This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
DataDelta Master Data Management (MDM) Workshop
Session 2:Session 2:MDM Case Studies & Business Benefits
Contemporary business strategies mandate flexible infrastructure –propelling MDM into “Top 10” initiative
‘Top 5’ Justifications for MDM Initiatives
1. Catalyzes market leadership & dominance2. Provides increased ROI by “leveraging / rationalizing existing
infrastructure”3. Significantly increases shareholder value4. Provides a disruptive technology for new business models5. Enables compliance & regulatory reporting
• Maximizes walletshare across product lines & BUs– Upsell via sticky bundles– New markets via cross sell to existing customer base
• Increases understanding of large customers by grouping all buying• Increases understanding of large customers by grouping all buying orgs into B2B or B2C hierarchy
• Enables “co-opetition” & electronic storefront models – e.g., B2B2C by integrating partner data with internal dataintegrating partner data with internal data
• Protects brand integrity by increasing customer satisfaction due to more focused marketing & service campaigns – e.g., “blended agent”
bilitcapability• Actualizes “consistent customer treatment” by blending channels to
MDM Provides Increased ROI by “Leveraging Existing Infrastructure”
• Enables integration of new & old channels – e.g., collections, f d t t t ith ki k ATM IVR & li lf i
Existing Infrastructure
fraud, contact centers with kiosk, ATM, IVR & online self-service• Minimizes architectural complexity to simplify application
solution design, deployment, & maintenance g , p y ,• Reduces number of interfaces between applications &
increasing reuse factor to save substantial integration costsI i f t t fl ibilit d t l t h ll• Improves infrastructure flexibility and control to enhance overall system performance
• Accelerates ROI of enterprise CRM solutions p• Reduces overall project risk through increased flexibility &
• Drives costs of “dirty data” out of the info supply chain• Accelerates revenue growth via more intelligent cross-sell & up-sell
enabled by complete understanding of customer (profile, accounts & interactions) to leverage bundling opportunities& interactions) to leverage bundling opportunities
• Facilitates quick scales of economy in mergers & acquisitions (M&A) – e.g., shortening customer, desktop, & billing integration
ftimeframes while providing scalability to support rapid assimilation of new block of customers
• Measures, manages & grows “customer information” as a key , g g ycorporate asset
• Drives fundamental operational savings & efficiencies – e.g., “once & done” enterprise wide services for key customer processes such& done enterprise-wide services for key customer processes such as account changes (name, address)
MDM Provides a Disruptive Technology for New Business Models
• Enables self-directed customer experience for sales & serviceP i i h i t t d 21 t t l h i
New Business Models
• Provisions hyper-integrated 21st century supply chain – e.g., outsourced manufacturing, outsourced service (“he who owns master reference data, dominates supply chain”)
• Automates customer transactions that flow across systems• Enables contingency planning for future technologies – e.g.,
biometrics smartcards etcbiometrics, smartcards, etc.• Provides much more than “just another customer interface”
– 21st century business application development platform– Service-oriented architecture – a.k.a, “first foray into SOA”– Web services– Integrated analytics– Integrated analytics– Near real-time materialization
• Centrally manages privacy preferences for consistent rules of visibility & titl tentitlements
• Enhances “evergreening” of customer data accuracy – e.g., continuous customer data improvement by providing self-directed customer care p y p gportals … which in turn integrate customer info across business units
• Enables regulatory reporting compliance– i.e., large customers’ material events (SOX BASEL II)events (SOX, BASEL II)
• Facilitates compliance with AML, OFAC, USA PATRIOT, et al
Customer view must prevail over product view as higher margin customers dictate common set of products & services via rapid
adoption of MDM
Retail Banking – Business Objectives
• Provide single view of customerProvide single view of customer– Across all channels – Across all products
