You Can Have It All The Siemens PLM Journey to Self-Service BI Bryan Hunter | Siemens PLM Software Program Director Business Intelligence
Todd Talkington | Tableau Sr. Tech Partner Manager
For internal use only / © Siemens AG 2013. All Rights Reserved.
Siemens PLM Software BI Presentation
Disclaimer The BI Strategy discussed in the following slides are unique to Siemens PLM Software and are not necessarily representative of Siemens AG Corporate BI Strategy.
Restricted © Siemens AG 2016 Page 3
XX.XX.20XX
Siemens PLM Software
Once Upon a Time there was a company that loved Excel…
Restricted © Siemens AG 2016 Page 4
XX.XX.20XX
Siemens PLM Software
…they owned every Excel-based reporting tool known to mankind…
Restricted © Siemens AG 2016 Page 5
XX.XX.20XX
Siemens PLM Software
…acquisitions were sad because of the complex reporting silos…
Restricted © Siemens AG 2016 Page 6
XX.XX.20XX
Siemens PLM Software
…but the ever-wise CIO knew there must be a better way. So he formed a new, highlyskilled team to change the culture… Restricted © Siemens AG 2016 Page 7
XX.XX.20XX
Siemens PLM Software
…the challenge was issued… • Self-service and Cross-application reporting new goals • Friendly for end-user and developer to use • Analytics and Visualization would be the way forward • One Source of the Truth - reduce data movement and shadow IT by creating Enterprise Data Lake • Fast - utilize Greenplum database’s speed Restricted © Siemens AG 2016 Page 8
XX.XX.20XX
Siemens PLM Software
…the steps set… • ETL - procure enterprise ETL tool to meet data movement and transformation needs • Reporting – find enterprise reporting tool to meet end-user self service, ad-hoc and visualization reporting needs – host reporting tool “bake-off” • Data Management - minimize data transformation while bringing in disparate data sources; provide consistency across enterprise
• Governance - protects data yet is flexible enough to allow business to do self-service and ad-hoc reporting Restricted © Siemens AG 2016 Page 9
XX.XX.20XX
Siemens PLM Software
ETL – Extract, Transform and Load • Informatica PowerCenter was obvious choice – industry leader - no “bake off” needed • High-speed connectors can move millions of rows of data very quickly out of source systems such as SAP, BW and SFDC into Data Lake • Informatica Cloud also used for some smaller, less complex needs
• NOTE: Informatica’s footprint continues to expand – has become “tool of choice” for data movement Restricted © Siemens AG 2016 Page 10
XX.XX.20XX
Siemens PLM Software
Reporting Tool “Bake-off” Vendor
Tableau
WebFOCUS InfoAssist 8
Business Objects
Qlikview
PROs Easy enough for business analyst to use Direct data connectors Reporting experts really liked product Analytics and Visualizations tool of choice Does not slow Greenplum DB down – low overhead Good support for row-level security Scales for enterprise rollout Decent Admin tools Reasonably priced
CONs
Focused on analytics and visualizations – not necessarily a good choice for operational reporting
Existing WebFOCUS environment Inexpensive Operational Reporting tool of choice
Limited Ease of Use IT support required
Existing BO environment Great Operational Reporting tool
Limited Ease of Use IT support required Current environment not scalable Expensive
Produces nice-looking reports
Limited ease of use - scripting is required In memory engine required – cannot use Greenplum DB directly Expensive
Restricted © Siemens AG 2016 Page 11
XX.XX.20XX
Siemens PLM Software
Data Management • As mentioned, Informatica uses high-speed connectors to bring millions of rows of source data into Data Lake • Data Staging process transforms transactional data into “analyticready” data: • • • • • •
Flatten out / de-normalize data Add friendly, consistent column names Ensure column values consistent Snapshot data for trending analysis Perform automated error checking Rollback updates if data updating fails per group
Restricted © Siemens AG 2016 Page 12
XX.XX.20XX
Siemens PLM Software
Governance Data loads from source systems and ease-of-use verified by IT and signed off by the business
Greenplum Data Lake
Tableau Workbooks can only be created or modified by authorized users. Authorized user can use Certified or Uncertified data sources
Tableau Desktop
IT publishes all Certified Data Sources at request of business
Tableau Server
Tableau Workbooks are published only after data is certified and Workbook has business SME approval
Tableau Workbooks
Restricted © Siemens AG 2016 Page 13
XX.XX.20XX
Siemens PLM Software
Put it all Together…
Tableau Desktop (Super Users)
SFDC
Tableau Server Other Reporting Systems
O2O
OpenAir
Informatica Extract, Transform and Load (ETL)
Greenplum Data Lake
SAP Restricted © Siemens AG 2016 Page 14
XX.XX.20XX
Siemens PLM Software
There were bumps in the road… • IT Team very uncomfortable – wanted more detail when we were in fact “making it up as we went” • Difficult to manage expectations
• Impossible to manage timeline – art not science • Lots and lots and lots of technical hurdles to overcome • Plenty of people challenges as well – strangely, each business department has a different personality Restricted © Siemens AG 2016 Page 15
XX.XX.20XX
Siemens PLM Software
…some good decisions too… •
CIO – project sponsor
•
Professional Services helped verify technical installation of product and gave us excellent “best practices” help
•
Training IT Team was extremely important to get team up and running as quickly as possible
•
IT Team “sequestered” for first year as we worked through standing up reporting solution
•
Had real-world Use Case to work through
Restricted © Siemens AG 2016 Page 16
XX.XX.20XX
Siemens PLM Software
Business Reporting Team •
Executive Reporting Leadership Team – sets overarching direction for Enterprise Reporting. Standing this up now – not needed in the beginning
•
Business Stakeholders / Reporting Council– responsible for reporting in a specific area of the business. Provides IT with Reporting Demands (data and, in some cases, Workbooks).
