Waterstons’ Business analytics specialists Dan, Chris and Michael will present Waterstons’ latest thinking and experience around the drivers behind analytics and intelligence in the business environment, and the current business analytics marketplace.
They will discuss Waterstons’ Business Insights Maturity Model, which sets out the methodology we use to help our customers derive competitive advantage, improve productivity and management control, and provide support for better business decision making, before using case studies to explain how real businesses are leveraging the power of modern analytics tools.
5. What does this mean to me?
@ Largely, probably nothing
@ Most businesses don’t care about Facebook, Instagram, Twitter, etc
@ It’s hard enough interpreting your own data, let alone anyone else’s
@ Maximising the value of your business information is vital
@ Once you understand what you already have you can worry about the rest
6. Understanding Your Data’s Value
@ Many businesses fail to capitalise on the value of their data
@ Predefined reports are useful, but the real value comes from the insights
hidden in an organisation’s data.
@ Business analytics is not all about answering questions
@ It’s about establishing the most insightful lines of enquiry
@ And supporting decisions with facts, trends and predictions
@ Shocking fact: few businesses are really good at using their data
7. Waterstons’ BI roadmap
@ An approach to delivering an analytics project
@ A series of delivery actions in a theoretical progression
@ A cradle to grave process for embedding, designing and developing
intelligence
8. The Utopian Process
Approach Ramp
•Selling BI as a concept
•Understanding the need, problems and solutions
Business
Background
•Understanding the business strategy, drivers, reporting requirements and KPIs.
•Decision Point
Conceptual
Design
•Understanding KPI definitions and data sources required
•Defining the data warehouse contents
Readiness and
Approach
•Analysis of systems and data availability, approach and technology selection
•Break-out Point – gap analyses, system implementation, consolidation
Functional
Design
•Reports, visualisations, dashboards
•User-group specific requirements
Build
•Build the data warehouse and ETL routines, implement chosen BI tools, develop and deploy visualisations
9. Waterstons’ BI roadmap
@ Not necessarily a linear progression; focus on business needs
@ Iteration reflects real business
@ Project experience demonstrates a different story
10. BI in the “Real World”
Approach
Ramp
Business
Background
Conceptual
Design
Readiness
and
Approach
Functional
Design
Build
11. Time for a change
@ Analysis of business data has moved on significantly
@ New analytics tools and discovery products are available
@ The popularity of, and desire for instant analysis has exploded
@ A new roadmap for delivering insight to our customers was needed
13. Where did you come from?
@ In the beginning, there was the operational report. With it came
answers to questions.
@ What am I making today?
@ Who should be at work making it?
@ Are all the parts in stock?
@ What orders am I supposed to dispatch?
@ Snapshot answers to questions – focussed on activity
14. Stage 1: Monitor
Operational Reporting
• What happened, when and where?
• Manual process
• Transactional
• Line-of-Business systems
15. Operational Reporting
@ Traditional reporting scenarios
@ Real Time
@ Detail
@ Line of business management and staff
@ Operational dashboards
@ Manage intra-daily business processes (low-latency)
@ Legislative reporting requirements
@ Traditional reporting technology is still relevant in an analytics context
@ Frozen historic snapshot
@ Controlled or limited access to interactive analytic features
@ Specific or specialised requirements for data visualisation
16. Where did you go?
@ Operational reporting was good. But not good enough. Enter the
world of Business Intelligence.
@ What has my performance been vs. KPI target?
@ How does our performance compare with last year?
@ Am I making enough things/enough margin?
@ Are the decisions I made last week/month/year making a difference?
@ Temporally related questions and answers – focussed on results
17. Stage 2: Question and Control
Business Intelligence
• What can I do about it?
• Consistent data sources
• Scorecards and dashboards
• Metrics-based
18. Business Intelligence
@ Operational Reporting
@ Data Warehouse
@ Multiple Data sources
@ One version of the truth
@ Defined KPIs
@ Targets
@ Last Year
@ Actuals
@ Historical Trends
@ Accumulation of Data
19. What’s current?*
@ Knowing what’s been going on gives way to working out what to do
next. The Business Analysis train arrives.
