Here’s what no one tells you about leading a Business Intelligence revolution.

Contents

Key Points. 1 Accomplishments: 1 Excel Only Gets You So Far. 2 Getting to the data. 4 The Doldrums. 5 A Path forward. 6 Sprinting to Success. 7 The Final Barrier. 8 Conclusion. 9

Key Points

· Technical and professional leadership is critical. You cannot be only technical or managerially focused. You must be able to get into the weeds and know what good looks like and shift to high-level summaries and build influence.

· The goal of BI is to provide a holistic view of the business.

· Success requires a substantial personal investment of time and energy to learn and apply new skills.

· Finance must move away from a task focus and being reactive and focus on systems thinking and being proactive.

Accomplishments:

• Designed and delivered 10+ hrs. of daily data automation of analysis in Power BI with a cost savings/productivity enhancement valued at $154k/yr.
• Successfully led a pivotal project to refocus business processes around measuring critical KPIs in controlling, procurement, production, & supply chain.
• I rebuilt the controlling function, capabilities, & processes to ensure accuracy, trust, & completeness.
• Was able to answer long-outstanding business questions around product/customer profitability, break-even volumes, machine efficiency, labor efficiency, and material yield.

My Story

July 2018- Oct 2020

At Savencia, before my start, the previous controller had been fired. Although this person had been with the company for several years, the controlling function’s deficiencies were a career trap. This person tried to take the previous mess of short-term controllers before they had left and failed. She reported variances that made no sense, publishing month-end results that were inaccurate, and did not connect with the business’s needs on any level. Although she had general business experience, she lacked the essential skills and abilities that modern finance runs on.

It wasn’t her fault entirely. Before the last controller, there were at least three others in the position within four years. When I later peeled back the layers in this project, there were traps and pitfalls at multiple levels, some that even I fell for. For example, weak and disparate systems, poor business processes, a lack of access to business data, a misconfigured ERP, a lack of IT support, and a weak overall foundation were all present. The result was that business users were dissatisfied and expressed deep distrust with finance.

To solve this problem, the organization hired a new CFO and gave him the order to make fixing this mess the highest priority.

This is where I came in.

As Operational Controller at Savencia, I would oversee three manufacturing plants and a team of two other controllers in a high-visibility role where we serve a critical need to report and communicate the manufacturing performance to the plant leadership teams.

The first few weeks and months were interesting, to say the least. Just about all processes were lightly documented or, better yet, not documented at all. I had to work my way into understanding what we were doing and how it was being done. Each Excel-based process for doing anything was a labyrinth of confusion. It was like an insane person designed a map that only they could read.

Reporting was backward-looking, flat/static, and only came at sporadic intervals. Without understanding product profitability, plant efficiency, and the key cost drivers, the leadership team drove by intuition. Each department had its own unique (see tedious) system to get the data and insights needed to do their jobs. This resulted in there being multiple “numbers of truth” for any single question.

The business was out of control.

My job, the reason that I was hired, was to bring in my experience of fixing and professionalizing accounting functions to the role. Clearly, much had to change.

Excel Only Gets You So Far

July 2019- January 2020

The first step, of course, was to understand what was currently being done and why. I spent a great deal of time understanding and documenting each process and task to grasp what was being done, how, and why. In this investigative stage, I discovered the whole gamut of common Excel user errors; incomplete data, incomplete summaries, incorrect formulas, un-updated inputs, and processes that involved writing over data. In the end, nothing surprised me.

Sidebar: Here is where many professionals go astray and continue to do what has been done before. However, if you understand the potential of the tool you use, you can drive real improvements. To this day, it still baffles me as to how Excel is so quickly bastardized. Instead, take the time to understand best practices and to master all the essential functions. As a golden rule, if you can’t explain your spreadsheet and your process to a 10-year-old, then something is seriously wrong.

Once I understood the current processes, I seized the opportunity to improve them. Every Excel process would be simplified, automated, discarded, or upgraded as I saw fit. The critical decider was whether something added-value or not to the business. As we needed new processes to serve business needs, they were built, validated, and shared.

However, Excel improvements can only go so far due to inherent limitations in the software. It’s great for a few things but not as a replacement for everything. It can’t be a database, calculation engine, and presenting tool at the same time. For example, once you hit a specific file size around 30–40 GB or add several complex formulas, the file begins to bog down and suffer performance issues. Calculating 41%…42%…43%…

The limitations of Excel also result in the business analysis segmentation into separate files saved in separate folders. This results in the age-old problem of not being able to see the forest through the trees. Each process was still labor-intensive and thus only produced weekly. Still, this was better than when they were previously done once a month or not at all.

