How to add 200 analysts without hiring anyone new

Making fact-based decisions is a challenge of many organizations.  Business and clinical leaders struggle with getting access to timely, relevant data to empower their decision making.  At Seattle Children’s we struggled with the same challenge.    We had several BI tools, a developing data warehouse and a great team of analysts and developers, but no matter how hard we worked, we simply could not keep up with the demands of our end-users in a timely way.  Resources were tight and simply adding more developers or consultants did not address the challenges.

We needed to think differently.  We needed ways to empower more self-service access to data to meet end-user needs while maintaining control and governance of standard organizational data.   We needed a better way to pull together all of our data, tools and requirements to improve our ability to create visual systems, a key tenant of the Toyota Production System, a core component of Children’s operations.  We also could not add any more staff to our team.   In our search for a better solution we found Tableau Software.

Tableau helped address many of our challenges by empowering end-users to answer questions on their own.  Users were able to leverage many of the reporting tools and data marts already in existence.  The IT and Knowledge Management teams were able to focus on building a robust underlying warehouse with timely, high-quality, governed data.   The collaboration of IT, business and clinical departments improved Children’s ability to become a data-driven organization.

From zero to 180 in 14 months

I attended the Tableau Customer Conference in August 2010 to take a better look at the tool and solutions being delivered.  I was impressed, but wondered if we really needed another tool?   There were probably only a handful of people who knew the difference between dimensions and measures and would really benefit.  After a short trial period and several passionate business users on board we purchased 20 licenses to Tableau desktop and the Server tool in October.  This should have been more than enough to get started over the first year.  We were very wrong.   By the time I left in January 2012, we had more than 180 active Tableau desktop users.

Quick wins

Very shortly after installing Tableau, business managers, directors and data analysts were able create compelling visualizations that changed the way leaders looked at data.  Answers to complex problems were delivered quickly.  The answers to long-held questions had everyone salivating for more.  Everyone got excited about data.   One doctor commented that he learned more in 5 minutes than he had in 5 years through the use of a visualization put together by a Director, with no technical development skills, in a matter of hours.  This became a universal theme where end users were learning more about their business while interacting with the tool.  Questions evolved very quickly and compelled increased adoption.

We didn’t hire new developers, but rather trained leaders and analysts in different departments who knew what their teams wanted to see.  Training was limited to a few face to face demonstrations and pointing users to online training resources.   As a versatile solution Tableau helped drive solutions to problems across Children’s including clinical operations, finance, supply chain, marketing, quality, patient safety and research.

Benefits to IT

One worry of establishing self-service access to data is that it will serve as a workaround to IT, negating great work the team has done for many years.  As a direct report of the SVP/CIO, I found the opposite was true.  Tableau helped advance our mission in several ways.

• No Increase in staff -   We carved out about .25 of an FTE’s time to support the server and get users access.
• Leveraged existing data – Users started quickly by using data readily available from other solutions such as Business Objects, Crystal and Microsoft Analysis Services cubes.
• Enabled rapid prototyping – End users could prototype and build complex solutions in Tableau without needing to learn complex coding languages or calling IT.   Users were used to exporting data to Excel and conducting analysis, but Tableau was faster and easier to use.
• Enabled focus on core infrastructure – Our limited staff was able to increase availability of integrated data in our warehouse.  The team spent more time on integration and less on time consuming requirements documentation and report development.  Data visualizations developed by end users became a new form of requirement gathering, we could focus on integrating the most pressing and well documented needs first.
• Improved security and governance – The historical approach of downloading data from standard reporting tools into Excel created several governance problems.  Once in Excel, governance and security controls were lost.   To improve their analysis end users pushed us to improve conformed dimensions and availability of warehouse data.  In the past they simply asked for more reports.  New advances in Tableau such as the data server encourage even more users to stop exporting to Excel and leave the data under the supervision of IT.

Empowering the “Nerd Herd”

I initially thought end users wouldn’t be able to figure out a complex analytical visualization tool without extensive training.  Analysts and managers have been building complex Excel spreadsheets for years.  An intimate understanding of complex business challenges and deep expertise in Excel served as a great starting point for developing visualizations.  Starting with simple problems helped users increase competence with the tool and the underlying data.  As competence increased, questions evolved and analysts could get deeper at their own pace.

To help enable adoption, we started monthly user group meetings.  I liked to call it the “Nerd Herd” because data nerds across the organization came together to talk data (as a data nerd myself, I’m allowed to say that).  The excitement and enthusiasm in the room was contagious and they started bringing more friends into the conversation.   By using simple visuals, analysts were able to engage leaders to talk about problems.  The tool disappeared and real conversations began occurring.

Bringing it all together

Bringing Tableau into Seattle Children’s helped bring all of the pieces together to empower end users and enhance IT efforts.   We still needed all of the people, data warehouses and business requirements, but we needed away to make them all come together better.   Children’s has really just begun this journey and will continue to evolve over time.  It will be exciting to see what the future holds.

Attending HIMMS 2012?

If you are attending HIMSS in Las Vegas, I will be presenting the Tableau story at Seattle Children’s on Thursday, February 23^rd at 2:45 pm.  I look forward to seeing you there.

Ted