The core purpose of “Analytics connect “is to give a brief overview of Analytics. Data is the prerequisite for any analytical activity, and it has to be clean and accurate. The more accurate is the information, the better insight would be. To get the competitive edge out of the analytical recommendation, all the contributing factors responsible has to be thoroughly studied and analyzed. Any “miss” has direct impact on the quality of recommendation. Analytics connect Analytics provides the business with the cushion time to meet the actual scenario with optimum service level. It’s not only improves your cash flow, rather it also provide signal well in advance so that business take appropriate measure in terms of aligning their strategy with their target, and making best decision upon their working capital. “In God we trust, all others must bring data” - W Edwards Deming “There is a striking correlation between an organization’s analytics sophistication an competitive performance” “The biggest obstacle to adopting analytics is lack of knowhow about using it to improve the business”. - 10 insights: A first look at the new intelligent enterprise survey on winning with data, MIT Sloan Management review Vol. 52, No1, 2010 What is Analytics and BI “Science of Analysis.” “Analytics is the science of examining raw data with the purpose of drawing conclusions from the information. Analytics is used in many industries and functions to allow companies and organization to make better business decisions and in the sciences to verify or disprove existing models or theories.” “The purpose of analytics is to take existing data collected from either a single source or multiple sources and use it to arrive at the optimal decision. Essentially, analytics can be best defined as a science of analysis.” “Business intelligence is the ability of organization generate knowledge” Analytics makes the difference (Company - ABC Inc.) Business Intelligence: Customer retention (75%) Business Analytics: Factors that influence customer retention (Why customer retention is only 75% and how to increase it to say 85%). Analytics & Companies stake Gartner Says Worldwide Business Intelligence, Analytics and Performance Management Software Market Surpassed the $10 Billion Mark in 2010. Business analytics market grew at 11.2%, and currently at $19.3 billion. A recent study by Gartner predicted advanced analytics as one of the top 10 strategic technologies for businesses in 2011. IBM has invested over 15 billion dollars in the last 4 years to acquire more than two dozen companies in the field of business analytics. Wipro is acquiring Promax Applications Group (PAG), an Australia based analytics company, for 35 million Australian dollars. Oracle bought Hyperion for $3.3 billion; SAP acquired Business Objects for $6.8 billion; and Microsoft purchased Fast Search and Transfer for $1.2 billion. There have been several other smaller deals too. Even Twitter admitted recently that analytics has become “an increasingly crucial part of improving our service” and acquired Smallthought Systems, the company behind an analytics service called Trendly that tracks real-time changes and user trends for website owners. Why analytics? Though analytics is not a new field, it has shot up in popularity in the last decade or so. Businesses have woken up to the power of analytics and more and more enterprises are moving to what is being termed as ‘fact-based decisioning’. There are three main reasons behind this phenomenal shift in businesses towards analytics and reliance on data over intuition. 1. There is a lot more data now than there used to be. 1.2 trillion gigabytes of data was produced in 2010 and it will double in the next 18 months and double again in the eighteen months after, and so on. Wal-Mart, a retail giant, handles more than 10 million customer transactions every day. The data generated by these transactions feeds into Wal-mart’s gigantic databases that hold over 2.5 million gigabytes of information already. Google processes million gigabytes of information every hour. 2. Lot more data is being collected and stored. Storage costs have gone down at an astonishing pace in the last 2 decades. In 1990, the cost of storing 1GB of data was around $9000. In 2010, this has come down to a mere 8 cents. Costs have gone down by more than 100,000 times in 20 years. This is what has really fuelled the growth in analytics, allowing businesses to store gigantic amounts of information at a reasonable cost. See this website that has captured the fall of storage costs since 1956. 3. Analytical tools have become more user-friendly and better at dealing with large data. With faster processing power and more sophisticated tools, users are able to sift through larger mounds of data and apply complex statistical techniques to draw insights enabling them to make better decisions on the fly. More data is being produced. More data is being stored. More data is being analyzed by businesses using better tools and statistical techniques. All of these factors are leading to better predictions being made. Better predictions mean better results and hence the growing popularity of analytics in business. What can analytics do for businesses? Analytics is enabling a shift in how businesses support operational decisions: from intuition to data-driven. It is becoming possible to run simulations or create models to provide insights and predict the future, rather than to simply describe what happened in the past, and to do this in real-time to support each individual business action. Thus, analytics enables businesses to make better, more informed decisions and compete more successfully. Marriott International A successful revenue management system has helped Marriott to achieve strong financial performance. Marriott’s analytics tries to identify most profitable customer through its rewards program. Marriott earned $4 billion through its web analytics Capitol one Capitol One runs over three hundred experiments per business day to target individual customers Increased customer retention to 87% and lowered the cost of acquisition by 83%. The value of capitol one’s stock has increased by more