• Reduce operational costs• Improve cross-selling• Improve net credit loss• Improve net credit loss• Increase marketing lift• Manage privacy centrally• Provide operational view for Basel II compliance• Provide improved customer service
Aggregation within industry –exacerbated by big commercial banks
Integration of new & old channels –e g collections fraud contact centers with
Business Drivers Technology Challengesexacerbated by big commercial banks emulating major monolinesHeavy reliance upon direct mail marketing & inherent increased
smartcards, etc.Complex hierarchy mgmt – e.g., household-level risk mgmtScalability e g ability to integrate new326, OFAC, SOX, …
Support future business objectives –e.g., M&A
Ensure consistent customer service
Scalability – e.g, ability to integrate new block of customersInfrastructure costs of integrating new data sources & channelsEnsure consistent customer service
across all channelsPrivate label cards needing lifestyle event based differentiationevent-based differentiationReduce IT infrastructure costs
Challenge is to mo e to po tfolio le el integ ation despite the
Consolidation (M&A) Abilit t bl d h lBusiness Drivers Technology Challenges
Consolidation (M&A)Deregulation/re-regulationIncreasingly "portable" customersSelf-directed service to drive down
Ability to blend channelsComplex supply chainsOnerous regulatory mandatesSelf-directed service to drive down
customer care costsReal-time marketing & integrated campaign management using
di ti l ti
Extreme scalability in call centers, provisioning, etc.Flood of data due to ‘datapredictive analytics
Fraud detectionBill presentation
Flood of data due to data services’“Plan anywhere, build anywhere” investment strategies in newinvestment strategies in new technologies that enable quicker new product introduction (NPI) at lower cost without sacrificinglower cost without sacrificing quality
Telco evolution will be radical as intense competition in wireless, LD, Internet & local service commoditizes products & slashes profits
• CSPs must acknowledge state-of-IT-affairs resulting from continuous M&AM&A
• CDI provides proven means to insure ROI of M&A – especially in CSP industry with historically fragmented customer & product master DBsMarket leading CSPs are planning to apply CDI as disruptive• Market-leading CSPs are planning to apply CDI as disruptive technology to “outmarket & outservice the competition”
• “Build vs. buy” has become “buy mega vendor” vs. “buy best of breed” ITO d t d t f h CDI l tias ITOs are mandated to move away from home-grown CDI solutions
to COTS solution• During 2005-06, CSPs will focus on deploying data-centric CDI
i f id ifi d i l i linfrastructure to provide unified customer view across multiple channels & LOBs
• During 2006-08, CSPs will move to process-centric CDI infrastructure to deliver increasingly accurate, complex & just-in-time unified customer views to enable bundle mktg & self-directed customer care
CSP evolution will be radical as intense competition in wireless, LD,
CSP evolution will be radical as intense competition in wireless, LD, Internet, & local service commoditizes products & slashes profits –
not to mention VoIP
Pharma/Life Sciences Business Imperatives
• Improve productivity of pharma sales reps
360° View of Customerreps
• Coordinate sales & operations planning via common infrastructure & goals/metrics Contracte Specialty
Primary Care Physicians
goals/metrics• Increase customer/physician
acquisition, retention & profitabilityR d i f t
Contracted
Accounts
Specialty Physicians
HealthCare• Reduce info management
overload/costs• Prepare for uncertain future
CareProfessional
Group Purchasin
g Orgs
Pharmacy Directors &
OtherInfluencers
Wholesalers / Speciality
Distributors
Recent survey indicates physicians need info to make more informed buying decisions – but only about 1/3 consider pharma sales reps impt sources; 1/3 found sales visits helpful, addt’l 1/3 want more
current comparative or clinical data/analyses relevant to practices, plus objective info on usage & side effects
Life Sciences / Pharma Technology Challenges
CRM alone is not enoughDW+CRM is not enoughImportance of flexible workflowDis-intermediated data sources
Data service providers (IMS, NDC, AMA - cleansing, matching issues)Industry standard identifiers (IMS, DEA, state licensing) not always accuratey ( , , g) y
Where does content mgmt system end & where does the structured data mgmt start?
