Exclusive portal designed for our Customers and Partners

Log in

Request for Access

logo
logo
countries flags

Blog post series: Plan your predictive analytics and machine learning journey with CCH Tagetik - Part 1

Jun. 5 2020 by Prof. Dr. Karsten Oehler, Solution Architect - CCH Tagetik DACH / Marco Van der Kooij, Managing Director - ForSight Consulting

Performance Management

Artificial intelligence and machine learning in particular, are the wind of change blowing through data-driven decision-making. So welcome to this series of blog posts about how you can plan your journey into predictive analytics and machine learning. You will learn practical advice on how to get started and how CCH Tagetik Finance Transformation Platform supports your journey.

The need for speed in deploying new developments tends to be a big issue for the Office of Finance. It really does help CFOs to be “fast followers” as Ventana Research put it. This means being ready to grasp the opportunities machine learning offers, while being in a position to learn from the mistakes of others. So lay your foundation ready to build on, comprising a centralized storage of high-quality financial and operational data, plus a modern enterprise performance management platform and the right people with analytical skills.

We’re often asked “how can machine learning and predictive analytics support Finance?”. Here are some major areas where AI and ML could make the difference. As the illustration below shows, they’re all deeply connected:

  1. Data quality improvement and data collection
  2. Profit and Loss forecasting
  3. Contributor analytics
  4. Driver based simulation

All the above are among the most typical pain points, which are identified in this case by business experts in a survey from the German BARC Research. Let’s face it, good data quality and data collection are simply essential for successful Sales and Revenue forecasting, particularly if you need to go in depth in a cost forecasting process, play with advanced simulations, etc.

As we said before, if you want to get a successful predictive analytics project off the ground any time soon, lay the foundations of data quality now.

In this first blog of the series, we will focus in the first area: Data quality improvement and data collection.

Improved data collection

To apply machine learning successfully to large data sets, you need to bring both financial and operational data together.

How can machine learning help to improve the acquisition and the quality of the data from different sources? Established ETL-Tools provide functions like recoding, filtering or pivoting. Yet there are plenty of inaccuracies which cannot be handled with conventional ETL, such as:

  • Duplicates (often quite complex to unearth)
  • Data fields not completely filled
  • Data not properly assigned to periods and other dimensions, which raises the risk of outliers in the data
  • Structural breaks limit the length of time series and bias driver relationships

Outliers have to be spotted to train the software, based on every field provided from the data source. This is where sophisticated enterprise performance management software comes in because when used for planning, consolidation and reporting processes, validations are used to check the quality of the data. Mostly, this is at an aggregated level for the periodic consolidation.

But with more granular financial and operational data, different validations are needed to make sure the data is correct (in other words, registration of transactions must be done correctly). This is where machine learning steps up as it can be applied to train the software on outlier detection. It can automatically apply translation rules to ensure data is attached to the right spot in the data model to achieve advanced analytics.

So what are the key areas where machine learning (ML) can offer a meaningful boost? Here are some items to specify for ML support:

  • Imputing
  • Outlier detection and adjustments
  • Intelligent assignments
  • Text crawling

And the data sources:

  • Financial and management accounting
  • Customer / prospect data
  • Economic data
  • Supply chain data

CCH Tagetik steps in facilitating data preparation

CCH Tagetik’s powerful Analytic Information Hub is the foundation for a successful predictive analytics project. This native data engine allows collection of any kind of data – financial and operational – from any source – structured and unstructured – and manage this information directly in a single solution. Its powerful capabilities for data preparation through Data Transformation Packages (DTP) mean you can load and validate financial and operational data from various sources via automated business rules, for e.g. outlier detection, and use it to accelerate data intelligence.

However, data preparation is only the first step. Stay tuned to this series to discover other areas where AI and ML could make the difference in supporting Finance!

To discover more download the white paper "Machine Learning for Controllers. Use cases for Forecasting, Planning, and Simulation" click here.

 

Share this post!