Free cookie consent management tool by TermsFeed Process Mining, Start to Finish [Process Mining v5.8]
Process Mining, Start to Finish
This content applies solely to Process Mining, which must be purchased separately from the Appian base platform.

Before you embark on a process mining project, you may want to get a sense of the time and effort it will take. Process mining not only involves a significant amount of data and analysis, but also coordination and communication across your organization.

This page provides a high-level overview of what happens in a process mining project. Where applicable, we've listed the module you'll be using to complete that step.

To learn more about process mining in general and how Appian can bring you from knowing to doing, check out Process Mining or refer to the following diagram:

pm-overview.png

Step 1: Plan your project

Your goals for process mining should align with corporate goals. The business problem you identify directly determines the data that is needed and who to involve on your team. Work with those stakeholders and process owners at your business to select a process to analyze. You'll also want to work with your system administrators and other IT decision makers to identify the data sources and sets that will provide you with the relevant information.

As part of the planning process, identify metrics that indicate success. For example, if the number of days it takes to fulfill an order is important, this is a metric you'll want to investigate during process mining.

Step 2: Upload data sets or connect data sources

Where? Mining Prep

To start, you'll want to get your data into Mining Prep, either by connecting a system or by uploading a file directly.

It's also recommended that you determine the viability of the data collected. If necessary, continue working with system administrators to refine data sets or gain access to additional data sets.

Step 3: Create a transformation project

Where? Mining Prep

A transformation project is a space to organize building blocks that contain data sets and transformation actions.

Step 4: Add transformation actions

Where? Mining Prep

Open the data set in Mining Prep to begin creating transformation actions. The guided experience makes it simple and predictable to prepare data for process mining.

During this step, you'll want to:

  • Validate your data to correct outliers or missing values.
  • Clean the data (remove stray commas, spaces, etc.).
  • Add or rename columns as needed.
  • Consolidate or standardize data, such as currencies or dates.
  • Add data from other sources to stitch information together.

Transformation actions are a preview of how the data set will be changed in the next step, transform and load. With transformation actions, you can test and confirm your actions without changing the data set.

Step 5: Transform and load data

Where? Mining Prep

When you're finished with transformations, send the cleaned data to process mining for modeling and analysis.

Step 6: Analyze the discovered model

Where? Process Mining

Process Mining visualizes the existing process for you as a discovered model. Use this model to see where common alternative pathways occur.

Step 7: Upload or create a target model

Where? Process Mining

Build a target model that represents how you want the process to be. You can then use Process Mining tools like root cause analysis and deviation insights to identify the differences between your current process and your ideal process.

Step 8: Filter your view

Where? Process Mining

If you find that the data is overwhelming or not useful in the current format, you can filter the information for more targeted analysis.

Filters can help you analyze one data set in multiple different ways. Case attributes that you define during Mining Prep are available as filters in Process Mining and can be used as dimensions when creating charts in a dashboard.

Step 9: Analyze process variants and deviations

Where? Process Mining

If you're looking to analyze the process more deeply, take a look at variants and deviations to see where the process occurred in atypical ways.

Variants and deviations can help provide additional insight into your key metrics. For example, if you notice it takes two more days to fulfill an order when a certain activity is skipped, this deviation might signal to you the activity needs to be changed.

Step 10: View insights from Process Mining

Where? Process Mining

On the Insights page, Process Mining provides ideas into why some issues or patterns might occur.

You can conduct root cause analysis on this page to help identify what the underlying cause of a deviation may be.

Step 11: Visualize and share the results using dashboards

Where? Process Mining

If you're responsible for sharing metrics on the process performance or other attributes, you can create dashboards to share with individuals or with your organization.

Step 12: Identify what you want to improve

Based on your analysis and your business's priorities or abilities, you can identify a list of activities to change. Process Mining may provide you with many insights or ideas on how to improve your process. However, change may need to be implemented gradually, especially for larger organizations. Carefully plan how and when to introduce these process changes based on your project goals.

Step 13: Implement changes, analyze the results, and repeat

After you identify what to change and how, it's time to make the new or updated activities reality. Engage stakeholders and process owners to influence the process as needed.

Regularly evaluate if the new process is performing as expected. You can analyze those key metrics individually, or repeat process mining to continuously improve.

Open in Github Built: Fri, Mar 22, 2024 (05:03:50 PM)

Process Mining, Start to Finish

FEEDBACK