Market-leading Pharmas must unify physician/customer information f l i l & b i i i b d d l
from multiple systems & business units in a robust data model specific to pharmaceutical industry
Improve Productivity of Pharma Sales Reps
• Enable sales reps to improve sales effectiveness by leveraging client sales histories & profile info from diverse sourcesclient sales histories & profile info from diverse sources
• Automatically capture physician info during sales meetings• Spend more time in the field – serving physicians – rather than in p g p y
the office recording & analyzing meeting results• Enable physicians to make more informed purchasing decisions by
the physician profile, mktg activities, etc. database centric quasi batch CDI is first step
Coordinate Sales & Operations Planning Via Common Infrastructure & Metrics
• Monitor real-time performance of sales force to quickly make strategic & tactical adjustments necessary to meet sales goals &
id d t d t i t
Common Infrastructure & Metrics
provide adequate product inventory• Employ RFID technology & the EPC Network to improve info
mgmt in pharma supply chain• Make M&A work via vital agglomeration of customer & supplier• Make M&A work via vital agglomeration of customer & supplier
master data• Integrate IT infrastructure of diverse acquisitions• Leverage info insights & relationships to expedite & enhance• Leverage info, insights & relationships to expedite & enhance
product intro & adoption • Overcome cultural obstacles to creating better collaboration
between Clinical & Marketing organizations• Insure supply chain integrity protection for producer & patient (&
everyone in between)• Protect the pharma enterprise involved in ever more
complicated contracts with external partnerscomplicated contracts with external partners
CDI needed that not only provides excellent support for sales reps,
CDI needed that not only provides excellent support for sales reps, but also makes customer info available to marketing, R&D & supply
chain functions
Increase Customer/Physician Acquisition, Retention & Profitability via Closed Loop Marketing
• Enable creation of segment-specific drug detailing aids
& Profitability via Closed Loop Marketing
• Deliver custom detailing to physicians thru field sales channel• Collect responses to product messages at point of customer interaction • Analyze responses to further refine marketing campaigns based on• Analyze responses to further refine marketing campaigns based on
Both physician & consumer data are intricately related as part of information asset programs which must keep regulatory risk &
validation costs to a minimum
Prepare for Uncertain Future
• Leverage digital packaging (RFID, ECM) for compliance monitoring & supply chain optimi ationoptimization
• Optimize consumer self-service – Apply advanced technologies to expedite info sharing via guided search, co-browsing,
tetc.• Helps consumers benefit from full range of info about health issue they’re researching • Enables pharma to deliver targeted info to customers at precise moment needed or requested
Examples: medical education; disease mgmt; patient compliance; product awareness– Examples: medical education; disease mgmt; patient compliance; product awareness• Drive & manage regulatory compliance across diverse legal domains• Prepare for increased outsourcing of both commoditized & specialized functions • Exploit strategic opportunities created by bioinformatics/ nanotechnology• Leverage re-importation trends
Asked “What kinds of surprises & challenges did you face, and how didAsked What kinds of surprises & challenges did you face, and how did you overcome them?” – in areas such as:
Other general topics:Other general topics:• Project goals & Key Results/highlights• Who uses/depends on the results where & how?• When was value delivered & how was it measured?• Single "Big Bang" project or multi-stage phased implementation?
B i C & E ti S• Business Case & Executive Sponsors• Project Ownership, Management & outside assistance• Solution Description (people, process, technology & data)Solution Description (people, process, technology & data)• Surprises & Lessons Learned
• SIs & consultants can add strong value even at this initial stage by enhancing structure & credibility
“Phase Implementations” are proving easier to win funding and• Phase Implementations” are proving easier to win funding and successfully execute
• CDI-MDM now being so widely identified as a critical strategic business objective that some business cases are coming from, instead of being sold to Sr Mgmt (e g “Why aren’t we doing thisinstead of being sold to, Sr Mgmt (e.g. Why aren t we doing this yet?”)
• But mid-level mgmt sometime using business case to avoid support (“I need a business case justification before I can change my priorities”)
• You are guaranteed to find new problems & issues, despite past investments in data quality
Extensive data profiling early is essential• Extensive data profiling, early, is essential
• Trust, but verify, “trusted” third-party data sourcesust, but e y, t usted t d pa ty data sou ces– Extensive data profiling is again warranted– Ensure vendor’s definitions map to your definitions
“Total employees” by subsidiary or total hierarchy?• “Total employees” – by subsidiary or total hierarchy?• “Total revenue” – by subsidiary or total hierarchy?
