HEALTHCARE MEETS DATA & ANALYTICS

The Problems

The current state of Supplemental Files (SF) analysis revealed that for quality measures appearing in the file, accounting for exclusions was absent, resulting in inaccurate number of patients being computed for the denominator. Moreover, lookback periods were the same irrespective of the measures, resulting in erroneous population counts being considered. For some measures, the age eligibility criteria were inaccurate. Many a times, rejection of the SF would happen due to inconsistency in formatting as per the payer’s request. Additionally, the time taken to develop and the knowledge to create it was not shared and resulted in long times to develop and QA a new SF or revise an existing one.

Our Solution

Even though Breast Cancer Screening (BCS) is a completely different measure than Colorectal Cancer Screening (COL), NCQA describes each of the HEDIS measures in terms of its components viz. Measurement Dates, Eligibility criteria (age, gender, lookback period), diagnosis, numerator, denominator, and exclusions. Hence, a component-based measures architecture was created to define each measure with its nuances. Similarly, a data structure encompassing the data loads from NCQA was created which required minimal changes in the ETL process to load the measures and their value sets specified every year. Additionally, the relationships between SFs, Measures and sub-Measures, Health Plans, Payers, etc. were designed to be table driven. 

EMBARKING ON ANALYTICS JOURNEY BY HARNESSING DATA IN A DATA WAREHOUSE

» Self-service reporting
» Well defined governance & organization structure
» Better predictability
» Increased productivity & alignment
» Better negotiating leverage
» ​Strong data harnessing foundation for future

INSTRUMENTING CAMPAIGNS FOR MARKETING ANALYTICS

» ​Healthiness campaign architecture
» ​Understanding of Patient states 
» ​Marketing metrics for any campaign
» ​Strategic thinking anchored in analytics
» ​​KPIs to explain the real value of campaigns

HEDIS QUALITY MEASURES COMPLIANCE AND MM$ SAVINGS USING ANALYTICS & AUTOMATION

» ​Improved compliance from 43% to 100% : $1.5M+ shared savings per year!
» ​High degree of automation
» ​70% complete for new SF
» ​Better predictability
» ​Better negotiating leverage
» ​Measures-driven architecture
» ​Risk-based QA
» ​Strong data & analytics architecture foundation for future

INSURANCE DIRECT MARKETING SAVINGS

The Problems

Client gets Leads information from 40 feeds across 112 data sources from more than 30 associations. Ability to handle the data loads in different data formats arriving at different frequencies involved mostly manual processes, along with some combination of SQL Scripts, MS Excel, MS Access, and Bulk Mailer Software. Name identification and Matching were manual in nature, without much of data quality checks.
 
Top 10 associations accounted for more than 90% of business, but management had no visibility on completion of the Loads. Hence, it became difficult to plan outbound marketing campaigns for their members without much certainty. To make matters worse, overloading of tasks due to adhoc requests coupled with availability of personnel, contributed to a huge backlog of feeds, some of which never got handled.

Our Solutions

» ETL processes resulting in turn around time of Load process in few hours versus greater than 3 days!
» More than 95% of Leads automatically loaded as STP in seconds resulted in estimated cost savings of 50x over 5 years!
» Complete management transparency in Load process
» Automated Lead Name and Address identification and matching
» Strong Data Architecture and Analytics foundation for future

INSURANCE MARKETING EFFICIENCY WITH AUTOMATION
» Faster Customer Acquisition – Quick STP processing with no breaks or restarts due to distractions or other priorities!
» Well defined Governance & Organization Structure - Analytics Architecture, Processes, Scripts, and well-defined roles have enabled right governance structure for success
» Better predictability – Better Planning for future feeds and appends resulting from increased transparency into Load process
» Maintaining History – Improved Association interaction communicating data changes over time
» Accurate Mailings – Less disgruntled members by reducing human errors in name identification with exact and fuzzy matching
» Reduced Loss of Sales to Competition – Monthly mailing to Prospects enabled, Elimination of backlogs of feeds, enhanced documentation, reduced learning curve
» Strong Foundation for future – Additional Associations and feeds could be automated lot faster now due to automation 

MANUFACTURING - HOW DO WE LAY A BETTER DATA & ANALYTICS FOUNDATION FOR IMPROVED FORECASTING?

