This isn’t going to be a typical comparison between two different solutions. I am not objective here. In a different lifetime, I might have become a great advocate for the other solution. After all, it has great visualizations, and a fantastic community of professionals and fans. You might also argue (especially if you are a fan of the other product), that I have no real claim to share my thoughts here. After all, I have never used the other product.
Nevertheless, if you are a fan of the other solution (and by now may be a bit angry that this blog post was shared with you, by someone from the other side), I think you should read this blog post, and the series which will follow. I am not going to convince you to stop using the other product. But it’s definitely time for you to start using this new emerging product. In the next 10 years, it will conquer the world. It has already started. And you cannot afford to stay behind.
In this series, we will discuss why Power BI is going to change your organization’s data culture, and provide you a rare opportunity to join the evolution.
As a community member, I have started to feel the momentum in the second half of 2016. The demand of Power BI was steadily increasing, and since 2017, it is accelerating to exhilarating proportions. Last week in Microsoft Data Insights Summit, James Phillips announced Power BI Premium, and shared a series of game changer features which will be released in the next three months.
Especially interesting for Tableau users who start considering Power BI projects, in the next 3 months, Power BI will release lots of storytelling experiences, that were typically associated to Tableau’s unique value proposition. The Power BI new features include timeline custom visual, quick insights, drill down and Visio integration. And there are also the new Data Bars that were just released in Power BI Desktop June update and the global availability of ArcGIS Maps.
Update (Sep 8, 2017): Power BI released an improved sampling here, which better addresses the limitation. We can now officially say that the claim above is obsolete.
So, here are few snippets of the things we will discuss in this series:
- Demystifying Power BI misconceptions.
- Share resources & best practices for BI teams that start considering using Power BI, or are starting their Self-Service BI & Data Governance journey.
- Demonstrate the power of Power BI to change your organization data culture (Sneak Peek can be found in my session at the summit here).
Ready to start?
Demystifying Power BI misconceptions – Part 1
According to a Tableau marketing page here (Which you can find as the first Google search result of “Power BI vs. Tableau”), there are 8 ways Power BI falls short (One personal note for Tableau product marketing team – Can’t you be less negative here? Don’t you think some people will think you are too drastic, and perhaps even in panic?).
According to Tableau, the first way Power BI falls short is the missing of outliers in scatter charts: “For starters, Power BI limits data visualizations to only 3,500 data points; it automatically filters your data, at random, for any points exceeding the limit.”
So, I had to test this claim (I am sure it was a correct claim in the past, but with the massive monthly updates by the Power BI team, it’s difficult to track the improvements).
Update (Ssp 8, 2017): This section was written before September update, which addresses the limitation above, by sampling high density scatter charts. Read more here.
I created an Excel workbook with more than 10000 rows, and a single outlier, whose location in the dataset can be randomly selected. When I loaded the data to Power BI, the outlier was always detected (Check out the cyan colored dot in the top right corner), so even if the visual is limited to 3500 points, the outlier is shown.
Now, even if Power BI is limited to 3500 values, you cannot rely on Scatter Chart to detect outliers. You better use box & whisker custom visuals or R Custom Visuals in Power BI, and with few mouse clicks, and zero knowledge of the programming language R you can detect clustes with outliers.
In the next post in this series we will continue the journey in demystifying the other seven ways Power BI falls short according to Tableau.