The benefit of Excel, however, is that you can create a robust process framework that can be leveraged later on by answering the questions of:

1. Where does the data come from?

2. How is the data transformed?

3. What additional reference tables supplement the data?

4. What are the output and data summary?

From this point, we completed our review of all critical processes and began to understand what the future looked like. Instead of performing a tedious analysis once a week, we wanted to automate all the steps to be performed once a day. Instead of having disparate files in a folder somewhere, we could create an analysis tool that would consolidate all analysis and periods over time. Instead of folders for each plant, there could be one centralized tool that served as a one-stop-shop.

A key point at this stage was to understand the source of the data. We had an enterprise tool that would produce flat reports from a data warehouse connected to the ERP. People would export reports and then manipulate them further in Excel, but this lacked the depth required to take controlling to the next level. What we needed was the data, not the summaries performed in a black box. Getting behind the scenes and accessing the data to query it when and how we needed it would be the end goal. Instead of using the tool to produce static reports, we would need to repurpose the tool to produce data that the business could use as needed.

To add more functionality and depth to what finance could offer, this would require more human resources and resources that were currently non-existent. We had stretched everything we had at our disposal to the edge. Getting to the future and addressing the business’s critical and still unanswered questions would require a very different approach. Instead of working harder, we needed to work smarter.

Getting to the data

Of all the steps involved with creating an analytics organization, accessing the data has been one of the most painful and laborious processes. As is typical, the keepers of the data are IT. Unless IT is willing to adopt a customer service mentality to understand and respond to the business’s needs, then everything in-between becomes much more difficult. And boy, was it difficult.

In the organization, the IT team was staffed by a group of people who provided limited services and support to the business and were hesitant to step outside of a limited box. There were trust issues galore. I cannot begin to tell you how many meetings and discussions that often turned heated were involved in bridging this gap. We would have conversations intended to explore the options available to access the data whether, through SQL, an ODBC connection, or something else, and the answer was no. “We will not share access to the data as our reporting tool has worked just fine for all the previous controllers” was spoken many times in different ways.

To solve the problem, the first approach was to work with IT to explain that the status quo wasn’t cutting it. Instead of being ok, the business had huge variances that they couldn’t explain and a costing system out of control. We explained our pain points, our goals, and our shared desire to help the organization. Still, the answer was no.

The second tier of this approach was to partner with another senior manager in the organization who ran the supply chain. He had long grown exasperated with the IT team had given up years ago when his data access requests fell on flat ears. Tensions ran so hotly that he refused to speak with IT. What had been a typical request for data that was quickly granted in his past experiences became a brick wall here. Instead of access, we were constrained with draconian controls.

Despite these approaches, IT remained steadfast in refusing to partner with the business in their quest for data access. They referenced previous failed business initiatives to access the data, security “risks” involving access, and they also suffered from a lack of talent and know-how in their department. Even if they wanted to change how things were done, they didn’t know where to start and wouldn’t admit it.

We were stuck.

Prague

May 2019

Throughout improving the current controlling outputs and KPIs, I worked with my manager, the CFO, and frankly, anyone who would listen to explain the goal of accessing the data and how it was a critical linchpin in creating a modern accounting function.

At a standstill with IT and stuck from progressing, we worked with other executives in the global organization to arrange a sister company visit that did allow their people to access their data, automate reporting, and deliver excellent service. I was to understand best-in-class practices and to bring the knowledge back to the US.

Sidebar: Days before I was set to fly, I had to undergo minor surgery that resulted in the excruciating tooth removal. With a bag full of antibiotics and pain medicine, I decided to press forward and not risking losing or deferring this opportunity. If I had further health complications, I was in the unknown, but it was worth the risk.

So, bags packed, I jetted off to the Czech Republic.

Landing in Prague, I met my contact and explored the country office to understand how they use data access to run the business. We soon drove to a rural cheese factory in the countryside, and I observed how the controlling function was run.

Sidebar: To many Czech’s at this factory, I was the only American they had ever met. I have a regional accent and a few quirks that they probably now think are typical. I apologize.

I asked many questions in the Czech Republic and probed deeply to understand the relationship between users and data. Ironically, this Eastern European country was light years ahead in their sophistication than their American cousins; just about all-controlling processes involved automation to the point where they would just refresh a pivot table to get the business’s answers. They were not exporting, cleaning and manipulating, and then summarizing Excel data. Their IT team had built a data warehouse connected to SAP, just like we had.

It was now clear.

Both companies used the same software, and only one was using it appropriately. I had the support needed to go back to the US and drive the change needed. IT would also hopefully see the light, bring down the walls, and work with the business users.

The Doldrums

May- September 2019

Full of energy, refreshed, and enthusiastic again, I returned to the US. I understood what needed to happen and the tools required. Instead of using Excel manually and using our data tool to produce reports, we would need data access. No longer should the data restrictions apply if the same restrictions were not in place with another Savencia organization. We would build processes that would perform the ETL (Extract, transform, & load) functionality with a refresh.