• “Match Accuracy” is different, but highly dependent upon, data quality
• Sometimes success can occur without a formal Data Governance program & full-time Data Stewards
But in general the more complex the environment (# data siloes• But in general, the more complex the environment (# data siloes, # LOB users) the more important to have formal Data Goverance to resolve conflicts
• Data Governance requires data domain experts (business users) who are a scarce resouce and thus guaranteed to have frequentwho are a scarce resouce and thus guaranteed to have frequent priority conflicts
• Strong value being reported from metadata tools (as opposed to “just using Excel to define & map fields”)
• Internal Education & Project PR– Not just during business case, but ongoing recruitment– Not just for initial collaboration recruitment, but also to help maintain
positive momentum & excitement around projectpositive momentum & excitement around project– Not just internal use, but recruited by Aaron Zornes for future CDI-
MDM Summit (leverage that work for more value)Leverage CDI MDM Summit material for own ppt– Leverage CDI-MDM Summit material for own ppt
• Sophisticated metadata managementp g– More than just “mapping fields in Excel”– More than just a “web wrapper around Access”
More Sr Mgmt is more often recognizing the value of modern– More Sr Mgmt is more often recognizing the value of modern, sophisticated metadata tools in the business case
BackgroundBackground• 2006 started CDI project with $7M budget• 32 global customer data sources• Encouraging results half-way through• Realized opportunity to address other problems
D id d t “t k thi t th t l l”• Decided to “take this to the next level”• Use full data governance and full data management• Opportunity to save lots more moneyOpportunity to save lots more money
Business case:Business case:• Analyzed all 2007/08 data projects• Determined cost of siloed approach of $22M• Estimated could deliver for $14M over 2 years via CDI-MDM• 1 “data champion” developed case in just 2 months
Full time time job– Full-time time job– get the numbers, crunch them, validate them, get rates from vendors, create
ppts, etc.– Had to press for numbers – used sponsor name for meeting leverage “X will
be facilitating the meeting for CIO…”– Some people would panic at thought of their local data silo eventually going
“Oh God, let me give you an example of the kinds of things we wereOh God, let me give you an example of the kinds of things we were finding:”
One of the groups only had a 40% 30% complete address• When asked, response was “when missing, we use our HQ address”• Controller response “Are you out of your mind? We can’t calculate our• Controller response Are you out of your mind? We can t calculate our
taxes and we’ll be fined!”
“So the process they were doing locally could have caused a major impact to the tax group and nobody knew about it and the dollar amount put on top of that could have been huge in fines.”amount put on top of that could have been huge in fines.
Full escalation process established to address conflicts:• “My business doesn’t operate that way, that’s not the way I report my
numbers to my boss” – Local Business“This is the way it’s supposed to be that’s the way we report our• This is the way it’s supposed to be, that’s the way we report our numbers to the street” – Enterprise
Steering committee• One of the first things done
D fi d h i i f h fi k l• Defined the organization as one of the first key goals• Started meeting even before program 100% formalized• Established full escalation hierarchyEstablished full escalation hierarchy
Data Governance Hierarchy• Steering Committee (8 people, monthly) - program decision rights
makers around funding, prioritization, staff and escalation issuesData Governance Council: reps from legal tax credit finance audit• Data Governance Council: reps from legal, tax, credit, finance, audit, stocks, etc. (6 people, on demand) - data decision rights group that is purely making decisions on data, data policies, etc
• Subcommittee - a representative of the council yet at a lower level…someone who has more hands on experience with certain issues based on the subject of the dayj y
• Data Management Team– Business “data champion”– IT manager– Project managers, data stewards, data warehouse architects, ETL
• Early success with initial customer hub project was critical“Wasn’t just about creating a customer hub because anybody can buy a
product and load data”
• Value in understanding:– how good or how bad the data was– how good or how bad the processes were around that data,– how inconsistent people within different groups and different systems treat
customer data
• True top down type governance – exec sponsor(s) making sure that l d t d th ’ itt dpeople understand they’re committed
“We couldn’t do this from the bottom up, it would never work. There are too many people in the organization over the last couple of years where the
Business CaseBusiness Case• Strong Sr Mgmt support allowed faster, general business case –
detailed, long-term ROI wasn’t required
• But middle mgmt less satisfied with strategic value– Often challenged for hard ROI numbersOften challenged for hard ROI numbers– Focused on adding value now instead of 1 year from now
• Phase 2 – Incorporate Analytical CIF– More complex hierarchical household matching– More likelihood of conflicts = greater need for governanceMore likelihood of conflicts = greater need for governance– Defering data augmentation until Phase 2
• A single, source system used for decadesA single, source system used for decades
• Everyone assumed there were no surprises left in data
• In fact, there were a number of surprises
“Originally we had no intention of doing much data profiling, but in hindsight that was a bad decision . . .”g
“A lot of people in the project understand the technology but don’t understand the data and how it’s used on a day to day basis acrossunderstand the data and how it’s used on a day to day basis across our channels and across our users.”
