Process Mining in Power BI
Process Mining 101 series: Part 4
In this part of our series, you’ll learn why we chose to build our Process Mining tool in Microsoft Power BI. For us it was a logical decision with a lot of advantages that come from this combination. But some of you may wonder what Power BI even is. What can you do with it? Why is it a perfect combination with Process Mining? What advantages do you get from using it? And what do we mean, when we talk about democratizing BI and Process Mining?
A bit on Business Intelligence
Business Intelligence (BI) platforms include procedures and processes for collecting, evaluating and presenting data with the aim of increasing added value, reducing costs and minimizing risks. All possible KPIs can be evaluated – information about your own company, competitors, customers or market developments. The prerequisite for this is that users must specify exactly what they want to investigate – and what the goal of their analysis is. BI brings data together, shows patterns and discontinuities and answers previously defined questions.
Features of a good BI tool
Easy to use
UI-based functionality that allows users to drag and drop, has click-and-point features and easy selection functions
A good BI tool also offers a mobile version. It allows for accessibility no matter where you are. Plus, you can receive alerts and access monitoring dashboards on the go.
Single Platform Model
But, no matter which device you are using, you should look for a single platform approach. You gain an end-to-end solution that covers gathering, analyzing, and interpreting data. With a single platform model everything is working together so you don’t have to worry about compatibility issues.
Enhanced analytics help users in gathering, analyzing, interpreting, and conveying information by simplifying and automating tasks.
This includes, for example, automated data preparation, collecting information from multiple sources, and reducing the chance of errors.
You can also use embedded machine learning and AI features for predictive analytics and forecasting without the need to know coding, for Q&A with natural language, easy access and quick and smart searches.
Diverse Visualization Features
Your data is automatically transformed into pie charts, graphs, or other types of visual presentation. You can quickly and easily see and understand patterns, relationships, and trends with no need for specialized training to interpret the graphics.
You can decide between numerous options which graphic is suited best for presenting the data or use automatic recommendations based on data results.
Collaboration and Accessibility
Cloud-based BI with prebuilt connections is the fastest and easiest way to get started and receive results.
When you choose a cloud-based BI tool, it is very easy to access by users across your company and on the go. This makes it easy to collaborate with coworkers and communicate issues to the right person.
A BI tool with prebuilt connections to other applications also reduces the time to get started and is accessible to people with no IT background. This also takes work off from the backs of your IT team, so they can focus on other, more complex tasks.
The Blind Spots in BI and Traditional Process Management
When you just look at BI and what it can do, you might think that there is no need for a Process Mining tool. Afterall, BI already comes with great analysis and visualization capabilities. Plus, you could just use traditional Process Management, such as workshops and interviews, as a basis for process optimization projects.
But there are a number of disadvantages to this method.
When you use interviews and workshops you likely fall into a trap that’s very common in social science research:
Social Desirability Bias and Prestige Bias
Both forms of bias impact the answer you get from the people you interview.
Social desirability bias describes a tendency of interviewees to respond with what they think is the socially acceptable or favorable answer. This can lead to them either over-reporting ‘good behavior’ or under-reporting ‘bad’, or undesirable behavior.
Prestige bias describes interviewees who want to be perceived ‘better’ than they actually are. This can mean, for example, richer, cooler, more competent or tougher.
This type of bias can lead to completely wrong numbers or pictures of reality. For example a telephone survey in the US in 1992 on gun-use found that about 845.000 American citizens had used a gun to defend themselves from burglaries over the year. But crime statistics only reported 430.000 burglaries where someone was at home and awake to defend themselves.
So what was it? Did people want to appear more heroic or did they think it was more socially acceptable to report that they actually use their guns for self-defense?
Either way, interviews and questionnaires produce unreliable answers, so they are unsuited to set up KPIs and monitoring dashboards.
The same can be said for only using BI tools because you use them to focus on the things you know are there. But this will not lead to the desired results.
This example should make things clearer. Let’s pretend you’re coaching people on business success and you’re only giving examples of highly successful individuals such as Mark Zuckerberg, Steve Jobs or Bill Gates. Would you tell people to drop out of college to start their own tech-business?
The answer is most likely ‘No’. This would be falling into the trap of survivor bias. You only look at the people or businesses who ‘made it’ but you’re missing the mistakes of those who didn’t. Seeing what they did wrong will tell you a lot more than looking at the stuff others did right.
That’s actually the biggest weakness of standalone BI tools: Do you always know in advance what is important? What if you completely overlook the things that really matter? What if you ask the wrong questions?
Combining Process Mining and BI
Process Mining takes a different approach. Instead of looking at isolated KPIs you look at your processes from end-to-end. That means, you don’t need a predefined question to start your analysis. You can take an unbiased look at your as-is processes, identify previously unknown problems and optimization potential and use this knowledge to directly start optimization measures, define relevant KPIs and monitor your progress.
Does that mean Process Mining is better than BI and can actually do more? Not quite. Process Mining works best with event data. BI tools will likely be stronger with any data that’s not event-based. Alternatively, you could transform all your data into event data, but that takes a lot of effort.
