10 Great Use Cases In Big Data and Big Data Analytics

Learn Aster
Teradata Employee

Multi-Channel Behavior Analytics

Multi-Channel behavior analytics requires data integration across brands and across catalog, call center, brick and mortar, and web channels to better understand customers’ preferences and behaviors.  Need to promote across different products and brands. Analysis includes:  360 degree customer intelligence, triggered customer programs and offerings, improve personalization cycles and micro-clustering, identification of customer/visitor segments and their paths across sales channels, and more effective out of stock analytics.  Tangible benefits include multimillion dollar impacts to revenue and cost avoidance though score-carding ad spending and effectiveness, increased revenue due to better recommendations engine conversion, understanding channel effectiveness and investment.  Customer satisfaction and attrition can also be measured.  Industries include: retail, mobility, hospitality, and government.

Analytics:  Teradata Aster Analytics:  Path and Pattern, nPath, Collaborative Filtering, Graph, Naïve Bayes, Support Vector, Machines, and Sessionization.


Advanced Market Basket Analytics

Affinity analytics is a very valuable type of analytic for framing retail recommendation engines for next best offers and product placement.  People who purchased product x also purchased a, b, and c.   This type of analysis is also excellent for other vertical applications such as healthcare, engine parts failure, and genomics. The data that feeds this type of use case can be very large and sometimes take days to complete analysis. Asters capability with affinity enables great flexibility and performance.  It is very simple to slice and dice and look at various segments of a market or subset of nodes.  Tangible benefits include tens of millions of dollars in retail product sales.  It also includes manufacturing cost reduction/brand loyalty by avoiding equipment and automobile warranty defect.  Benefits include:

  • More customer relevant promotional up-sell and cross-sell
  • Increase retention and loyalty of high value customers
  • Grow traffic in store and on the web
  • Vendor intelligence and optimization of promotional events
  • Merchandising and Product Placement/Price Optimization

Analytics:  Teradata Aster Analytics:  : Minhash, kmeans, Collaborative Filtering, Graph, Personalized Salsa, Vector Distance, Cosine Similarity, and Generalized Linear Models.

Clickstream Analytics

Clickstream analytics gives our customers the ability to deeply analyze customer web behaviors to better understand: golden path analysis, shopping cart abandonment, web search optimization, site abandonment, marketing effectiveness, and many other metrics.  Aster has been used to develop models for predicting persons that are in an active state of churn by looking at their search topics and running them through a machine learning predictive model.  Clickstream analytics also enable analysts the ability to locate and find operational blocking issues on their website.  Tangible benefits of clickstream analytics include multi-million dollar revenue increases to revenue though streamlining web traffic conversion.

Industry:  Retail, mobility, television content providers, Banking Financial Services

Analytics:  Teradata Aster Analytics:  nPath, Naïve Bayes, Decision Trees, Graph, Personalized Salsa, Vector Distance, Cosine Similarity.

Churn and Customer Defection

Subscription churn is customer attrition prediction.  This type of churn is easy for customers to exit a product or service where there is a monthly fee or charge for that service.  Vertical industries that use Aster for churn prediction include: mobility, banking/financial services, and television content providers.  The Aster Churn process takes advantage of data from multiple channels including: call center, clickstream, brick and mortar, billing, and IVR.  Aster enables the application of path and pattern analysis to develop the path to churn across these channels.  Aster then uses that data as a training data set for developing models to predict deflection.  Tangible benefits included: hundreds of thousands to millions of dollars of revenue. Benefits of Asters churn use case include:

  • Identify broken processes where customers exit, get stuck or cycle back
  • Understand and fix broken processes by analysing individual customer experiences and identifying root cause for failures
  • Outcomes are lower attrition (upto 10%) more revenue from future sales, and avoided customer acquisition costs

Industry:  Mobility, Television Content Providers, Banking Financial Services, Insurance

Analytics:  Teradata Aster Analytics:   nPath, Naïve Bayes, Support Vector Machines, Graph, and Sessionization.


Predictive Parts Failure and Affinity

From automatic teller machines to automobiles; manufacturers want to understand machine breakdown and failure.  Knowing break down affinity enables manufacturers to improve brand loyalty but also to lower warrantee repair costs.  Tangible benefits can be enormous and into the hundreds of millions of dollars. Aster stores and analyzes sensor data to impact product quality and performance, squeezing variation and margin from manufacturing that was hidden in datasets.  Aster is able to find and reduce manufacturing variation hidden in 100s of terabytes of sensor data.

