Struggling with Your Analytics Reputation?

As someone who helps clients with analytical subjects, I see the full spectrum of people’s successes and frustrations. Often, people will confidentially share their struggles with their external consultant when they wish to have a sounding board but they don’t feel comfortable having the discussion with their colleagues or leadership.

It is also not unusual after I give a speech at a conference, to have analytical professionals walk up to me and say, “Thank you, you’ve helped validate what I’ve been trying to accomplish at my own company.” It is comforting to these people to know that they are on the right track, even if they haven’t reached the point where their own colleagues understand the bigger picture and see the value of analytics. At this point, they are feeling alone on an island with no supporters, especially if they are the only analyst within their functional area.

Some people translate their frustrations into a need to organize and define. Over the years, I’ve seen many articles of people attempting to define analytical terms and / or analytical job titles. One author suggested that you can’t be a Data Scientist unless you are versed in applying scientific techniques like hypothesis testing. I guarantee you that the people in HR writing job descriptions and using the title of Data Scientist will never worry about the suggested definitions on the internet. In the real world, HR will do the best it can when writing technical job descriptions and we will always see a variety of titles in analytical roles. It’s far more important for them to capture the needs of the role than to worry about the specific title. But, I digress…

For those who are frustrated, I would like to provide some assistance. I have seen approaches to analytics which begin with floating a business case to the executives in order to get “buy in.” To be honest, anyone can make almost anything look good in a PowerPoint presentation. And if you’re in a large company like I was a few years ago, it would have taken me a full year to cycle through all of those executives. It’s just not a practical approach and those who know me, know that I’m all about aiming for practical and realistic when it comes to analytics.

I’m more of a “prove it” kind of person, so business cases based on general information you obtained from the internet hold no value. Instead, I prefer the bottom-up approach of proving what can be done with analytics to show value. I would always seek out someone who was “feeling major pain” and would be open to accepting my offer of, “Do you mind if we give analytics a try in this application?” Those “feeling major pain” are more apt to try new approaches.

As a second piece of advice, allow me to use HR for this example. You may have noticed that most large companies are structured such that HR Business Partners are aligned to business areas. They represent their business area to the HR team. As such, these are the people that need to be advocates of analytics in HR. However, until you educate them in the capabilities of analytics, they won’t be able to understand the value they can bring by suggesting HR analytics to their business area. By educated HRBPs, I’m not suggesting a mathematical education on analytics, but more of a non-mathematical introduction to show them what is possible. For anyone out there who has heard my talks at HR conferences, it is always non-mathematical. HRBPs don’t need to DO the mathematics. They just need to know WHAT you can DO with analytics. With this approach, HRBPs can become your advocates and you will struggle far less with your analytical reputation.

If you are not in HR, you can still think of your struggle within the same context. Who are the people feeling pain where you think analytics can help provide insight into one of their issues?

I’ll leave it to the reader to contemplate their own situation and approach.

Until next time,


Tracey Smith is an internationally recognized business author, speaker and analytics consultant. She is one of the most highly respected voices when it comes to business analytics and HR analytics. She is the author of multiple business books and hundreds of articles in a variety of publications. Tracey has worked with and advised organizations, both well-known and little-known, on how to use data analytics to impact the bottom line. If you would like to talk to Tracey about consulting work or speaking engagements, please visit  or contact Tracey Smith through LinkedIn.

Visit Tracey Smith’s Amazon Author Page


User Experience Test with Power BI

As an entrepreneur who makes a living in analytics, it is crucial for my company to keep up with the best tools on the market. One of the most important tool categories for my work is data visualization. My tool of choice for the last six years has been Tableau.

That said, for the past 18 months, I have been occasionally testing Microsoft Power BI. You may have seen them appear in the upper right quadrant of the last Gartner report. Backed by the budget size that Microsoft can offer, and pricing themselves below Tableau, I can see it being a matter of time before Power BI competes head-to-head with Tableau.

