One of the great things about KPMG though is that it gave me a lot of networking connections. I leveraged that to land a position at the Piedmont Heart Institute (PHI) doing project management and business development work. I did manage to greatly increase my Excel skills while at KPMG and that was pretty useful in helping me succeed at all the analysis work I found myself doing at PHI. I started following a few Excel and data visualization blogs to help me further enhance my skills. Some of them included chandoo.org, clearlyandsimply.com, peltiertech.com, excelhero.com/blog and junkcharts.typepad.com.
While at PHI I built some pretty sophisticated stuff in Excel. I built things that helped me optimize physician schedules, track patient throughput in our cath and EP labs, identify candidates for EP implants, do project portfolio management, identify new physician office opportunity areas, and many more interesting things. But with each new tool I developed in Excel, maintaining everything started to become onerous. I had macros built that would automate data pulls, formulas that would merge data sets and charts that would auto populate into PowerPoint for presentation, but it still was difficult to maintain. And God help the person who would come after me that would need to figure out the complex web of technology that I had woven. I needed a better approach.
Reading through comments on the blogs above, I kept seeing people recommend Tableau as an easier way of accomplishing the same data presentation tasks. I was very proud of my Excel skills and took it as a challenge to prove that Excel was a superior BI tool. So I started re-engineering my Excel work in Tableau. Being new to the tool, I was fairly limited in my imagination of what was possible. I tried to make Tableau be Excel (not the best idea). Still, there was enough familiarity with the features to accelerate my learning because a lot of Excel skills are easily transferable to Tableau. Tableau's VizQL interface is essentially pivot tables on steroids. At the same time, I discovered authors like Stephen Few and Edward Tufte who opened my eyes to the science of data visualization. Up until that point, I basically displayed data in a way that I thought looked cool in a presentation.
After converting a few of my projects to Tableau I was convinced of my own ignorance and submitted that I was wrong in assuming Excel was the best BI tool. (side note: Excel still has a place, but mostly I use it for light data capture and data transformation work. I leave the presentation layer to Tableau.) Now that the scales were removed from my eyes, I started gaining huge efficiencies in my workflow. I was able to completely automate a lot of my work and offer the end user an interactive data exploration experience through dashboarding. Subsequent questions were built as drill-downs to more detail. And if I didn't anticipate all the questions, I had a lightening fast ad-hoc analysis tool to answer questions on the fly. I no longer walked into meetings with a PowerPoint presentation with just static aggregated information. I walked in with my laptop and an interactive dashboard. No longer would I be in a situation of having made a wrong assumption which required a few days of re-work and another meeting to address. I could fix the issue on the fly, in the meeting. The incredible speed to insight meant that decisions could be made in a single meeting instead of multiple meetings. My audience went from sitting back and talking broadly about ideas to leaning forward and engaging in the process of finding the answers to specific questions. A question would be asked and immediately answered. That would lead to more questions that would also be immediately answered. And usually in the same meeting we would have a draft of a dashboard that would automate that analysis process with new data each day. It was life altering and I was addicted.
My desire to understand the science of data visualization had a big impact as well. It enabled me to shift perspectives and offer a fresh look at the same data sets leading to new insight. Tableau was like a natural extension of my mind in this new world that helped me fail quickly as I iterated through different designs. On the data modeling and data quality side of things, I was able to easily identify mistakes and correct them quickly because the data was at my fingertips ready to tell me everything in an instant.
I finally knew what I wanted to do with the rest of my career. It wasn't just a job. It became a passion. My passion and project success eventually got me noticed by the corporate Financial Planning and Analysis team. They needed some help with monthly operating reports that took a week to compile each month using Crystal Reports, Excel and Adobe Acrobat. The results were printed in a 133 page book, one for each of our five hospitals, and shipped on a truck 2 weeks after month end close to company leadership. I wanted to help, but I had a bigger vision. I had experienced the work that goes on in the trenches by analysts trying to make sense of data. I wanted to help everyone in the company do the same things that I could do at PHI. My new bosses decided to take this journey with me and let me run with the idea.
So I accepted the position to lead the BI program for Piedmont Healthcare. We quickly automated the monthly operating reporting so that it updated every day in the form of interactive dashboards. People subscribe to them and receive them each day in their email. If they have subsequent questions, they just click a link and use the drill-downs to answer subsequent questions. And if they need further ad-hoc analysis, their analysts have access to the same deaggregated data sources that drive the dashboards. Life is good!
Now we've moved on from financial analytics to other areas of neglected or onerous insight. There are more than 130 people across the enterprise who have been trained and have the ability to develop reports from 100+ modeled data sources containing more than 25 billions data points. They have published more than 3,000 reports/dashboards that are being consumed by more than 350 people each week. We are doing more clinical and operational analysis now, which is really exciting to me because Piedmont Healthcare's mission is to better the lives and healthcare experience of our patients. If my work can have a positive impact on our core mission, I'm thrilled! We've made a ton of progress, but have only scratched the surface of what is possible in the almost 3 years I've been in this role. The future is bright with possibilities to transform healthcare with analytics and I'm so fortunate to have stumbled into this career. For those looking into this field, I say go for it! There has never been a better and more exciting time to be in this space.