Industry:  Manufacturing

Analytics:  Teradata Aster Analytics:  nPath, CoxPH, Support Vector Machines, Graph, and Sessionization.


Micro Segmentation and Individualized Next Best Offer/Next Best Action

Retailers, television content providers, and banking customers are always looking for upsell and cross-selling opportunities to increase revenue.  To that end it is important for organizations to better segment their customers and achieve individualized next best offers.  For retailers this means building out better recommendation engines and marketing campaigns that reflect the individual.  The banking side of this is about understanding customer life stages and determining what mix of products to offer up next.  The other result found in banking clients was discovering what product mix and count of banking products reduced churn risk. Tangible impacts include millions of dollars of revenue gain from product sales lift and reduction in customer churn.

Industry:  Retail, mobility, television content providers, Banking Financial Services

Analytics:  Teradata Aster Analytics:  Machine LearningNaïve Bayes, CoxPH, Text Analytics, Graph, and Sessionization.


Customer Satisfaction Indexing

Historically organizations have had to rely on net promoter scores(NPS) in-order to gauge customer satisfaction.  NPS relies heavily on surveying customers and may not be a full representation of actual customer satisfaction.  Many times satisfied customers do not participate thus causing skew in results. The Teradata Aster solution not only can use NPS but also can look at data across many different channels. Customers also want to be able to look at data across the entire enterprise in order to find and repair material defects in customer touch-points.  Aster comes with a full framework for harmonizing data into a set of data structures.  There is a web based interface that allows the user to create specific business rules that define satisfaction and dissatisfaction events.  From there custom analytic runs can be performed by implementing the constructed satisfaction/dissatisfaction rules.  When reviewing the analysis customer are able to understand what rules generated the most satisfaction/dissatisfaction scores as well as interrogate customer journeys.  Material impact includes revenue gains and proactive customer churn prevention.

Industry:  Mobility, Television Content Providers, Banking Financial Services, Hospitality, Healthcare

Analytics:  Teradata Aster Analytics:  Machine Learning, Support Vector Machines, Vector Distance, CoxPH, Text Analytics, Graph, and Sessionization.

Online Search Optimization

Online customers typically abandon a retail website when they are unable to find what they are looking for quickly.  One avenue customer use is to search online for a product.  Most customer leave if their search result does not contain the product on the first page of a query search result.  This process is called silent abandonment.  Aster uses a machine learning technique for determining what search terms were used during online conversion activities.  The Teradata Aster Onsite Search framework can be used to improve search query results thus rendering a dramatic improvement in conversion and online sales.  Material Business impact has seen some online retailers with a 3-5% improvement in conversion.

Industry:  Retail, mobility, Television Content Providers, Banking Financial Services

Analytics:  Teradata Aster Analytics:  Machine Learning, Naïve Bayes, Support Vector Machines, Vector Distance, CoxPH, Text Analytics, Graph, and Sessionization.


Communication Compliance


Banking, healthcare, and other vertical organizations are challenged with new regulartory compliance requirements every year.  Teradata Aster has developed an analytical solution to ensure client communication compliance.  Compliance costs in the banking financial sector exceed over 1 billion dollars a year. These costs include process support, IT support, training, examination, and forensic support.  The Teradata Aster Communications Compliance solution makes it easy to:

  • Detect prohibited financial arrangements,
  • Detect deceptive marketing practices
  • Identify illegal sharing of information
  • Detect violations of corporate ethics policies

Industry:  Banking Financial Services

Analytics:  Teradata Aster Analytics:  Machine Learning, Naïve Bayes, Support Vector Machines, Vector Distance, Named Entity Recognition, Levenshtein Distance, Graph, and Sessionization.


Fraud Analytics


Fraud is a constantly evolving mechanism and requires constant model tweaking.  Fraud detection is generally reactionary and is a highly manual forensic process.  Aster provides a change fast solution to a perpetually changing fraud problem by looking at fraud across multiple channels and touchpoints.  Aster fraud use cases have been deployed using path and pattern analysis, and various machine learning techniques. 

Industry:  Retail, Banking Financial Services

Analytics:  Teradata Aster Analytics:  Machine Learning, Dynamic Time Warping, Naïve Bayes, Support Vector Machines, Vector Distance, Named Entity Recognition, Levenshtein Distance, Graph, and Sessionization.