That’s why combining state-of-the-art BI and Process Mining tools is a natural choice. You get the best of both worlds.
What’s more, Process Mining works best and can achieve the best results when there are a lot of connectors to databases and other data sources for your event log and action connectors to share, collaborate and trigger workflows. The greater the variety of connectors, the more versatile are the uses of a Process Mining tool. BI tools provide the right framework with all these connectors. And Power BI has an incomparable set of connectors.
Process Mining in Power BI
Since its introduction in 2014, Power BI has grown into the market leader in the BI market. But even before that, Gartner has listed Microsoft as a leader in BI solutions, now for the 13th year in a row.
Power BI is part of Office 365 and Microsoft’s Power Platform which are used as the primary office suite by the majority of businesses and organizations across the globe.
For all these organizations, it makes sense to use Power BI instead of third-party solutions, as it is part of Office 365 and meets all the criteria of highly advanced BI tools:
- It is cloud-based, so users have access to cutting edge intelligence capabilities and powerful algorithms from anywhere within a company structure
- Various visualization possibilities and templates allow businesses to really understand their data
- Built-in machine learning features help users to analyze data and to spot trends or make predictions
- Alerts can be set up on KPIs to keep informed about important metrics and measurements.
- Data security plays an important role in Power BI and there are security measures such as granular controls on accessibility both internally and externally
- Power BI comes as an app for Android, iOS, and Windows 10 devices, as well as a free desktop version and different plans for pro and premium versions.
Power BI Advantages
But there is more to Power BI than just great BI capabilities. Usually BI solutions are mostly used by business analysts and data scientists. But Power BI can be used by a wide range of people in a company because of a very user-friendly approach.
On top of that, users highlight the visual appeal, reasonable licensing cost and customer experience as some of Power BI’s strong points:
- Power BI can deal with especially large quantities of data
- The intuitive interface combined with a lot of personalization options make it easy to access and analyze data quickly
- As part of the Microsoft family, Power BI integrates easily with other Microsoft tools such as Dynamics or SharePoint, and with non-Microsoft products such as SAP® or Salesforce.
- Power BI can be used as a self-service tool and helps less experienced users with automatically generated dashboards, AI features such as Q&A or QuickInsights and Machine Learning capabilities
- It works with any data, so whatever you want to report on, you can do it
Considering the user-based approach and great feedback, it’s not surprising Power BI experienced a great growth of its user base throughout the last year.
Power BI and PAFnow
Now add to that the Process Mining features of PAFnow and you have a perfect match.
Every Process Mining Applications needs connectivity to backend systems as well as analytics capabilities and visualization capabilities. PAFnow is simply built on the Power BI platform, allowing clients to use the best of Power BI and PAFnow Process Mining in one solution.
In contrast to other Process Mining tools, PAFnow does not compete with the BI reporting infrastructure implemented in the company. It’s actually the opposite: PAFnow fits seamlessly into the existing reporting infrastructure. It runs smoothly inside Power BI. For all clients using Power BI, PAFnow is a natural choice.
You can use all PAfnow Process Mining advantages within your infrastructure. There is no need to learn a new tool. Users profit from a very short familiarization period, which leads to a high acceptance of the product.
In addition to the PAFnow algorithms, PAfnow Process Mining users can also simultaneously use Microsoft AI, machine learning and Q&A for their process data. And Monitoring, Alerting and Collaboration capabilities help you to convert process analysis and insights into key metrics that you want to keep track of.
Power BI’s numerous data source connectors allow you to extract data from various information systems while a variety of action connectors enables you to inject your Process Mining results into other applications, collaborate and take appropriate actions for process improvement.
The findings can be integrated and displayed as dashboards or reports on a website, on a mobile app or another application. So they become available very easily and quickly via the digital workplace Office 365 exactly where process problems arise and can be eliminated or prevented.
The integration of PAfnow in Power BI also renders the discussion of ‘What is better – BI or Process Mining?’ redundant. Now, it is about making Process Mining easy, convenient and affordable and thus accessible to all end users via the leading BI platform. This is exactly what PAFnow does in Power BI.
Democratizing BI and Process Mining
Microsoft’s vision of Power BI is to make the leading analytics / BI solution as easy to use and accessible as possible. This matches the PAF philosophy of democratizing Process Mining, in terms of accessibility, usability and intuitive operability.
This means listening to user feedback, lowering entry barriers, and focusing on a low-code or no-code solution, that can be used by people with no coding background.
It also means access to Microsoft AppSource and a variety of visuals for your report. Overall there are more than 150 visualization types (such as bar charts, pie charts, treemap, maps, area chart, etc) which are certified and provided by Microsoft, and community members for the global Power BI community. You can also create and offer your own custom visuals in AppSource.
In the next part of our series we will talk about common misconceptions and misunderstandings surrounding Process Mining as a technology.
If you have any questions or want to learn more about Process Mining, try out our demo version, contact us, get the free visual, or just contact us. We are happy to help you, answer your questions and tell you more about the mission of the Process Analytics Factory (PAF).
Part 4: Process Mining in Power BI