Home Electronics Planet, a big-box retailer, has digital marketing campaigns that are failing. Their Chief Marketing Officer gets some analytics and data science help from Business Scenario Investigators who recommend changing their search keywords mix, creating tighter customer segments based on product purchase sequencing coupled with real-time web page personalizations, and revising their e-mail marketing to improve business results.
10. SCENE 1
At Home Electronics Planet HQ
Problem: Katie shows Jodice the KPIs and explains her current
marketing efforts. Jodice accepts the assignment to help.
20. SCENE 2
Back at BSI Labs, Mercedes and Frazier tap into Planet’s
data stored on Teradata, and use a new discovery tool called
Aster to discover what’s really happening with the
customers
35. SCENE 3
Jodice provides a Readout and Architecture
Recommendations for Katie and the VP of IT, Lincoln Duckett
You’ll find details in this section on the
•Teradata Unified Data Architecture™,
•Celebrus Technologies
•Teradata Aster and
•Teradata Applications, specifically Integrated Marketing
Management
38. TERADATA UNIFIED DATA ARCHITECTURE
System Conceptual View
ERP
MOVE
MANAGE
ACCESS
Marketing
Marketing
Executives
Applications
Operational
Systems
Business
Intelligence
Customers
Partners
SCM
INTEGRATED DATA WAREHOUSE
CRM
Images
DATA
PLATFORM
TERADATA DATABASE
Audio
and Video
Machine
Logs
Data
Mining
TERADATA
DATABASE
HORTONWORKS
HADOOP
INTEGRATED DISCOVERY
PLATFORM
Frontline
Workers
Business
Analysts
Math
and Stats
Data
Scientists
Text
Languages
Web and
Social
Engineers
TERADATA ASTER DATABASE
USERS
SOURCES
ANALYTIC
TOOLS & APPS
Start outside the UDA
This is a “system view”, for introductions. There are more layers behind this.
Ultimate goal of supporting every user/application with the right data
Support any analytic technique – seamless to end user
Data – ALL data, is on the left hand side – from traditional to newer kinds of data
End users and their tools are on the right.
Shift inside the UDA
3 general categories of data-based tools
Why data flows left to right
Data can be loaded directly into any of the systems
The value of Data Platform – integrated for storage
The value of Discovery Platform – integrated for analytics
The value of IDW and its evolution – point of value integration/realization
Arrows/connectors in between are important glue
Management and governance tools across the top –value-added software to make the system easy to manage.
Many web tools provide aggregated behavior of web customers. By contrast, Celebrus collects extremely granular data about each and every customer’s every action on the website
CRM – Customer Relationship Management, coming up in a couple of pages. The real-time component is called “RTIM”.
Tagging-free – many web monitoring systems require substantial manual effort to “tag” information elements on web pages; Celebrus avoids this effort
SAX - Enables Machine Data Analysis, such as analysis of sensor data in Manufacturing. Identify anomalies in manufacturing production process or performance of devices.
Sequential Pattern - Automatically identify frequent patterns in sequential data.
Attensity ASAS - Entity/event extraction, classification, sentiment analysis.
Confusion Matrix - Used in machine learning for quantifying the performance of an algorithm and helps improve predictive models. Returns true/false positives/negatives.
Single Decision Tree - Build and apply a single decision tree for classification. Identify important variables (and disregard irrelevant) that play role in making a decision.
Distribution Matching – Test the hypothesis that the data is distributed from a certain distribution and estimate the parameters of several distributions that may fit the data.
LARS – Selects a set of variables that are the most important in the context of a linear regression analysis. Can be used as LARS or Lasso.
Fpgrowth – An association mining algorithm for recommendation engines. Discover elements that co-occur frequently in large datasets.
Histogram –Counts the number of observations that fall into each of the disjoint intervals.
IP Geo/Mapping - Identify the location using IP address
New Slide: Synergistic multi-genre analytics
Combine:
Mfg Yield Management = SQL + Statistical + Time Series + …
Location analytics = geospatial + time series + sql
Digital Mktg analytics = sql +time series + statistics + Text + Graph..
Social Media Analytics = SQL + Graph +Text (Attensity) +Statistics
Churn Analytics = SQL + Time series + statistics + text
Recommendation/Affinity engines = SQL + Statistics + Graph + Time Series
Fraud Analytics = SQL + Statistical + Graph
This slide represents the high-level Aprimo vision for Integrated Marketing Management. We have been in this space for 13 plus years and we recognize the value and the viewpoint around a vision to create integrated marketing for organizations so they can successfully communicate across multiple channels to reach their customers efficiently and effectively.
Looking at the center circle of this slide – we realize there are many functions within marketing - and we have to support those various levels within each organization. From the corporate marketing levels to the field and regional managers, to brand managers and business units. You also have to be able to reach and collaborate with the external suppliers and other external functions within the company yet outside of marketing (c-levels).
And, after 13 years in the IMM space, we have also recognized the ability to utilize the number of channels continues to grow and the ability to communicate and how you communicate has evolved over the years.