ULTRA PRECISION MARKETING SCIENCES (UPMS)
Diploma Program for Educational Institutions
talents that are
committed to the understanding of customer relationship, improvement in
targeting, acquisition of customers, increasing the value of customer
predictive analytics and extending the lifetime value of the enterprise.
The present business paradigm is an adaptive, inter-active, relevant and highly personalized paradigm that drives us to develop the tools, methodologies workflows and technologies required to manage digital elements (data, imagery, graphics, blogs, etc.) across a suite of integrated systems and relational databases.
This program will give a better understanding of the true value of predictive marketing sciences, the complex nature of relational data structures, how to nurture customers, retain best customers, grow customer share. Academically this helps marketers define “the art-of-the possible”. Inclusive of - CRM, Big Data, Small data, Good Data, Bad Data, Customer Journey, Analytics, ROI, LTV, CLTV, Artificial Intelligence, etc.
Who should attend?
subject is an 8hr
interaction, need to complete core subjects before advancing. Need to
to be certified.
The program is comprised of 12 subjects with a final examination essay evaluation from a selection of case studies.
framework subjects (total
Evaluation is through an “on-line” examination (passing mark 75%)
Ecosystem of the Marketing Experience
There are multiple objectives in this first session: the first is to get a grasp of the operational environment that all marketing analysts, marketing communications experts and executives must operate in; and secondly is to set a strong foundation in the breadth and scope of data which marketing experts must be comfortable with acquiring and analyzing.
The session will cover
traditional data sources but will also cover
business data feeds, social media and operational cyber-based data as
– in essence the operational
paradigm of marketing information. This not only includes
original feeds of data such as POS, audits and surveys, but further
derived data from analytics; the resulting input feeds from online
such as social media feeds; and error feeds such as return email and
In this session you will
learn how to develop the integration of data between our familiar
“Bricks and Mortar” circumstance all the way through to that of the
and “Cyberspace”. We investigate what we need to build from
perspective in order to be able to determine which competitors and what
econometric and socio-geo-demographic forces are responsible for
our business and, to what degree.
Here you will also learn through comprehensive case studies, the required data framework and illustration of the required marketing sciences that would enable you to quantify the potential loss (or gain) of business due to the spatial distribution of the competition. Very insightful!
Client Marketing Groups Through
Using a technique called “Cluster Analysis”, we create tightly defined homogeneous segments that are effective discriminators over a wide range of consumer behavior. Leveraging such techniques provides you with the ability to classify your customers into groups where members are as similar to one another as possible, while being significantly different to the members of other groups -perfect for strategic marketing initiatives for any product or service!
To develop high target profitable marketing groups, you will learn how to:
• Create a segmentation system that will find distinct groups of customers for whom differential communication and offer strategies may be deployed.
• Ensure that members of each group are as similar to one another as possible, and significantly different from members of other groups.
• Incorporate financial services, sales, response, and conversion potential in the segmentation to find under performing high value customers.
• Create a segmentation scoring methodology that is measurable, identifiable and actionable.
Looking for Unique Differences
Between Marketing Groups
You will learn how,
regardless of what
business vertical you operate in, profiling helps
visualize how one group or entity compares against
another. In this “How To” session, you will learn how to use profiles
identify high performing segments.
Marketing experts often deploy ‘Gap Analysis’ in profile development in order to help differentiate the unique essence of each group. This session also looks at how profiles leverage techniques like ‘gap analysis’ to illustrate the power of our marketing discoveries.
Participant will learn how to build and develop indices, non-0 distributions and how to determine the impact of a group or segment by its removal from the universal distribution. Participants will learn how to build a comparative segment performance framework around 8 key index types reflecting sum, mean and median variations.
Predictive Sciences subjects (total 5)
In this session you will
learn how finding
your “Top Performers” can be accomplished
through a technique called Multivariate Interaction Detection [MID]
that help marketers predict product sales,
preference, attrition, etc. Learn
how, using an iterative "splitting" process, MID recombines
the attributes, variables and data to rules that recombines our clients
sub groups that provide the most
outstanding indicators of performance! Once the splitting
is complete, a series of "rules" are developed, where
each rule yields a particular
performance score, response rate, with an understanding of marketing
Through detailed Case Studies you will learn how Financial Services, Retail, Telecommunications, Insurance, Death Care Industry, Not-for-profit, Religion, Education and others generated a series of performance models determining the characteristics of the best performers whilst simultaneously outlined the components of poor performance so as to develop an action plan to help migrate customers up the value scale.
Learn about how an American retailer used a broad and diverse data frame to generate performance nodes with over 300% more buying power than the average client; but also outline an approach to activate moderate shoppers to behave more like the best performers.
