Skip to main content

Siemens BT: Every analytical transformation requires a “Jonah Hill”

One of our newest SCRC partners, Siemens Building Technologies, recently shared their insights on creating an analytics strategy.  This effort was led by the Chief Procurement Officer for Siemens BT, Carl Oberland, but has recently been rolled out across the global organization.  Carl shared some of his insights from the experience of what it was like to start an analytics strategy from scratch, and provided a number of suggestions on how organizations can begin to create an analytics culture.

Carl noted that “Analytics strategy is like giving the whole organization a full body scan.  We were also highly motivated by the movie “Moneyball”, and in fact, we used multiple clips from MoneyBall to help sell the vision.  In the movie, the new coach of the Oakland A’s baseball team uses analytics to build his organization from the ground up.  One of the most important characters in the movie is the analyst, played by Jonah Hill, who ensures that all of the data is brought to bear and that the player analytics are created.

Siemens is a 50 BN Euro company with multiple operating companies, and it has an $8B Building Technology division focused on making buildings comfortable and safe and energy efficient.   There is a lot of work going on in how to use the Internet of Things in buildings to control them, and so we have a lot of investments in this space.

To initiate their analytics journey, Siemens worked with Dr. Robert Handfield from the Supply Chain Resource Cooperative to create a white paper that helped define the vision and the business case.  In addition, Carl assigned a person who could work on access databases, and began by combining procurement data for 5 divisions into a single database.  “The rest of the organization was using “old tools”, but when they saw what I was able to do with a single integrated data lake for procurement, and they saw what I had, other functions and divisions said – “I want this!” – and they asked me to do it for the whole company.”

Carl noted that “People initially have many different interpretations of what is meant by analytics.  The white paper we developed with Dr. Handfield helped to establish the mission and vision, defining what we mean by the digital platform? What is digital governance and stewardship?  We established definitions for things like Business Intelligence, defined as analytics that will tell us WHAT is happening, whereas Business Analytics was about what WILL happen in the future (predictive).  We then had to explain to different divisions that had taken offense to these definitions in our white paper, and had to resolve conflicts that had arisen due to misinterpretation of these definitions!  For instance, the guys on the R&D side were saying that “we are the technical experts because and we are doing analytics on the buildings.  We had to explain that business analytics were about running the business, not R&D.   We created a visualization platform to focus on internal operational, and customer-facing metrics, which is NOT part of R&D’s scope of work – so we not on your turf.”

“As an organization, we decided to build our analytics capability at the Business Intelligence layer – based on our belief that this was the quickest way for a Value Proposition to emerge which would get the executive team to want to continue to invest.  Since then, our efforts have accelerated – and we have moved quickly into other areas of the analytics space.  It’s great to have a lot of data but we also recognized it was more important to have GOOD quality data.  Otherwise, you have isolated pockets of data that can be used.  In a sense, creating an analytics culture is about considering opportunities along two dimensions:  data quality, and speed to value.  Where you decide to take actions becomes a cultural problem – involving how to drive a change in culture in how we capture quality data to drive a business result, and where to begin the journey.”

The way Siemens resolved this question, was to seek to create a higher “digital IQ” across all business functions.  Because procurement had started the earliest, they were at the “8th grade” level, while everyone else was at the third-grade level – and the goal was to get everyone to high school!  Carl notes that “Creating a data platform was key for us – we were worried that if we just put a bunch of tools out there – it would be a free for all – and there would be 10 people working on different apps to solve a common working capital problem!  Instead, we wanted to get control of the analytics development process, and to seek to solve problems as a standard for the whole company, to minimize waste.  Our goal was not to move to “Excel on steroids” but rather we wanted some level of control.  To create a more robust BI capability, we also recognized that we couldn’t leave this in the hands of data scientist… rather everyone had a different perspective on the same subject – and we needed to pull these different views into a centralized viewpoint that reflected all these views.   This is the vision we sold it to the CFO.”

Siemens began its journey by focusing on data governance, beginning with core data on customers, and next focusing on establishing quality data for systems operating in different areas.  This required developing regional governance by area, using Hyperion that creates financial reports, as well as Project Management Tool used by offices executing projects at universities and hospitals, etc.  The initial goal was to make sure that the team interconnected data from different places while addressing data quality.  Some initial disputes arose regarding “whose data goes into the data mart first?”   Finance wanted Hyperion data for financial analytics, whereas, operations wanted the Project Management data – so the team had to go through a reconciliation process.

An important component of the roll-out was the development of Institutional Analytics, that is, analytics that have a common standardized look and feel.  To enable this, Siemens adopted Qlikview as their visualization standard.   In many organizations, analysts spend weeks to establish their “end of quarter” charts and graphs.  The goal was to enable an executive to generate exactly the metrics and charts they wanted, any time, with a “push of a button”.  Siemens achieved this goal, and created a platform that allows all senior executives to press a button to run a meeting anytime, not just at the end of the quarter.  One executive emphasized to Carl that this capability, more than any other, allowed him to completely change the way he managed the business.  Now, sales executives are grilled on what shows up on the dashboard, and can be engaged on a weekly basis, instead of at the end of the quarter!  Another type of analytical capability, which only applies to about 5% of the workforce who will become adept at it, involves Discovery Analytics.  This refers to the ability to drill down into datasets, and explore relationships.  This type of individual requires technically oriented individuals, as well as those who understand what types of data are required to address a business problem!  Both individuals are needed to produce Discovery Analytics.

