Competing on Analytics – Time for PLM based Analytics?

This recent newspaper article caught my eye: “Mu Sigma receives $108 million funding” []. “Mu Sigma, the analytics and decision-support services company, has received a $108 million investment round led by private equity firm General Atlantic. While $93 million came from General Atlantic, the remaining came from venture capital firm Sequoia Capital, which had previously (in April) led a $25-million round of funding of Mu Sigma.”

There has been some recent action going on in the analytics software market. In September, US-based predictive analytics provider Opera Solutions received $84 million in funding, which the Wall Street Journal said values the firm at $500 million. After acquiring crime data intelligence software developer i2, Big Blue bought Canadian risk management analytics software developer Algorithmics for $387 million in the same month.

Although BI tools have been there for more than a decade now, the recent fillip to analytics was in part provided by Thomas H. Davenport’s 2006 in his seminal article in Harvard Business Review “Competing on Analytics“.  This was followed by two books, “Competing on Analytics: The New Science of Winning” in 2007 and “Analytics at Work: Smarter Decisions, Better Results” in 2010. To quote Davenport: “It’s virtually impossible to differentiate yourself from competitors based on products alone. Your rivals sell offerings similar to yours. And thanks to cheap offshore labor, you’re hard-pressed to beat overseas competitors on product cost. How to pull ahead of the pack? Become an analytics competitor: Use sophisticated data-collection technology and analysis to wring every last drop of value from all your business processes. With analytics, you discern not only what your customers want but also how much they’re willing to pay and what keeps them loyal.”

“In God we trust; all others must bring data.”

The catch phrase “Big Data” has been bubbling up from different vendors in the past few months. While it is debatable whether PLM systems would have “Big Data”, in a large organization which has such PLM system for a few years would have amassed a huge amount of data related to typically BOM’s, Product Structure Metadata, various types of associated documents, change management and transactional data related to business process activities. Typically all PLM vendors provide a search tool (though some provide an index-search tool too). However most of the vendors seem to be weak on providing an analytics tool. A PLM based analytics/BI product is should not be a just high end fancy reporting tool with low TCO, out-of-box templates etc. – there are some fundamental difference between reporting and analysis. While both draw upon the same collected data, reporting and analysis are very different in terms of their purpose, tasks, outputs, delivery, and value. Without a clear distinction of the differences, a PLM customer may sell itself short in one area (typically analysis) and not achieve the full benefits of its analytics investment. Analysis focuses on questioning, examining, interpreting, comparing, and confirming. [http://www.adobepress.com/articles/article.asp?p=1811387&seqNum=3]

“If you do not know how to ask the right question, you discover nothing.” – Deming

A PLM based analytics solution should clearly be able to do:

1. Predictive Analytics and Data Mining – Example: predicting how different parts in a BOM affect MTBF, MTTF, Spares Consumption, etc.
2. Data Visualization – Example: Present a multidimensional analysis of volumes of product data to get information about the business in a more intuitive way.
3. Forecasting – Example: Analyze and predict future product needs based on historical sales patterns, planning optimized production and project schedules.
4. Product Costing – Example: Analyze how different substitutes and part variants (either in-house or from suppliers) can affect the cost of a product.
5. Quality Improvement – Identify, monitor and measure quality processes over time and provide insight into how changes to various parameters affect final product quality or customer satisfaction.
6. Statistics – Use statistical data analysis to drive fact-based decisions as in Statistical process control (SPC).
7. Text Analytics – Example: Text mining customer complaints for trends and identifying problematic business processes based on the analytics. Social media analysis capability is also slowly gaining ground.

Five Emerging Trends in Analytics

According to Jane Griffin there are Five Emerging Trends in Analytics:

1. Increasing sophistication of analytics software and the need for a simplified approach to presenting the results of complex analysis,
2. Transition of analytics out of the office and onto the mobile devices of executives,
3. Companies starting to realize that analytics gives them to gain a competitive edge,
4. Companies are having (problems) dealing with “big data,
5. And, the biggest sea change of all the trends, is that agile development techniques, new data quality tools and methodologies, and technical innovations like in-memory analytics are taking analytics (and business intelligence in general) out of the realm of the data warehouse.

In my opinion PLM vendors currently provide fragmented analytics tools – Most of them address only a very specific section of the addressable market need. Also the problem seems to be compounded by the fact that not only companies implement a closed-loop PLM system thereby making data collection much tougher. PLM vendors should start taking note of the emerging trends (specially the “big data” problem) and start providing value-added analytics products to customers.

This entry was posted on Monday, January 30th, 2012 at 4:29 pm and is filed under Analytics. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.