Olaf Larson - Clarity Solution Group presentation at the Chief Analytics Officer, Fall 2016

1. The Art and Science of Real Time Analytics Olaf Larson Proprietary and Confidential - ©2016 Clarity Solution Group, LLC. All Apache and related Apacheproject trademarks or service marks are the property of the Apache Foundation.
2. Collapsing the speed of decision making? It is not about real time data, it is about real time action An opportunity missed is an opportunity lost
3. What we’re going to talk about today Faster Time to Action • How can it help • What hinders success • What can you do
4. How can faster time to action help you? Leverage of text mining through unstructured data in real time provided rapid detection of brand damaging communications in social media • This enabled was quick reaction through PR and social media sites. • The rapid reaction reduced negative consequences significantly and prevented substantial corporate losses.
5. How can faster time to action help you? Streaming analytics of online sales volumes picked up pricing errors • System used pattern recognition to quickly identify negative behaviors • Reduced losses and preventing potential damaged reputation.
6. How can faster time to action help you? Leverage of near real time online activity to shape shopping cart recommendations for major retail vendor
7. How can faster time to action help you? ( Leverage of streaming online activity data merged with preformed models allowed for rapid identification of high risk activity and potential fraud.
8. Why are some efforts less successful? Fast data and processing with no abilityor need to act Analytics need to be aligned with business value • e.g. rapid reporting of information for actions that are not high value
9. Why are some efforts less successful? Fast data and processingwith no ability or need to act Faster data without faster decisions Fast analytics that aren’t fast enough (10 seconds isn’t actually fast) Fast data, fast decisions, fast actions, no impact Your decision framework needs to be aligned to your analytics • Need ability to shift strategy at speed Architecture or Technology misaligned • Using standard platforms vs.new architectures designed to optimize quick pass through and rapid processing • Trying to push too much computing into interaction instead of leveraging callable services or restful APIs
10. What you can do Propriet ary and Confidential - ©2016 Clarity Solution Group, LLC
11. Collapse the speed of decision making • Rationalize your decision framework • Change paradigms • Improve the decision process • Get ahead of the curve on detection • Speed up consumption • Match platforms and tools to the need
12. Rationalize Your Decision Framework Key question: what decisions need to be made when given your current business strategy? • Pushing real time data through is not without costand if there is no real need for real time analytics there is a loss of ROI • Align data availability, tools, and decision engines with framework •Real Time vs. Near Real Time vs. Daily Batch vs. Monthly Batch vs. Quarterly Batch
13. Rationalize Your Decision Framework Create an agile decision framework Decentralized decision authority allows better flexibility to leverage more rapid data andinsights
14. Change the paradigm Recognize our job is not just data exploration and reporting, but driving action Move from Insight DeliveryTo Change Agent
15. Really change the paradigm Deliver decisions instead of just insights • Reduce the time to action by capturingthe decision algorithm • Requires: • Trust in your data and engines • Trust in your analytics processes and teams
16. Change the paradigm • Stop measuringyour teams on time to insight • Measure success by actions taken and business outcomes driven S u c c e s s
17. Apply Process Engineering to Decisions • Embed analytics into processes • Make the non-automated decision process more efficient • Push decisions down streamas far as possible • Empower decisioning with guard rails and business rules • ThinkGolidlocks – not too centralized, but not out of control either
18. Get ahead of the curve on detection More time is lost waiting to find out you need to take action than actually waiting on analysis to guide actions oAutomateddetection of anomalies and trend changes oPushed alerts oForecasting oLead indicators
19. Speed up consumption Information Ergonomics Cut the clutter Put insights where they will be used not in separate reports Provide insighton demand or proactively
20. Align Infrastructure to Framework Assemble and position your data based on need Make sure you can move the data and process algorithms in time Movead hoc analytics capabilities near real time • Spark provides us a common data fabric for computation across batch data integration, Data Science, and realtime streaming data Tap your “big data” where it sits using rapid analytics tools Implement agile analytics
21. Propriet ary and Confidential - ©2016 Clarity Solution Group, LLC Change Your Culture • Thinkprocesses, not products • Thinkenablement and automation • Align analytics resources with decision processes (not departments or data sources) • Get and grow the right kind of people • Think Agile – move quickly and assemble on the fly
22. Collapse the speed of decision making • Rationalize your decision framework • Change paradigms • Improve the decision process • Get ahead of the curve on detection • Speed up consumption • Match platforms and tools to the need
23. Olaf Larson Partner olarson@clarity-us.com www.clarity-us.com