I met with the CFO and discussed the path forward, and then I met with IT. While the Czech team had built a separate database using OLAP cubes, we had an SAP BW data warehouse. Bridging the gap to allow business users to query and use the data was outside their realm of comfort and technical expertise, and they would not budge once again.

I got yet another no.

During this time, the business urgency to provide insights and answers reached a fever pitch. The executives and the CFO were growing frustrated with the pace of progress. Whenever something went unanswered or answered incorrectly, I could sense that the window of opportunity was closing. They did not understand data access’s criticality to reach the role objectives and invested their political capital elsewhere. Pushing for resources, they finally acquiesced and would grant me an intern beginning in September to run through the end of January.

The second bright spot here was that IT finally looked to upgrade its capabilities. The employee who had managed the data warehouse previously left, and an experienced and knowledgeable employee took his place. We would finally have the technical knowledge on the backend.

Although we were using all our available resources to the maximum capacity and facing mental burnout, we were still not able to give the business what they wanted. We needed a different approach.

IT, growing tired of my requests and number of service tickets, manufactured a plot to push me out of the organization when I called them out for sharing incorrect solutions to a problem that they didn’t understand in ERP. Rather than waiting to test a solution that didn’t work, I tested it as she spoke and asked her to explain the results. She grew furious, stormed out, and went home. The next day, we have a cross-functional meeting, and she accused me of malicious harassment. After tempers flared and rose, the facts came out that my point was valid. She quit for a few months, returned, and then we repaired the relationship. Calling out consistently low-quality outputs raised the standards for everything they worked on with us afterward.

In June, I would receive my first verbal warning from the CFO because “business stakeholders were dissatisfied.” I knew that this was the first formal step in removing me if I couldn’t deliver.

Would there soon be a sixth controller within five years?

A Path forward

September 2019 — January 2020

With the arrival of my new intern, Jean-Baptiste, I sensed that a new day had arrived. His arrival signaled new opportunities and new challenges. Coming from the end of his university training in France, his English would need some work, he would need time to adapt, and I would need to learn how to manage a growing team.

Fortunately, Jean-Baptiste proved highly intelligent and eager to learn. As I transitioned various tasks to him that consumed my daily bandwidth, I could now use this time to plot our path forward.

From my network and research, I had determined that Power BI would be a handy tool for us to use to achieve the ends I had in mind. In the back end, Power Query could perform the ETL of a data source file, while measures in Power BI would allow me to perform the calculations and summaries required.

Although we were still not granted access to the data, we created a workaround. I would use their reporting tool, but I would adopt it. I removed all unnecessary fields and subtotals from these data reports to produce clean data tables. I created SharePoint folders for the different business processes and saved individual months of data into each folder.

Next, I would have to learn Power BI. I researched and paid for a course and spend countless hours on nights and weekends going through the lectures and practice activities to gain a base of knowledge.

During my days, I would apply what I knew and then continue to research and learn what I didn’t. Google was a great resource that helped answer all sorts of problems that we faced where formal education had left gaps.

In iterations, I would use Power BI to replicate the Excel processes and validate against several months of past results. It quickly became apparent that Power BI was more potent than I could have imagined. Whereas previously, we would perform the same process individually for three manufacturing plants with separate files. Power BI could perform those same calculations, and filters would allow users to select which plant summary to view.

I worked closely with my intern as we both began to develop more profound knowledge in Power BI design and coding. Here, we mainly taught ourselves DAX and dabbled lightly in M.

By the end of September/October, we had begun to share our Power BI reports with the CFO, manufacturing teams, and other function heads. They also understood the value of Power BI, and Excel was phased out.

I continued to press the CFO for Power BI Pro licenses to take us to the next level in reporting. I asked for ten licenses to start to create my dream one-stop-shop in Microsoft Teams for all analytical needs. Here, different tabs in Teams linked to published Power BI reports would represent further analysis. To name a few, there were separate tabs for labor efficiency, machine efficiency, waste analysis, product profitability, production volumes, material yield, inventory on hand, and consumption.

Teams allowed us the ability to create a self-service analytics environment. Whereas previously, we would have to email out the Excel file to various teams to discuss performance, now we could guide these same people to the tool and show them how to use it. We built trust with them and worked proactively to design it to their specific needs. It was a hit! Even old-school employees who were used to traditional ways of working found it highly beneficial and user-friendly.

As each process was automated in Power BI, we gained valuable time back. We could and did reinvest that time to automate more reports and create the critical analysis that would answer the business’s unmet needs. It began as a trickle and eventually turned into hours a day of savings. Even though we were manually updating our data files in our folders each data, this would not be a bottleneck. We would build business acceptance first, and then with momentum and overwhelming support on our side, we would work with IT again.

Over time, we would continue to add more user licenses until we reached the mid-30’s as we built critical mass.

