Customer case studies

 

Diageo logo Cognos

Key requirements for the new software are overall system performance i.e. speed, being tailored to current planning processes i.e. providing relevant and convenient set of menus for data entry, manipulation and analysis.

Show/hide details

Diageo has decided to seek an advice on changing the application so that it can be deployed to users for data entry directly into the system or linking data in from Excel spreadsheets for Annual Operating Plan (AOP) and Risks & Opportunities (R&O), data consolidation and collaboration work, aggregation of data and upload into the BW.

The following is the list of client requirements implemented for the system:

  • Built AOP application using Cognos delivering improved application responsiveness and performance in preparation of the AOP;
  • Built R&O solution for the preparation of the monthly forecasting and ad-hoc updates;
  • Migrated detailed calculations for the various expense line items into Cognos from Excel i.e. ensuring that the application reflects current business model;
  • Improved maintenance and management of the system;
  • Deployed the solution to users within the organisation so that they can input data directly into Cognos;
  • Used existing structures from SAP e.g. SKU codes, improving integration with BW.

Originally Diageo Australia has been using Cognos since 2004 for their AOP only. It was intended to be deployed to users across the organisation. Since then the application has been predominantly used by the finance team as a staging point for collecting and aggregating data for AOP. Excel is still being used to calculate data that is then uploaded into the Cognos using functionalities available in the software.

The new solution has significantly improved on that.

The key benefits recognised:

  • Plans and forecasts are prepared by individual Business Units and consolidated at the company level removing need for data collating from multiple spreadsheets;
  • Done by SKU and month improving accuracy and transparency of plans and forecasts;
  • Actual data is being uploaded from SAP data extracts monthly for entire business from a single data source;
  • The solution allowed planning and forecasting down to Operating Profit level;
  • Planning for Net Sales is done with detailed calculation for trading terms and discounts eliminating possible calculation errors;
  • Annual Plans are then uploaded into SAP from a single data source.

Implementation and solution support at Diageo Australia / 2010 - 2017

 

Qantas logo Cognos TM1

Qantas Freight deployed IBM Cognos TM1 platform for forecasting and reporting. The solution encompasses a multi-week rolling forecast for every Qantas terminal around the world. The system is being used for weekly revenue and costs forecasts, based on historical data. Automated upload of actual data and data refreshing to the data warehouse were deployed to run on a weekly basis. The TM1 solution delivered capabilities of weekly forecasting to Qantas Freight, on-line consolidations in multiple currencies, and data source for BI reporting.

Show/hide details

The following is the list of client requirements for the system:

  • Build the web-based forecasting solution to collect data for each sales station on a weekly basis;
  • Automate updating of sales station, currencies, terminals and other metadata;
  • Pre-populate the forecast based on a client-provided algorithm using multiple prior forecasts and actual data for given number of years;
  • Deploy the workflow and monitor the progress of forecast submissions;
  • Automate exporting forecasts to the data warehouse in the specified format;
  • Deliver data in multiple currencies with data input being possible in any currency;
  • Provide ability to keep any number of prior forecasts for reporting and comparisons.

I have built the system that has delivered on all requirements and demonstrated exceptional performance with several historical versions. The system was designed and built with a modular approach to TI processes, ensuring portability between environments and ease of maintenance. The TI processes built were using extensive parameterisation and dynamically-created SQL queries to achieve best performance.
MDX code was used to maintain dimensions and subsets. Use of MDX was in line with IBM-recommended best practices.

The technical implementation of the solution encompasses the following:

  • TI process for weekly data archiving from current to prior version keeping several rolling historical versions for reporting and comparison purposes;
  • Weekly data uploads from the Data Warehouse, using TI processes for both metadata and data;
  • TI process that triggers pre-calculation of the forecast, based on the client-provided algorithm and then pre-population of the multi-week forecast;
  • TI process using IBM-built system TIs to reset workflow status to not-started.

The system has delivered the following benefit to Qantas:

  • Ability to forecast data on a weekly basis;
  • Ease of data collection and consolidation;
  • Availability of historical data;
  • Minimal support required as almost every process is automated.

Having deployed a TM1 based system, the client spends less time on data collection and processing and has more time available for analysing this data. There are more individuals involved in data collection and fewer in data manipulation and processing, reducing the probability of error and improving the quality of data for the management.

TM1 implementation at Qantas Freight/ 2011 - 2012

 

Shriro logo Cognos Planning

Shriro Australia uses Cognos Enterprise Planning for their monthly brand forecasting across Australia and New Zealand. I have been providing professional services to Shiro for a number of years, upgrading software, implementing change requests and optimising application performance. One of the technological challenges was to enable consolidation of forecast data, only for those businesses that completed and submitted their forecast, leaving others unchanged from the previous month. I used Cognos Framework Manager to deliver the solution, greatly improving data accuracy and saving time previously used for reconciliations.