Gartner Report – February 2017

With a change in position on the Gartner report, I felt it was time to yet again head into Power BI to test the user experience and capabilities. Keep in mind that the axes on the Gartner report are “completeness of vision” and “ability to execute.” It doesn’t measure the existence of certain capabilities and the customer experience. My benchmark for testing is my experience with Tableau Desktop Pro, Tableau Online and Tableau Server.

My previous tests of Power BI have always left me feeling very restricted in capabilities. In these cases, I was testing Power BI “Services” which is the online service tool. As an additional note, they also offer a desktop version.

Microsoft Power BI Services (Online Tool)

Heading back into the online service version a few days ago, I had renewed hope of capabilities based on the Gartner report. The first thing I noticed in comparison to Tableau is that the service connections with Power BI offer more on the social side. For example, Power BI says it can connect to a MailChimp account (email campaigns) to create a dashboard for email campaign statistics. I thought I’d start there.

Selecting the connection for MailChimp in the Power BI service was easy. I selected my simplest MailChimp account with one list in it. The connection was a success and to Power BI’s credit, it created a default dashboard and report. In Tableau, I would need to build my first visualizations from scratch.

I continued to Test Number 2 which was to switch the credentials (ID / password) of MailChimp to my Numerical Insights account. That account houses several lists, is much larger and has more activity (subscribes, unsubscribes, email bounces etc.). I switched the credentials and… Power BI still showed me the MailChimp data, report and dashboard from Test Number 1. I tried again and… a connection error.

I submitted this error to the Microsoft Power BI community and the response that came back was that you have to delete everything you have (MailChimp data, report and dashboard) and start over again. Did that work? Yes. Was it a good user experience? No. Do I consider that a “solution?” Definitely not.

Microsoft Power BI Desktop

The second response from the community was to use the Power BI desktop version. The community claimed that switching the credentials of the MailChimp account (ID / password) was possible on the desktop tool and that the desktop tool had greater capabilities. So, I headed in that direction.

About a year ago, I installed Power BI desktop on a Windows 7 laptop without any issues. I could easily connect to data and create basic dashboards. I could share dashboards with other people, as long as they didn’t have an email address which was designated as personal. The frustration here was that I couldn’t share a dashboard example with a small client who used as her official business email. I did have to laugh a bit since locking out is Microsoft locking out one of its own domains.

Today’s test was on a new Windows 10 laptop. I downloaded the desktop version and the installation produced no errors. However, launching the program was an entirely different story. A box opens, saying “Initializing model” and then I am left with a black box with some indecipherable text in the upper-right corner. At this point, the program is frozen and must be closed within the task manager.

The response from the Power BI community? It mentioned changing settings of GPU rendering within Internet Explorer (for some reason the program pulls its settings from this obsolete browser), stopping Nahimic for MSI if you’re running an MSI laptop (say what?), and installing the 32 bit version instead of the 64 bit.

I tried the first and third suggestions and managed to get a box to appear which allowed me to sign into my Power BI account. Sadly, it returned me to the same black screen. Clicking in the “blackness” at other locations revealed another menu. Sadly, I have to guess on this black screen where the menus might be. Clearly I have some sort of display issue.

I decided to attack this problem with the process of elimination. I am running two screens: one on my laptop and another with an external display. Each has a different screen resolution. Sure enough, if I disconnected the external display and then launched Power BI, it displayed correctly. So, the key to making Power BI work for me, is to unplug my monitor, launch Power BI and then plug my monitor back in. I suppose that with two different screen resolutions, some of my pain is self-inflicted, but surely I’m not the only customer with this set-up.

At this point, I headed back to my MailChimp test to run the same test that I had tried online. In the desktop version, a MailChimp connection is clearly labeled (beta). If I’d seen that label in the online version, I would have been more forgiving in my assessment of that tool.

I managed to connect to my simplest MailChimp tool. In the online version, I was spoiled by being handed a dashboard of my MailChimp information already made. In the desktop version, this doesn’t happen. No problem there since building visualizations from scratch is what I’m used to.