The Needs

One of the largest chip manufacturer in the world wanted to
​» Improve the forecasting of licenses & features
» Provide a way to input and compare forecasts and improve them over time as well as for the system to forecast and tell Consumers/Users
» Do a Trend Analysis (multiple reports) of historical usage of licenses & features (slicing and dicing across many dimensions)
» In a subsequent phase, also do better forecasting on CPU, Storage
» Eventually, reduce the time taken for milestones from prototype to final to sign off
» Build a scalable and simple solution to achieve the above for all business users

Our Solution

The client’s users, projects, data and analytics needs had grown tremendously and especially so in the last couple of years. In order to meet the needs of future growth, they engaged with us to explore how their data and analytics environment could scale up to better forecasting the current and future needs.
AAG conducted Phase 0 of Data & Analytics Assessment for client’s current & future needs. Apart from presenting the findings and recommendations, AAG delivered a roadmap with a proposal for a phased implementation of future state.

CHIP MANUFACTURER UNDERGOES CURRENT STATE ASSESSMENT

The Assessment resulted in the following Benefits for Client:
​» Comprehensive Current State Assessment & Phased Future State Proposal
» Current State Observations, spanning
   - Data & Analytics
   - Processes & People
   - Tools & Technologies
» Recommendations for Future State
   - Technology & Data Harnessing
   - Harvesting Value & Algorithms
   - Phased Tracks Approach
» High Level Data Flow Architecture
» High Level Requirements (BI Reports)
» Roles, Investments, Roadmap
» Strategy to address gaps
» ​Governance model

IT MANAGEMENT - "DATA SCIENCE DRIVEN"

The Needs

​Spiceworks needs included
» Identify data science to be a driving product
» Uniquely identify entities in their system
» All business consumers to have a “common definition” of entities/terms
» Build a scalable harnessing platform to achieve harvesting needs of all business users
» Create a system to do historical trending
» ​Define organization structure (Product, Engineering, Data Science etc.) to meet all the above needs and align with 3-year goals

Our Solution

Spiceworks embarked on a journey to spearhead data science to be more of a driving product than a reactive product through building an organizational structure and robust platform to harness and harvest data.
Analytics Advisory Group was able to recommend a robust and sustainable model to Spiceworks in order to continue addressing the needs of business users, clarify entities within the Spiceworks system, and increase value creation to keep Spiceworks in the forefront of the technological sphere.

SPICEWORKS WANTED TO BE LOT MORE "PROACTIVE" AND "DATA SCIENCE" DRIVEN

After conducting a Data & Analytics Assessment across different teams in Spiceworks, we were able to recommend the following:
​» Focus on Data Governance
A separation of “core” vs. “non-core” activities, identification of roles (especially in Architecture) following the C.A.M.P. model, and a separation of “what” and “how” is needed.
​» Employee Analytics Training
An emphasis on Tableau, advanced analytics, and text analytics (NLP exploitation) enforces a strong analytical background for future data science initiatives.
​» Enforcement on Business Identity and History Maintenance
​» Need to “Data Instrument” for “Personas”
​» Switching to More Robust Software Applications and Storage
Switching on premise to the Cloud, delving into Python, Kafka, and SQL, cementing a data management platform would prevent old hardware failure and update systems to be more sustainable for increasing data needs. 
​

AAG Clients

Past Clients - Experiences & Capabilities

Partial List of Analytics Clients served in the past 
INSURANCE
     North America:
     Hartford, London Life (GWL), Chubb, Liberty Mutual, State Farm, CNA, USAA, AIG, Farmers Insurance, GenRe, SwissRe, TransAmerica
     ​Europe:
     Aviva, Allianz, AXA (UK), AEGON

RETAIL
     H.E.B., Delhaize Group (Food Lion, Delhaize, Hannaford Brothers), Roundy’s, Whole Foods
     JC Penny, Limited Brands
     Walgreens, Rite Aid, Eli Lilly
     TxDOT, ToysRUs, Best Buy, Lowes

HEALTHCARE
     Kaiser, Aetna, CHS

BANKING AND FINANCIAL SERVICES
     American Express, Morgan Stanley, Bank of America, JP Morgan Chase

TELECOM
     ​AT&T, Verizon

© 2015-20 Analytics Advisory Group, LLC. All Rights Reserved.