Sprinting to Success

January 2020- July 2020

Power BI had become the backbone of operational controlling for Savencia by January 2020. To say the least, I could not perform my job without it. This experience demonstrated to me that BI tools are critical for finance who are interested in taking their performance and their deliverables to the next level over the Excel plateau. It wasn’t merely a nice-to-have; it was a must-have.

Working with IT now and their new and knowledgeable resource, we began to automate all reporting fully. What was once done weekly or monthly in Excel would be done overnight with zero human interference. We treated IT as a partner and worked to build a strong working relationship. Instead of using them in a task-based approach, we brought them into the business and showed them the impact of the work we were doing together. I was not just asking for a data connection; I showed him how things worked and gave him access to everything I could. Once we figured out how to fully automate one report, the others came quickly. As bugs popped up, we quashed them together.

We developed new reporting and analysis that satisfied the business’s need and solved new problems that we were not aware of before. By viewing the business and the results holistically, we had time to think and question what didn’t look right. For example, we discovered that a reconciliation between the ERP and the timeclock was not being done. We fixed it. We discovered inconsistencies in processing data from different people for the same purpose. We worked with them and fixed it. And with Power BI, we had a way of monitoring and controlling to ensure that what was fixed stayed fixed.

We were able to spend time working to understand and document the business processes that generated data and collaboratively design new and improved processes that would generate much higher quality data. We spent time understanding the various systems that were not connected to the ERP and pushed IT to integrate them and raise support.

Best of all, for once, we would answer the business’s critical questions. The customer that they were unsure of whether or not it was profitable was resolved. Product profitability was fully understood, which led to decisions to limit the number of unprofitable sku’s. We answered what the break-even volume was for each plant. Management could now drive forward without a blindfold in their decision-making process.

The time and resources that we freed up allowed us to improve the budgeting process, as well. We created a truly zero-based budget that the business had a high degree of confidence in for the first time.

With any remaining time we had available, we began to partner with the different support functions such as supply chain, purchasing, and sales to improve their reporting and analysis withs with Power BI. Where data structures were missing, we worked with IT to fill these needs. For the first time ever, the business would have one “number of truth” for any question. For most questions, with Power BI, we could answer them within minutes instead of deploying the far too typical, “I’ll get back to you in a few days.”

Word moved up the chain, and global executives noticed what we were able to achieve as a high-performing finance team.

The Final Barrier

July- October 2020

At the end of June, my intern and now friend Jean-Baptiste headed back to France. Over the few months that we worked together, he grew tremendously. His language skills were highly developed, he was technically savvy, and he learned how much finance function operates. Although he was slated to leave by the end of January, I convinced him to stay for the final push, and his involvement proved critical.

Unfortunately, the loss of 1/3 of my team resulted in more of my time being spent in the weeds and details instead of providing leadership and working with other functions. There were still systems and processes that we needed to fix, but this required political capital that the executive team was unwilling to invest. Now filling in the work left behind and working on the new zero-based budget, I faced burnout for a second time.

To address the issue, in September, I approached the CEO and proposed that my role be expanded from Operational Controller (Manager) to Director of Process Improvement & Business Intelligence. Some of the challenging problems ahead would require at least 1–2 more resources to resolve and formal authority for owning and directing business processes and systems. While the CEO congratulated me on my success and asked me to stay, he was unwilling to change my title but did offer a substantial salary increase. I turned him down. To me, money is only part of the equation. If I cannot fix the root causes of problems and instead apply band-aids, the work drains me mentally.

Conclusion

I turned down the pay raise and chose to enter the job market again. With my new skills and accomplishments, I was highly sought after for director-level roles and above. I quickly found a new role with new growth opportunities that would pay me 50% more than I earned previously.

I am incredibly proud of the legacy that my team left behind. For the next controller, all processes for month-end and routine activities were fully documented and tested by other finance employees to hold in the interim. All the Power BI dashboards continue to work to this day.

The experience at Savencia, although very challenging and draining at times, was critical in my growth as a finance professional and in understanding the impact of data and data analysis in the future. I kept my commitment and delivered sustainable improvement.

I have several takeaways from this experience:

· If you are a finance professional, you must learn and know how to use BI tools. Whether you use Power BI, Power Pivot, or something else, learn how to query and transform data and then summarize the results around the business’s needs. There is no value in exporting and manually manipulating data in Excel.

· Grit and perseverance are critical for getting to a solution. If you face no, find a workaround.

· Access to data will be contested. Learn what to ask for and understand how others access their data.

· Without good processes, you get poor outputs.

· Resources are very limited or non-existent. Learn to work without them by removing low-value work.

· Organizational leadership will not understand “how the sausage gets made” or required to get the desired outcomes. Getting support from the top is difficult.

· The skills gap is genuine and is a significant hurdle at multiple levels.

Originally published at https://www.linkedin.com.