I headed into the test of changing the MailChimp credentials, i.e., changing the ID and password over to my more complicated MailChimp account. If you recall, the reason I decided to install the desktop version was to get around a lack of functionality in the online version.

I found where to change these credentials pretty easily and asked Power BI Desktop to refresh my data view. Unlike the online version, the desktop version was successful in this task. I could see a 5 year history of all of my email campaigns.

At this point, I had used up my allotted testing time. I have seen enough that I know that the online version is not something I would recommend to my clients. As for the desktop version, I do plan to continue testing and will track the results against Tableau.

I am convinced that Power BI will morph into something impressive, but I’ll need to give it more time. I promise to conduct more testing in the future now that I have resolved my desktop installation display issue.

Until next time,


Tracey Smith is an internationally recognized business author, speaker and analytics consultant. She is one of the most highly respected voices when it comes to business analytics and HR analytics. She is the author of multiple business books and hundreds of articles in a variety of publications. Tracey has worked with and advised organizations, both well-known and little-known, on how to use data analytics to impact the bottom line. If you would like to talk to Tracey about consulting work or speaking engagements, please visit or contact Tracey Smith through LinkedIn.

Visit Tracey Smith’s Amazon Author Page

Critical Business Decisions, Quickly

By Tracey Smith
President, Numerical Insights LLC

We live in a crazy world where time is money and the faster you can make key business decisions, the more likely your business will still be here 10 years from now. Perhaps you’ve read this statement by Richard Foster of Yale University.

“The average lifespan of a company listed in the S&P 500 has decreased from 57 years in the 1920s to 15 years today.”

The radical change in lifespan on the S&P 500 has a lot to do with today’s increased pace of change and a company’s ability to react and / or be proactive in its decision-making. You may have also noticed a flood of companies actively assessing how to analyze their data to make good use of it, and in some cases, monetize its value.

Now, the story goes well beyond large companies and heads into smaller companies. In fact, smaller companies often have the ability to move faster than an S&P entity just by way of having to head through far fewer “approvers” before a decision can be executed.

Businesses of all types and sizes are now finding ways to deliver better business outcomes like increased revenue, lower costs, better quality and increased profit. Today, the need to gain insight into decisions affecting these outcomes is crucial for everyone. 

With today’s analytical tools, using business analytics to “see into” one’s business, is no longer out of reach from a budget standpoint. The battle in the business intelligence world is so fierce today that prices have become reasonable and tools are more accessible.

As one example, we’ve put an incredibly simple dashboard online for everyone to see. Click here to view it. We’ve greatly reduced the number of parts in this example since the real world application has over 10,000 parts to look at!

Even an example as simple as this one allows a company to make key decisions. If you didn’t click the link, it’s a look into a company’s product profitability where we can easily visualize which products contribute to the company’s bottom line, and which may need to be considered for deletion. Additionally, we can see that products that bring in the most revenue, don’t always bring in the most profit. Further, the company’s marketing department can use this information to target their marketing budget in the direction of increasing volumes sold on higher profit part numbers. This is just one of hundreds of examples we can present for valuable decision making.

The benefits of analytics are well documented:

  • Greater visibility and capability to analyze data and make important decisions with it,
  • Ability to measure detailed performance or products, services, warranty, customer service… almost anything you can imagine,
  • Company-wide access for employees and leaders to use information interactively and see data the way they need to see it for their own job, and
  • Ability to increase business results.

Insights lead to better decisions, which leads to better business outcomes.

Does this sound like something you’d like to hear more about? Use the registration link below and we’ll send you a free case study paper describing the business outcomes that can be obtained by looking at this information.

Register here to receive a Case Study paper. Available to new registrants.

Until next time,


Numerical Insights is passionate about using data to solve business problems. We have helped both well-known and little-known companies in multiple countries use data to make decisions which impact the bottom line.

You can find Tracey  Smith on the web at:
Numerical Insights Web Site
Find Tracey on LinkedIn
Twitter ID: @ninsights