Guest Story: Power BI Adoption

Finding value in data and dialogue

After reading Gil’s article asking if we were “Riding the wave of Power BI”, I wondered what we’d share about our own journey. Now down the road from our humble beginnings in Power Pivot, we’ve put pen to paper to share our experience.

Six years ago, I had a hard time believing that Excel would ever be able to connect and import 150,000 rows of data per second. My experience using Power Pivot would just be the beginning of what we’d later realize with Power BI. Over time we’d realize a subtle difference in approach that would help us realize big gains. We’d learn how to leverage data and visuals to drive better dialogue than just expect data and visuals to speak for themselves.

2013 – Excel on steroids

In 2012, I was fortunate to take a role in the Accounting Department at Industrial Finishes and Systems. We are a nationwide growing family-owned company that primarily serves the collision repair, wood products, recreation vehicle, and aerospace industries. A large part of our business services automotive body shops with paint finishes and supplies. We proudly claim, “If it rolls, floats or flies, we have the solution for you!”

I came into the company as an Accounting Manager with a finance and accounting background. In past roles, I had leaned heavily on Excel and automation. Having done my fair share of VLOOKUPS and pivots using legacy system data, I quickly felt at home and sought out the “Export to Excel” button. I diligently worked to automate our existing manual processes and provide trends and insights.

In 2013, I set out to find the next step in my Excel journey. I was midstream in a VBA course on YouTube when I stumbled on a mysterious Microsoft product called Power Pivot. I purchased the Power Pivot DAX book by Rob Collie. In the book, Rob writes that Power Pivot will be a revolutionary discovery for the Excel Pro., and by chapter three I was hooked. I became fast friends with our ERP vendor to help accumulate all the tables we would need to connect.

2015 – Power BI a whole new approach

Fast forward to the public release of Power BI in 2015. I’m thankful to have the tables in place and a solid start with DAX. Our student intern Hunter quickly picked up Power Pivot, and we’d eventually download Power BI to try out. We realized that this program gave us a clear advantage with regard to publishing and refreshing our analytics on a much larger scale. We are now able to bring valuable data closer to where our business actually happens, with less effort.

With Power BI, the accountant takes on more of a developer role, and the field sales manager takes ownership of what information they access and where and when they want to access it.

Where has this been my whole Excel life? One model; DAX is fun; automatic updates; and many people benefit…SOLD!

Cleared for takeoff

Next, we sought approval to publish reports and to encourage others to sign-up. This required internal discussions regarding cost-benefit analysis, data and system security, and our overall comfort with cloud-based data storage. We understandably encountered hesitation and push-back at times. All our information in the cloud? How about data security? Can we control who sees what?

After reviewing the Power BI Security Whitepaper, our senior management concluded that given our size, geography, and investment in Office 365, Power BI made for a great opportunity to build a competitive advantage and better see and understand our business at every level.

Given our decision and passion to implement Power BI, on top of an extended free pro trial offer, management fully supported us in seeking out additional learning opportunities. We jumped in and made an effort to deeply understand DAX and best modeling practices. Feeling somewhat confident with DAX, we next moved into Power Query and learn just enough M to be dangerous.

Knowing what we know today about the abilities of Power Query and M, we would have invested in Query earlier than we did. Having the ability to effectively clean, sort, and shape data ahead of time really took a load off of our DAX functions. Today with M and DAX we are more confident in our ability to build complex models.

Broadening access

Our rollout was slow and deliberate. To some degree, we fell into Stage 1 in Gil’s article, “Hiding behind the wave barrier”.

Here, rolling out access to data was something that IT (and management) provided to only the select few. When a third DAX developer emerged in the organization to blend operating data to create KPIs, he would need our sales data. This is still a careful decision, but the decision today considers cost and benefit, not strictly where someone falls on the org chart.

Testing the waters

I’d love to say that from here that we published perfect dashboards, everyone signed up, and we all lived happily ever after. The reality was that this effort required a large amount of evangelizing, training, and walking softly around existing paper-based reports and processes.

We set out to publish dashboards in four areas working ourselves from the inside of the organization out:

1. Accounting, Finance, Payroll – Data mining and reconciliations
2. Branch Operations – Trends, daily status reports, inventory adjustments
3. Field Sales – Sales mining, buying habits, and trends
4. External Customers – Impact / Scoreboards

Convincing report users (data explorers) inside the company to put longstanding reporting aside and pickup Power BI was challenging. We sat with people, held screen-share calls, and volunteered during live training events to share. We even came up with an acronym – T.R.U.E. – and promised that if a dashboard didn’t provide you with Timely, Relevant, Useful, and Easy to comprehend information, we’d get rid of that dashboard with no obligation.