“Phase 1 only has one source system the legacy CIF people are used toPhase 1 only has one source system, the legacy CIF people are used to using . . .”
• Currently do not have a Data Steward (in IT or Business)
Still h h d t l j t t hi & i• Still have had some struggles on just matching & merging– Resolved with weekly all-day meetings over a couple months
• Data Steward & more formal governance planned in future phase
• Business Case – stick to top handful of benefits (you can drive yourself crazy with a thousand different benefits)yourself crazy with a thousand different benefits)
• Executive support is criticalecut e suppo t s c t ca– It’s not a 12-24 month project and then your done.– It’s a 10-20 year investment
• Lay out strategy long before you start anything else (several people over 3-4 months))
• Planned tightly controlled rollout in 2008 (users, dependencies, etc)– but everyone now saying “me first!”
D f i l it t Ph 2 (h h ldi t ti• Defering complexity to Phase 2 (householding, augmentation, analytical support, etc.) – Greatly helped maintain focus & delivery deadlines– “one of the best things we’ve done in hindsight”
SIs may want to do things the quickest and easiest way which might• SIs may want to do things the quickest and easiest way which might not be the best way for us long term
• Had to ask Sr Mgmt to step in a couple times early on to resolve priority disputes for data domain experts’ time
Background:• Already had a centralized database• Loading a lot of data making broadly avail to the org• Originally only planned a simple “Data Steward” program• But quickly realized that there was a lack of governance throughout
the organization on datat e o ga at o o data
Business Case:• (some) Sr Mgmt realized need to transition from being just a “data
steward team” to providing data governance for the entire organizationg
• Used “grass root” support of middle-mgmt peers to help convince their bosses
• “We identified some major issues where the same field in the database is being use for four or five purposes… “database is being use for four or five purposes…
• “difficult to measure quality - we’re fighting every step”
• “if you really want quality data – it’s going to take years, it’s an iterative processiterative process.
• Trying to maintain independence of the business and technologybut partnering with both groups in order to get things done– but partnering with both groups in order to get things done
• Domain steward – 1 for each system of recordy– updated job descriptions so it becomes part of their ongoing
responsibility, such as maintaining metadata, and then it’s reflected within their annual performance appraisal.p pp
“The great thing is the way we’ve set up our governance structure b it t i l d th t k thbecause it goes to a single person and that person can make the decision rather than trying to get together 40 bodies”
• Get the governance structure set up and bought intoA i t d t l i l ll f t d t t– Appropriate processes and controls in place as well, from metadata to rolls and responsibilities
– Defining Data Stewards is not enough
• Pull in an SI to help get it done– “No sense bleeding through things you could have someone come in and g g g y
in 3-4 weeks, jump start your program by 24 months.”– Also helped in terms of credibility - more reassurance to Sr Mgmt
• Invest significant resources in metadata repository– Plan more time, effort & tools than “just mapping fields in Excel”
“You’re always at risk of losing that smart guy who knows how the system works and what each one of those elements meant. As soon as that guy walks out the door you lose all of that history ”
BackgroundBackground• Started as a “Data Warehousing” project• Evolved into an enterprise “Single Customer View” project• Early results were disappointing & hurt credibility• Uncovered need to resolve internal priority conflicts
350+ d d lt d• 350+ users depend on results every day
Business case:Business case:• Fundamental issue of “same” reports from different departments
providing different numbers• Informal business rules could change “on-the-fly” within departments, so
even reports from the same department may not agree month-to-month• Started as a focused Marketing effort to prove value for taking the nextStarted as a focused Marketing effort to prove value for taking the next
step at the enterprise level• Driven by Director of Business Intelligence• Implemented over a series of 3 project to incrementally achieve