Also important in rolling out an analytics strategy is the enablement to configure data and establishing an IT resource to do so.  At Siemens, SAP feeds Hana, and ERP data has to be manipulated into financial reports, using extraction and transformation approaches with Hadoop.  Hadoop feeds data into the data entry mart, which makes it accessible through Qlik Sense and Qlik View. Director Level executives emphasized that they want to be able to do their own analysis, but don’t capture the data anywhere to produce the analysis they needed.  This required mapping needs to data, followed by technical visualization, end user inputs and assignments of data stewards responsible for capturing the data.  One of the challenges was being able to hang it all together.  To coordinate the movement and capture of data, the CFO established a digital office for the entire organization, which moved the initial effort out of procurement.  The Program office at the local level will handle regional data governance.

An important component that was deemed critical for creation of an analytics culture was the technical visual team.  In the movie “Moneyball”,  “Jonah Hill” was the key individual who was able to pull together all the data required to assess ball players.  Every organization needs a “Jonah Hill”,  someone who can cover data flowing in from human resources, legal, operations, and procurement.   Carl notes that “when you first approach departments and ask them what analytics they need…. they don’t have a clue!  But once they get a flavor for how analytics can support business decisions, they want more!   And then things mushroom, at which point you need to establish each department’s responsibility for data stewardship that must reside within those departments.  Getting each department to understand what it takes will get them involves, and each department was instrumental in picking out which data they needed to sustain and stewart for their visualization requirements, and each one also identified who the resident ‘data expert”  and visualization analyst/expert would be at a local level.  The local Data Steward is one of the most important roles, because this system is responsible for understanding what data quality looks like when it’s right, and if it can be “shoved into the system”.

Carl identified a few “lessons learned” from this analytics culture experience.

  • Start on a small scale. Pilot your efforts first in one function to understand how to move data into a data mart.
  • Data stewardship and data governance are the most difficult aspects to establish in any analytics initiative. Establishing responsibility for data quality is key, and the role of the regional/local data steward is critical
  • Don’t underestimate the skill level change that is needed for personnel to think analytically. Until they see what it looks like,  they will be lost, and cannot tell you what types of visualizations are needed.
  • Be mindful of the program becoming another “Excel on steroids”. This occurs when you have multiple people trying to solve the same business issue in siloes with significant variation in look and feel.  However, if someone comes up with a “cool app” at the discovery level and business owner likes it, escalate it and put it into the institutional view for everyone to use in a standard format.

Some of the fundamental analytics used at Siemens included the following metrics, that created some basic but critical performance insights for senior executives.  To facilitate a “push the button” analytics capability, Siemens developed an institutional analytics landing page.  Executives can go in and get a dashboards pathway – which is loaded once a month.  This makes it easy for this executive to download the dashboard, and quickly publish it in an email to the executive team for discussion.  Some of the key metrics used by the divisions included the following:

  • Project Risk Analysis – which projects have the highest risk? Almost 20% of projects are high risk, and the system can now assess this and roll up risk to an overall score to assess which ones are “red” – technical risk.
  • Forecasting – what is our trajectory in terms of where will we end up on organizational costs vs. budgeted costs
  • Visualization deep dive to determine how do we drive competitive bidding? This metric identified those areas of the business where managers were NOT using competitive bids, and we were able to show significantly higher costs in those divisons where bidding was not being used.  This metric moved the number from 45% to 70%.
  • Contribution to net income (cost out)
  • Contribution to growth (save money before we book the job),
  • Purchasing volume (can click on division and drill down into it based on volume going to small business, p-card volume, etc.)
  • Business Reviews that include gross margin, absorption, utilization, and 22 other dashboards that cover all business-related issues. One executive noted that “the most impactful contribution to the business has been the BI platform that we developed this year”.  Every executive can establish what analytics are important to their own business.
  • Budget dashboard – allows an executive to see their internal cost center, both current and forecast, and determine how he can adjust salary and fringe to impact the forecast which rolls out to the GL, which is sent to finance every month.
  • Supplier risk. This is a project which engaged a group of NC State graduate students through the SCRC, and gauges the potential an executive has to impact project risk through alternative supply.  This has allowed managers to mitigate the problem , and allows us to have a discussion down to the component level.  The discussion may involve “here is where the risk lies – high, medium or low – and on high – do you want us to go to a second source?  Do you want to invest or not – and if not – don’t blame procurement if there is a disruption that shuts down your project!”  This is a great platform that enables a discussion on supply risk with the business, and allows the sourcing manager to look at geographic lanes, and do a deeper dive to drill down down to the root cause.

Carl emphasized that “you need to find two or three “Jonah Hills” for your transformation – and once you find them, you want to make sure you keep them around, as they are valuable individuals!”  NC State is proud to have worked with Siemens in rolling out this success story, and look forward to working with them in 2019, and maybe producing a few more Jonah Hills for our organizational partners!