Show/hide details

The change presented additional functionality for the Cognos Planning application that was not available out-of-the-box. The client clearly identified objectives for the functionality enhancement as follows:

  • IBM Cognos Planning is to be used to collect forecast data by brand from users across the organisation;
  • If a user wishes to seek an amendment to the forecast, they are required to make changes to the current brand forecast and submit changes to the management for review and approval;
  • The management reviews all submitted brand forecasts each month and can reject amendments for various reasons using in-built workflow functionality of Cognos Planning;
  • Only approved amendments will need to be consolidated and then accepted by the ERP system and used for company-wide reporting.

Technological challenge:

  • Cognos Planning system is required to use approved amended brand forecast, leaving the existing forecast for rejected or unchanged brands intact and form a new monthly forecast.

Solution:

  • The Cognos Framework Manager was used to implement the solution;
  • Data was published into a SQL Server database using the 'table publish' functionality;
  • Framework Manager Model was built that used a filter to select only the submitted brands;
  • Administration Link was then used in the Contributor Administration Console to copy forecast data into the ‘Latest Approved Version’ for those brands that were submitted and not rejected by management, leaving the forecast for unchanged and rejected brands intact;
  • The Cognos Package that was created from the Framework Manager was used as a source for the Admin Link;
  • The macro was used to execute the admin link and was made available in Cognos Connection. The system administrator then scheduled this macro to run during the month-end to prepare the current forecast.

Benefits:

  • The client now spends less time verifying current forecasts and comparing numbers for individual brands, ensuring that users do not change forecasts without submitting them to management and allowing unauthorised changes to be included in the current forecast. This saves considerable amount of time and effort previously used for data validation by implementing the system that prevents any unauthorised change;
  • The change brought the data integrity that the customer was looking for, allowing the management to rely on their current data to make their business decisions.
  • IBM Cognos products were deployed to deliver the complete solution, ensuring the client can maintain the system in the long-term.

Cognos Planning implementation at Shriro Australia / 2008 - 2012

 

Deloitte logo TM1

Deloitte Australia has deployed IBM Cognos TM1 to process monthly costs allocations and prepare a transaction journal based on the outcome.
The TM1 solution has achieved considerable improvements in transparency and performance, providing accuracy and reliability.
Managed by SMS MT Ltd, I was engaged as the solution architect and the lead developer. The TM1 project delivered on all requirements within the timeframe and the budget allocated.
Having implemented the original model using IBM Cognos Planinng in 2006, Deloitte used my expertise to migrate the solution to IBM Cognos TM1 in 2012. Having a thorough and robust knowledge of both platforms allowed me to facilitate the transition from one platform to the other with minimal disruption to the client.
The solution deployed advanced Turbo Integrator techniques, MDX, and high degree parameterisation to achieve considerable performance improvements. I followed best practices from documentation and naming conventions to code handling and user acceptance testing.
The solution recorded 20 seconds to process data compared to 40 minutes in the model it replaced.

Show/hide details

The TM1 project at Deloitte was initiated to deploy new solution in allocating costs from non-client facing to client facing cost centres.
Client requirements:

  • Run allocation for multiple versions i.e. actual, forecast, budget;
  • Run cascading allocations;
  • Ability to adjust any allocation using a range of techniques;
  • Achieve considerable performance improvements

I was engaged by SMS MT as TM1 Consultant to collect requirements, prepare the solution design document, and facilitate signing off by the client. I have built the prototype, led the development effort and provided quality assurance for the project. The project was completed within the timeframe outlined in the project plan and in line with the client expectations.

Technology:

  • IBM Cognos TM1 was chosen to deliver the solution;
  • Integration with the Data Warehouse was set up to update metadata, upload data and export results;
  • TI-created dynamic SQL was used to query databases for optimal performance;
  • TI processes were used alongside feeders in the TM1 server to achieve maximum performance;
  • Dynamic TI-created views and subsets were used;
  • MDX was used to maintain subsets;
  • Data output was processed in the format specified, ready to be posted as a journal;
  • Parameterisation was used to simplify migration between environments i.e. development to UAT to production.

The project achieved tremendous performance improvement over the existing model, above and beyond client’s expectations. The client has conducted thorough testing and I was responsible to addressing any issues. I also prepared the system document and conducted knowledge transfer so Deloitte became self-sufficient in making adjustments to their new cost allocation solution.

Achievements:

  • Delivered functionality to run allocations for 4 or 5 versions and up to 18 months;
  • Capability to use different allocation settings for each version;
  • Built functionality to perform cascading allocations;
  • Delivered error reporting for under and over-allocation, considerably improving the transparency of the process;
  • Delivered functionality to adjust any allocation for any version or time period using percentages or fixed values;
  • Delivered functionality to temporarily remove any allocation or change their sequence;
  • Achieved performance improvements for the solution to process 3 versions for 12 months in 20 seconds compared to previous model running for 40 minutes for one scenario at a time.

The TM1 project served as a testing ground for further TM1 deployment in different parts of the business.

TM1 implementation at Deloitte Touche Tohmatsu / April 2012

 

MORE CASE STUDIES COMING SOON