Business Analytics for Managers - Taking Business Intelligence Beyond reporting

1. All the way from strategy to data sources, this book will give you guidance on how to work with data warehouse information. • Read more at www.ba-support.com Business Analytics for Managers- Taking Business Intelligence Beyond reporting

2. Introduction Purpose of the PowerPoint… • These slides can be used for teaching: ”Business Analytics for Managers: Taking Business Intelligence Beyond Reporting” • We don’t imagine that the slides are covering everything, but they are selected based on the criteria below: • These slides contain the essential models of the chapters • In some cases exercises are suggested and examples included

3. Introduction Purpose of the book… • A guide to fuel what we refer to as the analytical age • The ability to make an information strategy • An understanding of Business Analytics (BA) as a holistic information discipline • An understanding of the ever increasing role of BA • A reference to most used BI concepts, definitions and terms

4. Chapter Introduction Content • What is business Intelligence? • Defining the term • The difference between BI and analytics • Defining more terms • What is an information system? • Business Intelligence is much more than a technical disciplin

5. Introduction What is Business Intelligence • Two popular definitions are: ”Decision support” and ”The process of providing the right people the right information at the right time” • Which customers should we send reminders (Credit department) • What advertising should we send to which customers (CRM) • What type of employees have the highest absence (HR) • Information about which products and customer segments are most profitable and therefore deserves focus in the future (Marketing) • This book definition of BI is a combination of the two definitions: ”Delivering the right decision support to the right people at the right time”

6. Introduction What is Business Intelligence • This book focuses on information from data warehouses, however the same decision could also be based on e.g. survey data, interviews with subject matter experts or external consultants. • In general however, it is not the source that is important, it is too which degree it enables the right persons to take the right decisions. • Good decisions are made; if the decision maker uses all the relevant input and analyzes this correctly at the time of the decision. • Business intelligence is typical humans taking actions based on IT system input. • IT automation is typical when IT systems take actions based on IT system input. • Another definition could be: decision support used for business decisions.

7. Introduction The difference between BI and Analytics Types of information (BA vs. BI)… • This book typically takes it focus on processes – that is the routines we do and how they interlink. • Processes typically evolves in steps over time • Horizontally when we manage it • Vertically when we improve it • The blue area represents a stateless than optimal • Can also be seen as the informationwe need before we run a new process vs. the information we useduring the process. Later in the bookwe will look at learning information,which is information we use after wehave learned from a process. E.g. amarketing campaign. Process Performance

8. Introduction What is an information system? • BI-department produce information, but that in itself is not value creating. Value is not created until people throughout the organization improve upon their decision making and business processes • BI should be seen as a part of an information system • Business processes • Technical solutions • Human competencies

9. Introduction What is an information system? • A value creating information system is characterized by three elements: • Some specific business processes are optimized by improved decision making – compared to a situation without an information system People must act differently and more efficient than they used to, before value is created by using information. You can say that people must improve upon their work processes, that means the way they act in daily work routines • The information system contains a technological element that collect, store and deliver information It can be IT-based, but also paper, papyrus scrolls, yellow sticky notes or heads with good memories • Human competencies form part of the information system to Someone must be able to retrieve data and deliver it as information in, say a front-end system. Even more important, those who make the decisions, those who potentially should change their behavior based on the decision support, are people who must be able to grasp the decision support handed to them

10. Introduction What is an information system? • When you create a new information system the order is the same: • Identify how you want the business process to work – who should do what and when, in which order etc. • Then create an information system that can deliver the right decision support to the right people at the right time • Train users in how to work the new process, including how to retrieve and use the technical decision support system

11. Chapter 1 Content • The BIA model • Presenting the central concept in this book • An example • Make it more concrete

12. Chapter 1 The BIA model Structure of the book • The BIA-model – from strategyto data sources

13. Chapter 1 The BIA model Structure of the book • The models five phases each of which is allocated a separate chapter • Ch. 2: The relationship between business strategies. The overall competitive position and strategy of the company set requirements for the information to be delivered and used • Ch. 3: Based on the overall company strategy, requirements for information deliverables are specified at the functional level , so they are able to reach their objectives. Business processes are improved in this area. • Ch. 4: How data, information and knowledge is created by analysts. • Ch. 5: How information is stored over time in a data warehouse • Ch. 6: What sources systems typically deliver data to a data warehouse

14. Chapter 1 An example

15. Chapter 2 Content • Strategy and information • Strategy is about solving short term issues and gaining long term competencies. Competing in the information age means that you in the strategy creation process also must be aware of how information can help you. • The relationship between strategy and business intelligence • Varying degrees of integrating between company strategy and the usage of information • Which information do we prioritize?