2. The Art and Science of Prediction and Forecasting
Predictive modeling is the process of predicting or forecasting the price point a customer may purchase a given product or service. Depending on the prediction circumstance, the model may also try to predict "when" a certain event is going to occur - as well as at what price point. By gaining control and understanding of when your clients would be likely to purchase given products and services and at what price point is remarkably powerful! In some ways this is one of the most important skills/techniques to understand.
In order to develop targeted profitable models identifying the best customer prospects, you will learn how to: “Identify those customers with a high propensity to purchase a specific product around a given set of price points and/or at a specific period of time. We can use a set of predictive models to score our clients to determine how and what to sell, or, how to win them back!
Predictive modeling provides us with the linear solutions for prediction and forecasting models. Through the use of a model, we obtain a list of relevant predictors driving a purchase, and, we also are able to determine the probability of this event. Should we have “time” as a predictor we can also estimate when these events can occur and the likelihood of that timing as well.
1. Learn when it is appropriate to use regression techniques versus MID
2. Learn what data, or, how to acquire the appropriate data for general linear modeling techniques
3. Discover how to use predictive modeling and forecasting in your marketing strategy and how it directly applies to your client LTV goals.
4. Discover how to contrast and isolate prediction and forecasting methods around your key marketing segments – and why this is so important.
5. Learn how to present your forecasting tables and charts in a corporate setting to illustrate your marketing objectives for a given fiscal year(s).
Retention and Attrition
In this session you will
learn how Retention
and Attrition are different sides of the same coin — on one hand
retention is the art of
understanding how to hold on to our customers, whereas, attrition is
the art of
understanding how to get back customers who have
left or who are about to move to the competition.
Learn how Attrition is the process where we look at the probability of when a client ceases to be a client. This style of regression model determines the probability a client will convert, respond to a given event. Clients are scored from 0 to 100% in terms of response, conversion, attrition where treatment is governed by this probability.
Learn how Retention focuses on how we nurture our client to the fullest; how to market; what to market; when to market; to make sure our clients are satisfied across all dimensions.
Some financial institutions around North America find clients moving their investments from one institution to another with the hope that they will provide a better rate and return. Learn how one company used attrition modeling to retain high value customers while continuing to gain new customers and generate a remarkable gain!
For each and every campaign we need to spend budget for the print, web site, email blast and so on. At the end of the campaign we generate revenue. ROI analysis helps us breakdown the performance of the campaign from a revenue standpoint against other campaigns, or within the campaign. Naturally the importance is we learn how much revenue we are generating but more importantly how much profit. Of even greater importance however, is that we begin to understand what campaigns are generating the greatest gain where we compare performance on a segment to segment basis.
Collectively then, we learn what it is we must do to successfully generate profit.
You will learn a number of different methods to calculate ROI - and much of this is dependent upon the array of data you have to work with. Clients in all circles and verticals must prove their campaigns work - learn how to calculate ROI under any and all conditions regardless of data source.
5. Lifetime Value Analysis [LTV]
Lifetime Value “LTV”: can be defined as “the future profits to be realized on a new or existing customer over a given period of time.” LTV is best modelled over a longitudinal database that spans many years of data: client sales and product data, econometric, socio-geo-demographic, competition; and spatial data. LTV is essentially looking at profit from two aspects: all data prior to NOW is a historical view of profit – and when we predict profit into the future we generate the forecast of potential client value (future profit). An outcome of LTV analysis is also the determination of attrition enabling the construction of programs to nurture high value clients to deter these clients from migrating to the completion.
However there may be
circumstances where client databases are not
comprised of transactional data points, and thus LTV can be established
single campaign or cumulatively across a series of campaigns. Lifetime
can be used as a decision tool for both prospects and existing
helps answer question such as (i) How much money will this campaign
over the lifetime of this customer, (ii) what is the appropriate budget
acquiring a new customer, or, (iii) what is the correct spend retaining
Learn how a company uses the power of AI and ROI and LTV calculators to estimate the value of a client, estimate the performance of future campaigns and how to forecast the performance of their latest campaign!
Final certification subjects (total 4)
1. Share of Wallet
Share of Wallet is the
percentage ("share") of a customer's
expenses ("of wallet") for a product that a firm sells. Different
battle over this share they have of a customer's wallet, all trying to
much as possible. In this session you
will learn how to define exactly how much of this wallet share you (or
customer) will have - and even more important how to garner a greater
Learn how share of wallet is probably one of the most important analysis a company can perform. This endeavor is naturally quite difficult due to the data requirements. In the case of financial and retail clients this can be delivered with special surveys, or, by the collection of huge databases.
You will learn how one company was able to determine how much share of wallet they acquired over three markets; essentially enjoying 33% in one smaller market yet only 7% in a larger more promising market.