@ Hmm, I wonder what happened to cause that spike in sales?
@ Do blue ones REALLY work better than the red ones?
@ Are we attracting the most profitable customer groups?
@ Why, why, why?
@ Interrogation to answer ad-hoc questions – focussed on insight
*This slide formerly entitled “Where did you come from, Cotton Eyed Joe?”
21. Business Analytics
@ Operational and BI reporting
@ Analysis of reporting to ask questions
@ Why trend occurs?
@ What can be done to change trend?
@ Power users performing analysis
@ Ad-hoc reports
@ Interactive drill down / slice and dice
@ Potentially discover new trends
22. Where are you going now?
@ What’s leading edge? Data driven forecasting, strategy and planning.
The world of Predictive Analytics is growing.
@ What will next year look like?
@ What would happen if we lost that customer?
@ Could we cope if a competitor entered the market?
@ Could we satisfy demand if the number of customers doubled?
@ Trend analysis answering ‘what-if’ questions – focussed on the future
23. Stage 4: Beyond Intelligence
Predictive Analytics
• What’s going to happen?
• What-if analysis
• Strategic integration
• Business forecasts and planning
24. Predictive Analytics
@ What-if scenario modelling
@ Manufacturing; forecasting and promotion planning
@ Workflow
@ Predictive
@ Tools for extrapolation of trends
@ Recognition of trends; automatic analysis of data
@ Data mining algorithms:
@ Compute a trend for sales data
@ What characteristics make a good customer
@ How likely a student is to drop out
25. Beyond Intelligence
Operational Reporting
• What happened, when
and where?
• Manual process
• Transactional
• Line-of-Business
systems
Business
Intelligence
• What can I
do about it?
• Consistent
data sources
• Scorecards
and
dashboards
• Metrics-
based
Business
Analytics
• What’s
happening
now and
why?
• Enterprise
level data
• Self-service
data
• Insight-
based
Predictive
Analytics
• What’s going
to happen?
• What-if
analysis
• Strategic
integration
• Business
forecasts
and
planning
26. Is there life after intelligence?
@ Predictive analytics is not the end
@ As scary as it may seem, there is a Stage 5
@ Automated control methods already exist
@ Extended automation will change decision making processes
@ The ‘self-healing’ business could become a reality
@ Dependent entirely on the risk-appetite of the business
@ Not for the faint hearted!
28. Next seminars
@ Mobile Device Management and BYOD – The major players
@ Wednesday 25th June, London Office
@ The Magical Project Manager
@ Friday 4th July, Durham Office
@ Sign up online via www.waterstons.com
Editor's Notes
Specialist temperature controlled distribution and storage company. Provide their traffic office with a map of where their trailers are at any point in time across the UK
(Usage model: Query then analyse - tactical)
e.g. Picking sheets for warehouse operatives
Typically produced of a transactional system and a data warehouse is not used; different architecture to a BI system
One thinks of paginated reports with lists of tabulated data
Up to the minute detail essential to the running of the organisation
Typically not interactive, but it can be: (temperature controlled logistics; where is a vehicle now on a map; will it be there on time?)
Operational dashboards:
- particularly common in environments where it is essential to act on opportunities and issues quickly: e.g. help-desk, supply-chain
Continuous view of what is happening in a business unit; maybe not long-term trends that have built up over months
More direct connection to source systems for low-latency
Usually a single data source; not consolidation of multiple data sources
Food manufacturing company
Need to quickly respond to demands of supermarkets who run promotions on certain products
Plan for seasonal events (Easter, barbeque weather, world cup – things that affect buying patterns)
Alterations to pricing, volume and launching in new markets
Sales cannibalisation
Needs much greater engagement and input from users
HE Client
- Looking at Bayesian algorithms against student engagement behaviour to identify students that are likely to struggle / drop-out