•
Super Users – Business users that can create reports and, in some instances, combine Uncertified data with Certified data
•
Data Czar – responsible for Enterprise Data Cleansing/Monitoring and reporting-specific business process around DUNS, Industry, Product Hierarchy, etc. Must understand and has authority to require changes to business processes and Transactional systems where needed to support reporting and other such initiatives. Do not have one today, but hope to convince Executives why Data Czar is needed
Restricted © Siemens AG 2016 Page 17
XX.XX.20XX
Siemens PLM Software
IT Reporting Team •
Demand Manager – conducts periodic project prioritization with Business Stakeholders to prioritize all Enterprise reporting demands; works closely with Business Stakeholders to ensure alignment in developing Enterprise reporting area vision
•
Project Manager(s) – works with different areas of the business managing business expectations and reporting projects
•
Data Modeler(s)– develops overarching database architecture for providing Certified and Uncertified data to appropriate Business users, using “data in place” and ETL to Stage and Normalize data into Data Lake
•
Reporting Tool Team(s) – creates reports, provides support and administers security for Operational reporting tool
•
Meta Data Manager – manages Data Certification, Data Cleansing, Data Normalization and provides Business users with Data Guides on all Certified data sources; works closely with and supports Data Czar
•
Data Lake DBA – technical resource that manages Data Lake
•
Data Management Team – provides support and administers security for Data Management tool(s); provides Data Blending, Data Staging, ETL, Cleansing/Monitoring, Data Normalization and supports Meta Data and Data Modeling
Restricted © Siemens AG 2016 Page 18
XX.XX.20XX
Siemens PLM Software
Overall Roles and Responsibilities Business – Data, Workbooks and Priorities: • Owns data and business processes • Verifies “ease of use” of analytic-ready data • Approves/denies user security requests • Responsible for creating workbooks from published data sources • Works with IT in training end users • Sets priorities for what data is transformed when by IT team IT – Data Transformation, Governance: • Performs business analysis on data needed per the business • Extracts, transforms and loads analytic-ready data into Data Lake • Implements security per business requirements • Ensures governance of data source and workbook publishing • Creates “starter” workbooks when needed • Works with business in training end users Restricted © Siemens AG 2016 Page 19
XX.XX.20XX
Siemens PLM Software
Data Alignment Matrix •
Gather each slice of data needed by each Business Stakeholder
•
Add to Data Alignment Matrix
•
Prioritize based on overall need
Restricted © Siemens AG 2016 Page 20
XX.XX.20XX
Siemens PLM Software
Priorities – More Work Than People… •
Each Business Stakeholder prioritizes their work
•
Meet with all Business Stakeholders – prioritize overall work
•
Resource availability based on overall priority
Restricted © Siemens AG 2016 Page 21
XX.XX.20XX
Siemens PLM Software
Data Governance Process • Uncertified Data – is often brought into Tableau as Super Users do self-service prototyping and ad-hoc reporting. This is encouraged because the business will iterate the Workbooks before requesting data to be Certified. • Certified Data – data must be verified and signed off by the business before it can be used in building and publishing a Workbook. • Data Requests – are made by the business in the Help Desk ticketing system • What Data is Most Important? – prioritizing what data moves into the Data Lake when is the most important item to prioritize. We look for “sweet spots” where multiple Business Stakeholders need the same basic data to help guide prioritization • ETL – is 80-90% of the total effort required
• IT’S ALL ABOUT THE DATA.
No Data = No Reports. Bad Data = Bad Reports
Restricted © Siemens AG 2016 Page 22
XX.XX.20XX
Siemens PLM Software
Security Request Process 1. Requestor fills out Help Desk Ticket and attaches Manager’s approval 2. Data Owner(s) reviews and approves 3. Tableau Account set up, Requestor added to proper Group(s) 4. Row-level security added (if needed)
5. Requestor notified, Help Desk Ticket closed Restricted © Siemens AG 2016 Page 23
XX.XX.20XX
Siemens PLM Software
Summary 1. It’s all about the data 2. See #1 3. Start small – try it out on a small scale first 4. IT and business must be partners or all bets are off 5. Governance must protect data yet allow for business to prototype and do ad-hoc reporting while not hurting anyone 6. Never underestimate the power of Executive Sponsorship Restricted © Siemens AG 2016 Page 24
XX.XX.20XX
Siemens PLM Software
And now…the Unicorn! Certified
Uncertified
Essbase
Products
SAP
Access
Write- Back
OpenAir
Operational
O2O
Discovery
GTAC
Excel 3rd Party
Data Management
SFDC
Data Lake
Restricted © Siemens AG 2016 Page 25
XX.XX.20XX
Siemens PLM Software
What’s Next? Check out more from the Tableau Enterprise IT Webinar Series tableau.com/learn/series/enterprise-it-summit-webinar-series Give Tableau a Go tableau.com/products/trial Informatica & Tableau tableau.com/tableau-and-informatica Informatica informatica.com
Restricted © Siemens AG 2016 Page 26
XX.XX.20XX
Siemens PLM Software