16. Chapter 2 Strategy and information • Postulate: ”A company’s core competencies lie in the field of knowing how to handle internal processes and knowing what customers want now and in the future” • These competencies include things the company is especially good at, and which can secure its survival, also an organization should be capable of evolving these competencies to meet the future requirements in the marketplace • The point is, that companies competes on continuous creation of knowledge and the ability to execute based on it

17. Chapter 2 The relationship between strategy and business intelligence • Separated: In these companies, data are not used for decision making on a strategic level. The BI-function only works on ad hoc basis. No prioritization in regard to what is relevant for the strategy. • Coordinated: The BI function performs monitoring of individual functions’ achievement of by using Dash boards etc. BI is use reactively to monitor the execution of strategy. • Dialogue: A learning loop is facilitated when the BI function is reporting on business targets. Analyses create learning from the differences between targets and actual. BI is used proactively to improve operational processes when executingthe company strategy. • Holistic: Information is being treated as a strategic asset, which can be used to determine company strategy. The organization constantly evaluates howit can compete on information when strategies are created.

18. Chapter 2 The relationship between strategy and business intelligence Coordinated • The BI function performs monitoring of individual functions’ achievement by using Dash boards etc. • Here the BI function is a reactive element, solely employed in connection with the monitoring of whether the defined targets of the strategy are achieved (performance management)

19. Chapter 2 The relationship between strategy and business intelligence Dialogue • A learning loop is facilitated when the BI function is reporting on business targets. • Analyzing the differences between targets and actual create learning • BI is used proactively to improve upon operational processes when executing the company strategy.

20. Chapter 2 The relationship between strategy and business intelligence Holistic • Information is being treated as a strategic asset, which can be used to determine company strategy. How information, in combination with strategies, can give them a competitive advantage • Competing on Analytics (Davenport) describes how a company can use information as a strategic asset/resource

21. Chapter 2 Which information do we prioritize? • The strategy part of the book is primarily inspired by Treacy & Wiersema’s article: Treacy, M. & Wiersema, F. (1993) Customer Intimacy and other Value Disciplines, Harvard Business Review, Jan/Feb • The article explains that companies compete on the dimensions: • Process Excellence (efficient in relation to production and delivery services, and which always focuses on optimizing internal processes) • Customer Intimacy (strong customer relations, that is, about being able to establish a psychological connection to customers) • Product innovation (strong in the field of product innovation and being a leading supplier of ’state of the art’ products) • If a company masters one of the three disciplines and matches its competition on the two others it can become a market leader

22. Chapter 2 Which information do we prioritize? • Discuss which information for the following companies would be expected to focus on when making a company strategy: • Dell – low cost web distribution of products • Apple – innovative and customer oriented • Siemens – world leading wind mills • Hennes & Mauritz, Boss and Wal-Mart clothing stores.

23. Chapter 3 Content • The relationship between this chapter and previous chapters • Keeping the BIA-model in mind • Establishing new business processes • If we start up a process from nothing, we have no data on it • Optimizing existing business processes • If optimizes and existing process, we have data to work with also • Which processes should we start with? • Depending on the overall company strategy, what internal information systems must be expected a natural focus for the company

24. Chapter 3 The relationship between this chapter and previous chapter • The previous chapter focused on how information is used at the overall strategic level • At the functional level, we identify how to get from having some overall objectives for a department / function to being able to specify the information requirements. • The example contains 3 function / business processes

25. Chapter 3 Establishing new business processes • The Rockart model explains, how to identify information related to strategy and objectives, this is called developing an information strategy • The department/function is given some objectives. To reach these objectives an operational strategy is developed. The critical success factors are the elements of the plan that must have a successful outcome, if the plan as a whole is to succeed. Thisalso defines what the Business Intelligencedepartment has to deliver Lead and lag information • Lead information is information or knowledge that is necessary for even beginning our new business activities. • Lag information let us monitor if we reach the strategic target

26. Chapter 3 Establishing new business processes An example, page 64.

27. Chapter 3 Establishing new business processes Exercise • Create 3 information wheels with lead and lag information for Human Resources (HR) in a IT consulting company • Assumption: overall strategy at company level: we want to increase the market share from 15% to 25 % • Assumption: objective for HR: Hire 20 new good sales managers • Your task as a HR manager  • Make your own local / operational strategy • Identify 3 critical success factors • Identify lead and lag information for each critical success factor • Develop and present your information wheels