2. Multi-Channel Attribution (MCA)
Multi-channel attribution is a set of rules that assigns credit for sales & revenue to specific channels, technologies and touch points. When we refer to sales & revenue we are referring to any income that could and would be attributed to the channels used in the campaign.
When we refer to channels they can be specific touch points such as direct mail, email, mobile or social / cyber. Social channels would include banner ad, page posts, sponsored tweets, search ads, landing pages for any given company but often through portals provided by Google, Facebook, Twitter, Minds, Gab, BitChute and YouTube to mention a few more notable.
Touch points are essentially the points of engagement that describes who, what, when and where of the given channel. Touch points are tangible describing the event: who purchased, content used, web assets, form, location, time, campaign and revenue. Attribution therefore refers to the rules and methods used to assign value across multiple channels and touch points: the actual revenue amount is determined by the touch point rules. Sales in this case would also refer to conversion…
The purpose of the course is to help program participant learn how to establish the multi-channel data points in order to create a usable and notable information source. This data source naturally tags the key input vehicle but also the type of media as well. These sources are then broken down by touch point such that a practical analysis can be generated by the program participant.
1. Discover the data requirements and components required for a complete attribution analysis.
2. Learn the various MCA models and when they are applicable:
a. Last Interaction/Last Click Attribution model
b. Last Non-Direct Click Attribution Model.
c. Last AdWords Click Attribution Model.
d. First Interaction/First Click Attribution Model.
e. Linear Attribution Model.
f. Time Decay Attribution Model.
g. Position Based Attribution Model.
3. Discover how to generate MCA reports showing channel proportions, percentages, performance indices and revenue distributions.
3. Econometric Zones of Influence: Understanding Forces Which Affect Buying Power
When we look at any given store location, what exactly do we see and what does it exactly represent? Naturally, as a bricks and mortar building it has an address – and clients from all around will come to this location. The larger the store typically the broader the geographic pull: but, is it just that simple?
Over the past 200 years there have been many attempts to understand what creates what we refer to as “the econometric zones of influence” – where each store is the central focal point of econometric energy drawing from the population surrounding it enacted upon by both the agglomerative forces that support sales versus competitive forces that detract sales
We study the ‘type’ of cities that our companies and firms operate in, and these subtle variances generate an econometric framework which then defines the zones of influence which our store generate. Our stores then – have a well-defined “zone of influence” which naturally is affected by competition, supportive industries, spatial elements (road networks and physical barriers).
These zones of influence then allow us to generate the most accurate models of economic performance possible.
4. AI (Artificial Intelligence)
Artificial Intelligence is many things to many people across many segments and walks of life and business. Consider when IBM’s Deep Blue (Watson’s predecessor per se) played against the Grand Champion Gary Karpov and when IBM’s Watson destroyed the best of the best in the Jeopardy final (Ken Jennings and Brad Rutter): probably one of the best illustrations of message parsing and quick search algorithms.
Today AI is the application
of many structural
& predictive models that have been put into a collection. These
are then applied as rules, essentially programming logic that is used
decisions, search tables or present output as content. Thus AI is the
intelligent operationalization of the outcomes of statistical routines
the output such as linear equations, cluster memberships,
content links or outcome data are recoded as logic: logic to be used in
to make decisions, present content and interface with their human
or other AI avatars.
As machine learning
artificial intelligence is becoming decidedly main-stream. Nowhere is
apparent than in marketing, where AI is becoming a virtually
Key AI takeaways:
1. Understand: Discover the underlying terminology, trends, approaches and currently technologies.
2. Strategize: Develop your AI strategic plan through understanding the processes, best practices, templates and resources available.
3. Activate: The development of an AI pilot – develop the evaluative criteria required to determine success and how to scale your AI.
4. Improve: Upon successful activation learn to monitor, measure, and adapt your AI. Learn what AI performance really is…
1. Program Implementation & Execution
This is where it all comes together and you can start to envision an implementation path for the enterprise. The Who, The What, The When, The Where, The Why, The How. From a marketing perspective we would refer to this as “The art of the possible”.
Program participants will be given the opportunity to select one of 5 case studies to work from. With the case study selected the student will need to discuss the effectiveness of the analytical techniques applied and whether other alternate methods could have been applied and the expected outcomes would be. Program participants will also be required to intelligently identify how the source data was used and how this data could have been further augmented to solve the research problem solved in the case study.
Program participants must debate the research conclusion presented in the case study: does the program participant agree with the conclusions presented by those generated by the case study or would the program participant provide a different conclusion using a different methodology. Program participants will be evaluated on their understanding of the case study and how they present their thesis agreeing and augmenting the conclusions of the case or by their presentment of an alternative approach and its defense.
For Information about how you can acquire
this program for your educational institution, please contact