Sitting here waiting to board my flight back to NC and reflecting on the last
Gartner Business Intelligence Summit session I attended today that brought it home for me.
It was
Blending Data & Expertise at Cisco: How Analytics can Boost Forecast Accuracy. Anne G. Robinson Ph.D and Senior Manager of Information & Data Strategy Customer Value Chain Management shared how Cisco re-engineered their customer value chain and planning process.
She started her session by defining analytics – advanced statistical, mathematical and algorithmic techniques used to solve a business problem. Analytics is not frequency, queries, OLAP or simple reports (Amen.) Coming from SAS and being in the BI realm for so long, this is always a good level playing field to start with - it eliminates translation issues.
And then she started telling her story....
For years Cisco had a top down marketing driven forecast and financial planning process. Faced with a complex and diverse product range, they would go through the business planning process and eventually roll down into the demand planning process – that wasn’t driving an optimized supply chain.
The company had a one year planning horizon plan that while global and representing dollar allocation did not have accountability, transparency, or instill a great deal of confidence in the process.
Then a major shift occurred. Cisco rethought their business value chain and married supply chain, customer quality and demand planning to form a Customer Value Chain that would outsource 95% of manufacturing. This new group employs over 8000 and manages more than 1000 suppliers.
Key to this new model was the move to a new planning process that would close the loop, integrate analytics, instill confidence and trust in the new process, and deliver timely forecasts so that demand could be managed strategically.
Dr. Robinson started with the people strategies that helped this project succeed.
• Find an executive sponsor who will remain involved throughout the initiative and drive key decisions.
• Create an analytics team: Robinson started with a team of 10 with a mix of statistical forecasting, supply chain analytics, BI and data management experience
• Define the vision. The Cisco vision was to improve forecast accuracy, to have a non biased data driven process that was flexible and could scale to the thousands and thousand of products needed.
• Reshape existing roles: The new process would create a closer relationship between marketing and the supply chain. Focusing on areas of expertise, respecting areas of knowledge while partnering in the process with IT and
As the analytics team started to look at their part of the project, a strong partnership with IT began at the same time. Dr. Robinson said she made it clear that she expected the business folks to understand enough to be able to effectively think through impacts to IT when a rep wasn’t in a conversation. Conversation and communication between the business and IT group was key to the initiative’s success.
After forecasting software, processes were identified, stakeholders brought in and the analytics team got to work on the identifying the best forecasting models, etc. an effective way of integrating analytics into the decision process began. The team evaluated which forecasts could be automated for products that were easier to manage and could then focus analytical methodologies on more complex product mixes. Cisco uses SAS for their
forecasting process.
Cisco instituted the following demand planning steps that include:
• Independent forecasts and analysis
• Consensus meeting with all stakeholders (decisions on thresholds, concerns and building trust in analytical approach)
• Demand plan generation
• Close loop with customer value chain execution and business (marketing, sales, etc)
Results of the new process include
• Comprehensive forecasts for ALL products globally (18-24,000)
• 12% forecast accuracy
• Forecast bias tightened up
• Statistical forecasting process completed in 10 hours and weekly execution and now days and hours
• Vision for the future was extended to multiple quarters view
The audience seemed to really enjoy the presentation - it was so encouraging to see the interest in how to blend analytically driven methods, like forecasting, into the business. Lots of folks had questions about how to help business and IT connect and communicate – or how to help the analytics experts translate business value.
Dr. Robinson had senior executive support and sponsorship that I think helped align everyone in the group – regardless of function – to identify with common goals.
Gartner BI Summit was a good event this year but I can see that the types of questions and sessions that folks have are evolving due to economic factors and other issues. I am interested to see what happens next year but for now – gotta get ready for
SAS Global Forum at the same hotel in just a couple of weeks!
Forecasting. What does it mean to you and your organization? For two days in June here in Cary we'll host the fourth annual F2009 and conference so that forecasting professionals who help their organizations plan and compete more effectively can share be
Tracked: Mar 17, 17:30