28. Chapter 3 Optimizing existing business processes • The model focuses more on the lag information by collecting and analyzing it to understand correlations to be able to improve processes in the future. • The model uses lag information to create lead information (maybe learning loops)

29. Chapter 3 Optimizing existing business processes Exercise • Optimizing two existing business processes of your own choice (perhaps you local canteen) • Identify lag information to create lead information • Create learning loops • Present your information wheels

30. Chapter 3 Optimizing existing business processes An example, page 71.

31. Chapter 3 Which processes should we start with? • Correlation between strategy and operational processes with significant analytical potential lead us to the fact that some analytical disciplines are more relevant for some businesses than for others. • Exercise • Which analytical disciplines do you think the companies below should focus on? • Dell – low cost web distribution of products • Hennes & Mauritz, Boss and Wal-Mart clothing stores. • Apple – innovative and customer oriented • Siemens – world leader in wind mills

32. Chapter 4 Content • Data, information and knowledge • Analysts role in the BI model • Requirements the analyst must meet • How to select the analytical method (The three questions) • Reporting • Statistical testing • Data mining • This chapters focus in on what links the technical part of BI and BA together with the commercial part of the organization. The focal point here is therefore the analyst and his or hers toolbox

33. Chapter 4 Data, information and knowledge Exercise • Discuss the difference between data, information and knowledge – Are these concepts the same thing?

34. Chapter 4 Data, information and knowledge The definitions in the book • Data is defined as the carrier of information. An example of a piece of data could be ‘‘bread’’ or ‘‘10.95.’’ Data is often too specific to be useful to us as decision support. Data rests in data warehouses and typically describes a transaction, action, status etc. • Information is data that is aggregated to a level where it makes sense for decision support in the shape of, for instance, reports, KPIs, alerts, tables, or lists. An example of information could be that the sales of bread in the last three months have been respectively $18,000, $23,000, and $19,000. • Knowledge is information that has been analyzed and/or interpreted. This means that the BI department, as an example, offers some suggestions regarding why bread sales have fluctuated in the last three months. Reasons could be seasonal fluctuations, campaigns, new distributions conditions, or competitors’ initiatives. It is therefore not a question of handing the user a table, but instead of supplementing this table with a report or a presentation. This means, of course, that when the BI department delivers knowledge, it is not a result of an automated process, as in connection with report generation, but rather a process that requires analysts with quantitative methods and business insight.

35. Chapter 4 Data, information and knowledge Exercise • Give some examples of how data can be transformed to information and from information to knowledge. • You can reuse the Canteen case if you want to.

36. Chapter 4 Requirements the analyst must meet The core skills of an business analyst • Business competencies • First of all, the analyst must understand the business process he or she is supporting and how the delivered information or the delivered knowledge can make a value-adding difference at a strategic level • Tool kit must be in order (method competencies) • Reporting • Statistical tests (to show any correlations that might be present in the tables.) • Data mining (spot pattern or correlation in data) • Technical understanding (data competencies) • If, for instance, an analyst needs new data in connection with a task, it’s no good if he or she needs several days to figure out how the Structured Query Language (SQL) works, what the different categories mean, or whether value-added tax is included in the figures. • Analysts spend about 80% of their time retrieving and presenting data, so we also have to place some clear demands on the analysts’ competencies in connection with data processing.

37. Chapter 4 How to select the analytical method Three question model • The aim is to present a model that can be used in the dialogue between management who wants information and the analyst who must deliver it: • Question 1: Determine with the process owner whether the quantitative analytical competencies, or the data manager and report developer competencies are required. • Question 2: Determine whether hypothesis-driven analytics, or data driven analytics can be expected to render the best decision support. • Question 3: Determine whether the data-driven method has the objective of examining the correlation between one given dependent variable and a large number of other variables, or whether the objective is to identify different kinds of structures in data.

38. Chapter 4 Reporting Three question model • Reporting (by using descriptive statistics) presents information and the individual viewer or business user will be the person to interpret and transfer this information into knowledge. • Reporting is typically based on a part of the available data and is cross tabulated (e.g. sales per week per region)

39. Chapter 4 Reporting Types of reporting • Ad hoc reports (one time only) • Manually updated reports (normally used in connection with projects and therefore have a limited lifetime) • Automated Reports: On Demand (standard reports) • Automated Reports: Event Driven (relevant information presents itself to the individual user at the right time) • Reports should be internally aligned (one version of the truth)

40. Chapter 4 Statistics • When working with hypothesis-driven methods, we use statistical tests to examine the relationship between some variables in, let’s say, gender and age. We must have some initial idea of the relationship • The result of the test will be a number between 0 and 1, describing the risks of being wrong, if we conclude based on the data material that there is a relationship between gender and lifetime. The rule is then that if the value we find is under 0.05, that is, 5%, then the likelihood of our being wrong is so insignificant that we will conclude that there is a relationship. • Since the 5% also means that if we make 20 random test between variables which has nothing to do with each other, will in average 1 time (in 5% of the cases) find a significant relationship which is a false true. For the same reason you must before you test make a sanity check in regard to whether it theoretical possible that there is a relationship between the two variables.

