Table of contents for BIWA Summit 2008
- Live from BIWA Summit – Competing on Analytics
- Powering Next-Generation Predictive Applications with Oracle Data Mining (ODM)
- Critical Success Factors for successful BI and analytic implementations
- From Data Warehousing to Strategic Data Assets
- Getting to the Right Price with Oracle Data Mining
- Oracle’s BI Strategy
- Intelligent OLAP: Data Mining and OLAP
- Fraud Detection with Oracle Data Mining
At the Business Intelligence Warehousing and Analytics Summit at Oracle today. BIWA is part of the Oracle User Group focused on BI, analytics and data warehousing. Jeanne Harris of Accenture (author, with Tom Davenport, of Competing on Analytics) started off the day. The subtitle of her presentation is “Building Competitive Strategies Around Data-driven Insights”.
Analytics are transforming all sorts of things from weather forecasting, analyzing crime to teaching. There is even a TV Show “Numbers” that is based on real cases actually solved with math as shown in the show. The math is not new but the use of it is exploding. As always people like sports examples: Tiger Woods, a sustained high performer, is an analytic type who analyzes everything he does as are sports organizations like the Red Sox and the New England Patriots. Increasingly high performing sports organizations are analytic.
Similarly businesses like Harrah’s, netflix, cemex and best buy are competing with analytics. They are enhancing customer relevance, out-executing competitors, differentiating a commodity or increasing agility. Harrah’s, for example, was in a very undifferentiated position – nothing special about customers or locations – and hired an external professor (Gary Loveman) to be COO and then CEO. Gary applied rigorous analysis and fact-based decision making to turn around the organization. He makes the point that a business that makes the same decision over and over again can get a huge return from a small insight – the basis of the profitability of applying EDM and analytics to small but highly repeatable decisions.
The research for Competing on Analytics was expected to be limited – the few companies like Harrah’s – but found hundreds. This research continues with more companies getting analyzed in more detail for the next book.
She defined analytics as “The extensive use of data, statistical and quantative analysis, explanatory and predictive models and fact-based management to drive decisions and actions”. She emphasized the importance of impacting decisions – no point to this stuff unless it changes the behavior of the company. Previous efforts to build Decision Support Systems, Business Intelligence etc were often marginal to the business – a sidebar. Companies competing on analytics embed this into the fabric of their business and their processes.
To show how this approach works widely, she talked about Will Smith. Will Smith is currently the only bankable Hollywood star, predictably helping films make more than $120M+. Will is a quant too – he was about to study math at MIT – and he and his manager analyzed the top 10 grossing movies and found all 10 had special effects, 9 had creatures and 8 had special effects, creatures and a love story! Guess what, these were the kind of movies he made. He has continued to do analysis, apparently, and he is rumored to be using the script analyzer product discussed in Super Crunchers to find the write scripts!
So why compete on analytics? Well, on what basis do you compete? Used to be monopolies, innovations, geographical access or proprietary technologies but these are all less effective these days. Analytics-driven optimization of key business processes is increasingly the only way to compete. Perhaps developing a distinctive strategy, find the best customers, optimize your supply chain or similar.
All of this is possible because:
- Data is available, especially in Enterprise Applications, and service providers provide standard external data.
- Technology can support the analytics
- Demand is there as a new generation of leaders as well as new compliance and customer-centric demands
Of these, perhaps the increasing demand is the most significant.
Why use analytics in a downturn? Well you can:
- Cut costs and improve efficiency with optimization techniques and predictive models for market shifts and changes
- Manage risk more effectively and in every transaction
- Leverage investments in IT and information
- Invest smarter so you can be ready when it ends
She showed the SAS graphic that shows BI moving from standard reports to ad-hoc reports and drill-down through statistical analysis and forecasting to predictive modeling and optimization. Increasing sophistication, increasing competitive advantage and business value.
Netflix is another example, a company built on the effectiveness of its movie recommendation algorithm. It analyzes your choices, customer feedback and more to both recommend new movies to you that you will like (so you will continue to pay the fee) and distribution of shipping requests across frequent and infrequent customers (because infrequent customers are more profitable they get better priorities). The algorithmic approach also helps Netflix correctly price distribution rights as they have the data and the movie studios don’t (because they are not analytic)!
Firms that are much less unusual are also analytic competitors (or aspiring analytic competitors) and the research found examples in every industry. Such as Cemex, the global leader in supplying cement – a highly perishable product. They use real-time data and predictive models to guarantee a 20 minute delivery window (taking account of traffic, weather, labor etc) instead of the day-long windows its competitors use. They saw a 35% increase in productivity, can charge a premium for the delivery window and have gained market share. The use of analytics is spreading in Cemex, even to the extent of using analytics to analyze how to manage the culture change caused by using analytics!
The number of analytic competitors is growing with the difference between 2002 and 2006 being very noticeable. Almost none now say they have little or no analytics. High performers are massively over-represented in the analytic companies. They value analytic skills, have above average capabilities etc. These companies find that they can deliver sustainable competitive advantage using analytics.
Making better decisions, over and over again, really makes the difference. The relentless seeking of new decisions that can be improved analytically throughout their business processes is what generates this advantage.
She outlined two paths to becoming an analytic competitor. The first is a fastpath approach, characterized by “in god we trust, all others must bring data” imposed by an analytic executive who is willing to invest and work to change culture. The second have managers who like to manage by their “gut” and this requires proof points with pilot projects, measuring benefits and gradually spreading the word.
The anaytical DELTA she said is:
Clean, common, integrated, accessible. Something new and valuable not just what’s easy. Create new metrics and measures.
Broad-based approach as you need to avoid fiefdoms around data ownership and develop some core competency. Personally I think this is an outcome of becoming an analytic competitor, not a pre-requisite.
Passion and commitment – “Do we think or do we know?” attitude needed at the top. Not just a CEO’s commitment, CFO needs to show quantative leadership around performance management, and functional leaders need to adopt analytic measures and approaches in their processes. CIOs and the IT organization must decide if they are going to just provide data or if they are going to help drive (or even own) the analytic capability.
Best target varies by industry – web analytics and recommendation for online businesses, fraud and credit for financial institutions etc. First go deep in one area, then broad. I have to say that this reinforces my thinking around centralization being an outcome not a prequisite.
Professionals and amateurs. Only 5-10% are professional analysts who can create algorithms and another 15-20% who can use them and create simple models. Everyone else is focused on spreadsheets and more straightforward analysis. You need a critical mass of analytic talent but they must be plugged into the business.
Her final example was Best Buy and she repeated a great comment – their CMO said “Marketing is now 70% science and 30% art”. At Best Buy, the analytics drive innovation – fact-based decision making based on data gathered through experimentation. They also found that you need an “industrial” approach rather than a “craft” based one.
She wrapped up by making the point that this cannot be done overnight so get started.
By the way, the BIWA SIG has its own blog – http://oraclebiwasig.blogspot.com though I don’t know what they will do with it yet.