During this time, we published some simple internal data mining dashboards. We worked to find layouts that were simple and functional. Most in our organization prefer to see the raw numbers, so we combined slicers on the top, a table/matrix of numbers in the middle, and finally a comparative year-over-year line graph on the bottom. This standard format was a great fit, and our data mining dashboards started gaining traction.

We started dealing with some that did not yet 100% trust the numbers. Numbers move and change quickly in Power BI, and no one can see all the data at once. I fully appreciate wanting to stop and validate. It was important to jump on each concern early and help validate or fix minor issues. We made it a point to embrace every question from users as an opportunity to further gain trust.

Hold on!

Our functional format and our drive to demonstrate data integrity allowed us to leap over Stage 2 in Gil’s article, of throwing pebbles in the pond and move straight to Stage 3 – Riding the tidal wave.

As we talked more about our dashboards and addressed the accuracy of the numbers, barriers fell, and pro account requests became more frequent. Talk of Power BI moves beyond a few power users promoting the program, to a larger majority talking about the benefits of the information and referring to it simply as “My Power BI” …. “I looked on my Power BI and it showed me X.

Exploring financials in the matrix

Our next conquest was to replicate our financial statement and give managers the ability to navigate through the financials and be able to drill in to view the underlying detail. We currently have measures that read through each line in the financial statement and call out areas of interest for the manager by creating a measure that scores each line and links that score to a conditional color.

About a year later, we are approached by an existing report vendor with an offer to add BI visuals including a financial statement package. This challenged our existing statement in Power BI. Not to say that we need Power BI to provide every BI report, but the decision to stick with Power BI, in this case, showed our commitment to a tool that we are able to develop as our business needs change.

Road trip!

Encouraged by the wins inside the organization, we decide to take Power BI and some data on the road to a few external body shop customers. We designed a simple purchase tracking dashboard and visit a few body shops to see if they saw value in accessing their spending. The response was okay, there was no real spark. We asked if they’d be willing to share information with us regarding monthly production numbers so we could generate some industry insights for them. Some agreed and we went back to the drawing board.

We came back with version two, we combined some of their key production numbers along with their purchase information we held on our end and generated different body shop Key Performance Indicators (KPIs).

For the first time, we’re able to walk them through the visuals and show them how the quality of their insurance-paid estimates trend, or if the problem is a result of overuse or waste of product.

We showed them how those numbers could improve overtime on the dashboard, and this is where we noticed the shop owner leaning forward and just say, “Wow” or “I had no idea” … We’d stop for a minute and just listen to the owner brainstorm what may be at the root of the results. Before we knew it, we were participating in a data-driven conversation and providing next month’s goals and solutions.

More than just a pretty interface

Let me be clear here, as much as I love Power BI, the key takeaway was not that Power BI was a magic number cruncher (although yes that is impressive), but rather that Power BI could help pave the way to lowering barriers and drive real conversation.

If the tool drove this type of dialogue – and this is where the value really is – we couldn’t just drop off the username and password and expect success from day one. It would be like me bringing a new bike to my toddler and showing him how great it is to ride. Would he be excited? Absolutely! But if I give him a high-five and walk away, he will likely go back to playing with his other toys. The bike is awesome, but only when I’m there to show him how to use it.

In other words, the stakeholder that sees the right demo of data AND dialogue feels hopeful that they are partnering with a solution provider that speaks their language. In hindsight, when given limited time and attention from a business owner, the spreadsheet at a glance, looks like a number puzzle, and Power BI clearly makes a better connection.

Stakeholder first

This is where we broach into parts of Stage 4 – Reaching the other side of the whirlpool.

We define our success in this area by our ability to set and track goals, as well as the quality of our dialogue when using the visuals. Starting with the stakeholder’s needs in mind, we work backward to the design of the model.

It was brilliant for Microsoft to generate a suite of ‘self-service’ programs that shift meaningful information closer to the stakeholder. The strong Excel user has an opportunity to become, as Donald Farmer spoke about in his 2019 Pass Summit session, the “New Analyst.” This person has evolved from data provider to someone that is creative in design, gets access to source data, and takes ownership of the finished product. There is a great opportunity for the analyst to step up to provide lasting value on a larger scale.

Now that we are this far in our journey, we look forward to new solutions that drive overall corporate improvement among our different regions and workgroups. There is little doubt that Power BI will be front and center in our company for years to come.

To feature your Power BI adoption story, contact Gil at gilra@datachant.com. Did you find Mark’s story useful? Did you have a similar journey? Share your thoughts in the comments below.