41. Chapter 4 Data mining • Where statistics is a hypothesis driven process with the aim of creating knowledge. • Data mining is more of a data driven process with the aim of finding actionable patterns in the data. • This means that we scan the data for patterns and we evaluate the quality of the decision support the process generates via testing the model on an unknown data set. In statistics we evaluate the quality of the decision support via monitor the level of significance and screen for whether it is a relevant test in the first place. • Data mining is typically used to give decision support on questions like: • Which of our engines will break down when and why? • Which customers will leave us when and why? • Which customers will buy what and when? • Which customers have high credit risk? • What is the price of our products next year?

42. Chapter 4 Exploratory techniques Four typical examples • In BI, we typically see four types of explorative analyses. These are methods for data reduction, cluster analysis, cross-sell models, and up-sell models. • Data reduction - We take all the information in a large number of variables and condense it into a smaller number of variables. • The cluster analysis also simplifies data structures by reducing a large number of rows of individual customers to a smaller number of segments. For this exact reason, the two methods are often used in combination with questionnaires, where data reduction identifies the few dimensions that are of great significance, and the cluster analysis then divides the respondents into homogenous groups or segments. • Cross-sell models are also known as basket analysis models. These models will show which products people typically buy together • Up-sell models are not looking at what’s in the shopping basket once; instead, they are looking at the contents of the shopping basket over time

43. Chapter 4 Data mining A typical data mining process • Step 1: Development of different models • Step 2: Selecting the best model based on criteria such as : • The model can be interpreted • The model’s prediction power on an ”unknown” dataset • Step 3: Score a data set – Implement model (production)

44. Chapter 4 Data mining Example • Question to be answered: “Is there a correlation between the inquiries from corporate customers and whether they canceled their subscriptions shortly after?” • Based on the model, an automated electronic service was generated that scanned the data warehouse of the call center every five minutes. If any ‘‘critical’’ calls were found in the log readings from conversations with customers—which could be that they had called in for a good deal combined with the fact that their contract was up—then the person who was responsible for this customer would automatically receive an email. With reference to the reporting section of this chapter, this is essentially an event driven report which is generated based on a data mining algorithm.

45. Chapter 4 Data mining About data mining projects • A data mining project often takes several weeks to carry out, partly because we are often talking about large volumes of data (both rows and columns) as input for the models. • The modelling fase of the project only takes up a fraction of the total time it takes to establish a new data mining process • After the data mining process is established and automated, it can be performed in a matter of hours next time.

46. Chapter 4 Choose appropriate analytical method for information wheels Exercise • HR wants' to optimize their future hiring process by using analytics. As analyst you shall help them choose appropriate analytical methodologies for their information wheels • HR wants to compare the new sales managers performance with the ’old’ sales managers performance – What’s the method? • HR wants to identify a profile of a successful sales manager – What’s the method? • HR wants to identify if there is a correlation between salary and sales results – What’s the method? • HR wants to identify a good 10 year carrier path for the new sales managers – What’s the method? • HR wants to know the average sales manager’s historical sales result – What’s the method? • Which of the above questions/answers are lead information and which are lag information? • Which of the above questions/answers belong to the information domain and which belong to the knowledge domain?

47. Chapter 5 Content • Why a Data Warehouse? • Architecture and processes in a Data Warehouse • How should you access your data? • BI Portal: Functions and examples • Exercise in Data Warehouse • This chapter will look into what a data warehouse or data repository is, and how it extract, transforms and loads data to other systems and front ends

48. Chapter 5 Why a Data Warehouse? • To increases the usabilityand availability of source data • To ensure consistency and valid data definitions across business areas and countries (this principle is called one version of the truth). • To hold documentation of metadata centrally upon collection of data • To avoid information islands • To create a historical data foundation • To avoid overloading of source systems with daily reporting and analysis

49. Chapter 5 Architecture and processes in a data warehouse Data floats from source systems to the BI Portal

50. Chapter 5 Howshouldyouaccessyour data? How does users get data out of a data warehouse.