using a new product manipulated. For example, independent variables
version compared to an old version.” include price, features, advertising, and usability.
Just because customers’ average rating on one Often, independent variables correlate with
product is a 4 and the rating on another product is a 2 dependent variables, but the correlation doesn’t equal
doesn’t mean customers are twice as satisfied. The first causation. Other variables that you’re not measuring
rating is definitely twice as high, but unless the scale is can “mediate” or be responsible for the relationship.
both ratio, and calibrated so the numbers correspond to For example, higher sales (dependent variable) might be
customer behavior, making such claims is risky. It’s best attributed to a new marketing campaign (independent
to simply say the rating was twice as high. variable) but the increase is actually just due to a
growing economy that’s helped all businesses (mediator ✓ Descriptive: The descriptive data becomes the
variable). template for whom you measure. It includes
demographic data like gender, age, geography, and
You can think of independent variables as the
income. It also includes self-described attitudes and
ingredients you use to cook a stew. The soup is the
preferences toward products, categories, and
dependent variable (what you care about), but adjusting
technology. From this data, you can create meaningful
the ingredients and their combinations is what you can
segments.
control.
You can collect this data from purchases,
Variables often come in the form of words
registrations, surveys, interviews, and contextual
instead of numbers — for example, new customer or
inquiries.
existing customer, male or female, high income or low
income. To make analysis of these qualitative values ✓ Behavioral: The behavioral data becomes the
easier, you can code them into dummy variables by framework for testing experiences. It is the general
assigning them a number (for example, new customers pattern customers exhibit when using your products
get coded a 1 and existing customers get coded a 0; and services. It includes making purchases, registering,
men get coded as 1 and women coded as 0 [or vice browsing, and using different devices.
versa]).
For example, customers of certain product
With your variables coded as 1s and 0s, you can categories, like consumer electronics or home furniture,
compute the percentage of customers with each tend to browse products on their tablet at night and
variable. make purchases on their desktop during the day.

✓ Interaction: The interaction data becomes the task


scenarios that you simulate and measure during a
usability test. It includes the clicks, navigation paths,
and browsing activities found on websites and software.

The classic usability test typically focuses on this


level of granularity by simulating real interactions. You
can use realtime data from A/B testing, Google
Analytics, and lab-based or unmoderated testing to
collect data for this grouping.

✓ Attitudinal: Preference data, opinions, desirability,


branding, and sentiments are usually captured in
QUALITATIVE DATA surveys, focus groups, and usability tests. This is where
questionnaires like the SUPR-Q (www.suprq.com),
Qualitative data is data that is not easily reduced to System Usability Scale (SUS;
numbers. Qualitative data tends to answer questions www.measuringu.com/sus.php), or the Net Promoter
about the 'what', 'how' and 'why' of a phenomenon, Score (www.netpromoter.com) quantifies how
rather than questions of 'how many' or 'how much. interactions and behaviors affect attitudes. These
attitudes will then affect some self described descriptive
Qualitative data is often helpful by itself to explain the
attributes quantified in the descriptive grouping.
“why” behind low satisfaction rates, higher sales, or
high customer turnover rates. Improvements you make that affect attitudinal
data, like increased trust and loyalty, drive further
For example, if you see customers complaining
buying behavior.
that they don’t know what the total price of their order
is in local currency or how to change the currency in a MANAGING THE RIGHT MEASURE
website shopping cart, you know what you can fix.
If you can’t measure the customer experience, you can’t
WHAT DATA TO COLLECT manage it. Improving the customer experience starts
with measuring. But you must be sure you’re getting the ✓ Likelihood to recommend or likelihood to
right measure (or usually measures) to manage. The repurchase: With the popularity of the Net Promoter
right measure(s) will: Score, it may seem like word of mouth is the only
measure you should care about. But if everyone already
✓ Identify problem areas
knows about and owns your product or visits your
✓ Track improvements over time website, likelihood to purchase again might be a better
measure of growth.
✓ Be meaningful to the customer
For measuring customer loyalty, I recommend
The wrong measure(s) can: using both repurchase rates and likelihood to
recommend. This provides a mix of both behavior and
✓ Identify wrong areas of focus
attitude.
✓ Miss problems all together
✓ On-time arrival versus on-time departure: Have you
✓ Lead to unintended consequences ever been on a plane that pulled away from the Jetway
only for it to sit on the tarmac waiting for mechanical
✓ Alienate customers issues or other delays? You then arrive at your
MEASURING EXPERIENCES destination late? It’s likely that the flight segment still
counted as an on-time departure. You can’t argue with
✓ Conversion rate versus number of conversions : the measure: The plane did pull away from the Jetway
Conversion rates are the central metric for testing on time and that does mean something. However, that
better designs, ads, and campaign effectiveness. The action just doesn’t mean that much to the customer
ratio of total users who purchase, register, or click sitting in the idled plane.
(convert) to all users who viewed the page is an
Measuring is good. Knowing what to measure is better.
effective ratio because you can compare low traffic and
Finding the right measure means taking multiple
high traffic pages.
measures and seeing which one best tracks customer
✓ Number of clicks versus time to destination : When sentiments and revenue.
you’re trying to make a more efficient experience,
reducing the number of clicks to accomplish a goal
seems like a good way to measure. Putting all
functionality and content on one page would certainly
reduce the number of clicks, but that probably is not
what your customers have in mind.

✓ Call time or call satisfactorily resolved: Wonder why


those often scorned customer service agents you call to
complain to speak so quickly? If you want to reduce call
time in a customer support center, you can instruct
agents to get off the phone faster, but have you really
increased service or quality if customers have to call
back? Often a simple follow-up question sent via email
can solve this problem.

✓ Customer satisfaction as a bonus motivator: Many


companies pay bonuses based on achieving and
exceeding certain customer satisfaction goals.
Unfortunately and not surprisingly, this can lead
employees to improve their chances for getting the
bonuses in ways that make measures less meaningful.

You might also like