Hide this madness

CDO Toronto Executive Summit – Jonah Group Summary Report

We were very proud to sponsor the inaugural CDO Executive Summit in Toronto on June 24, 2014. The following are some highlights from the sessions we attended:

  • From Raw Data to Actionable Insights – Pranay Agrawal, Wojciech Gryc, Paul ter Weeme, Eric Monteiro: The panel agreed that data analysis projects shouldn’t last longer than 10 weeks. If you can’t find the signal in 10 weeks then it probably isn’t there or you are asking the wrong question. You can calibrate your analysis by splitting historical data into old and new data sets then running the analysis on the older data set and comparing predictions with the newer data set to see if they match. The panel agreed that you need top-level buy-in and a clear mission to run a successful data analytics project. Another theme echoed in other sessions is not to fall into the “must have perfectly clean data” trap – clean data is always 2 years away.
  • Know Your Customer Analytics – Victor Dudemaine: Victor points out that companies who spent less on analytics also have a lower ROI. Companies must learn to use data to remain competitive. Sun Life operates 12 mature data warehouses. Victor’s team prepared a chart showing which data sets support each business imperative. Overlaps represent opportunities for re-use. He used the chart to get management by-in for a centralized data store that reduces development and maintenance costs. The chart also provides a road map and score card for the project.
  • Data Management Evolves – David Besemer: The session focused on Cisco’s logical data warehouse product called Composite Information Software. There are compelling use cases for the architecture. A logical data warehouse layer allows quick mapping of data sources without moving original data, effectively allowing a choice of isolating changes (e.g. when moving legacy data warehouses) or rapid response to business needs vs. traditional enterprise data warehouse approaches. The relatively low cost of a logical DW layer makes it worth consideration.
  • The Potential of Data – Eric Monteiro: The session was aimed at retail marketing and the idea of “interception” to send targeted offers and incentives to consumers based on their personalized needs. Eric gave good high-level examples of end-to-end consumer experiences (e.g. planning a trip, flying, and returning) that are inadequately served today. My takeaway is to look at how customers experience products and services before, during and after their interactions with the seller in order to find ways to improve their experience. If executed well this will create opportunities to increase customer loyalty.
  • Big Data Alchemy – John Weigelt: John compares data analytics to making base metals noble. He demonstrated new features of Excel 2013 including Geoflow, which automatically maps geographical data sets. Microsoft is working with a property management company in Chicago. They estimate they can save $100 million/year by monitoring and analyzing energy use in their buildings. The post office uses machine learning to teach computers to read addresses. Since 1997 automated routing has increased from 10% to 98% of all mail items.
  • Privacy by Design – Dr. Ann Cavoukian: Ann points out that privacy isn’t just about security, although security is necessary in any privacy scheme, it’s also about control. People might share their personal data if they get something in return but they need control over how and when their data is shared. The Germans call this information self-determination. Ann points out that companies can use privacy as a differentiator. Customers will reward companies who protect their privacy. Companies can use this customer goodwill to break free of the zero-sum game often attributed to privacy endeavours.
  • Shining a Light on the Dark Data – Daniela Crivianu, Adele Pugliese, Mara Lederman: Data that has not yet been illuminated by analysis – or possibly not even discovered yet – is dark data, and it is a strategic asset waiting to be mined. Daniela joked that her approach of performing a data survey to create a “database of databases” was perhaps banal, but nevertheless useful. Dissemination of information, ongoing communication, and clear roles and responsibilities were critical to overcome the internal human barriers against participation and adoption. Persistence paid off as users began to appreciate the value of the illuminated data. Additionally, the panel agreed that the overall approach must be business-centric, and must not be executed as, or presented as, an “IT” project.
  • Digital Process Management – Neal Oswald: This was my favourite session of the day. At CIBC Neal and his team developed a complete digital pipeline view of CIBC`s mortgage lending business across disparate data siloes and business processes. Within 22 months the team had developed dashboards, visualizations and deep drilldowns of process activity to measure productivity at every stage in the process. Using agile methods his team delivered new features and insights every 2 weeks, incrementally building on systems and delivering quick wins to gain momentum and credibility. The technology was straightforward; the biggest challenges were dealing with culture and business resistance to change.
  • Hand-Engineered Data Revolution – Rick Kochhar: Rick mentions that those companies that viewed Basel II as an opportunity to create new systems to generate rich data, rather than just a costly tactical exercise, ultimately achieved strategic risk management wins including reduced capital reserve requirements and increased profitability. It is important to view data as strategic on its own, and not just as the by-product of operational systems. Data quality is a process not a destination and data lineage is a property that is often overlooked. For some data, like financial transactions, good means 100% accurate while other data sets only require partial accuracy.

The Jonah Group has developed a data discovery process that supports agile development. We are currently integrating modern tools like Hadoop and MapReduce to further reduce the development time and operating costs of analytic systems. Please contact Drew Atkins for more information.

Add a comment

Comment feed
The better to greet you with
No one will ever see this
Your pride and joy
The reason this comment form exists

The crew behind ASOT

We're a team of interactive, software, and business intelligence experts skilled in the design, construction, and management of online enterprise systems.

Visit The Jonah Group site

Get in touch with us