than 1000% over 10 year period. Google Analytics Provides analytics as a service to everyone with a website. Google analytics helps advertisers to maximize their hit rate! Analytics in Sports A C Milan uses predictive models to prevent injuries to its players Oakland athletics use Analytics for Team selection (Won 4 super bowls between 2000 and 2008.) Customer Life Time Value is used by many baseballs teams in US to identify profitable customers. Three pillar of analytics approach Measurement frame work: How I am I doing, as is process. How is company performing? What are the current strategic goals and priorities and what metrics to measure success or failure towards those goals? Can these metrics be aligned with stated priorities, goals and strategies? Portfolio analysis: what are my business dynamics drivers of business? What are your business dynamics? How do your products perform in the marketplace? What is the driver of business, and how does it relate to company strategies and priorities? Customer analysis: Who are my customers, how do I engage them, how can I communicate my analysis to customer. What do they want? How do they behave? These are the crucial questions in your customer analysis. Basic questioning along these lines will create several categories for them. Demographics: What are your customer demographics? The relevant demographic metrics will depend on your specific product/service, i.e. gender may or may not matter, or all of your customers may even be of the same gender or age group. Needs: Determine your customers’ needs. Do they have a feature need incremental to your current product? Do the customers need installation or operating support? Behaviors: You can look at how your customers behave. Do they pick up your product at a warehouse or retail outlet? How many are convenience seekers always going for fastest shipping etc.? Are they mail catalog subscribers? There should be direct connection between the business and dashboard. Too many dash boards which basically doesn’t relate to the business objective is irrelevant, so understanding the specific business need is very important There are two important aspect of looking at the data from analytics point of view Metrics: the numbers that measures the characteristics of the past. Predictive: Way to estimate what will happen in the future. Metrics are classified into 2 types:1. Process metrics: A metric used to measure the characteristic of the methods, techniques & tools employed in developing implementing & maintaining the software system. 2. Product metrics: A metric used to measure the characteristic of the documentation & code characteristic. Metrics is an accurate no., since it is based on the actual data, where as the predictive has a certain percentage of the accuracy e.g. forecast accuracy of best in class industry is 78.7%. Analytical approach Hierarchy 1 Standard Report (What Happened? When did it happen?) 2 Ad HOC Reports (How Many? How Often? Where?) 3 Query Drill down (Where exactly is the problem? How do I find the answer?) 4 Alerts (when should I react? What actions are needed now?) 5 Statistical Analysis (Why is it happening? What opportunities am I missing? 6 Forecasting (what if this trend continues? How much is needed? When it will be needed?) 7 Predictive modeling (what will happen next? How will it affect my business?) 8 Optimization (how do we do things better? What is the best decision for complex problem?) Functional Domain of Analytics Supply Chain Marketing and Sales Media Web Finance and Accounting Human Resource Analytics Technique T E C H N I Q U E S Analytics Problems Analytics and Academia Business has recognized the power of analytics and it is time academia did the same. Today’s business managers need to learn how analytics can help them make better business decisions that can generate better business outcomes. They need to have an understanding of the statistical concepts that can help analyze and simplify the flood of data around them. They should be able to leverage analytical techniques like decision trees, regression analysis, clustering and association to improve business processes. Finally, it is important for them to have first-hand experience of some of the most widely used software like SAS, SPSS, COGNOS etc., and exposure to real-life business data. Analytics programs in India While India has become the global hub for analytics, there is an enormous shortage of programs offered by Indian universities to train students for this field. Some of the leading MBA colleges have started to offer courses in business analytics as part of the MBA curriculum. However, there are still very few specialized degree or certificate programs available in this field. IIM Calcutta and recently started at IIM-Bangalore a rare exception, offers a 1 year course in business analytics for professionals with at least 2 years of Work experience. In the US, North Carolina based Institute for Advanced Analytics also offers a 1 year masters course in advanced analytics. It is imperative that business education institutes recognize the emergence of business analytics as a powerful field that is changing the way businesses are run. It is also important to understand the crucial role India is playing as the global hub for analytic talent. Education institutes, especially in the field of business administration, need to tailor their curriculum in recognition of the growing importance of analytics and the shortage of skilled professionals in the field. Every business is unique in terms of the kind of challenges it faces, the information it possesses and the decisions it has to make. However, a manager who is aware of the scope of analytic applications for business is able to leverage the power of data to make better business decisions. Types of Qualification required in Analytics Analytics requires high qualification, with a research oriented educational or professional background. Master or PHD Operations Research Marketing sciences Masters Business management Mathematics, Statistics Econometrics, Finance Others Engineering, CPA, CFA Professional Hierarchy & Career Growth Manager (7+ Years) Consultant (5-7 Years) Senior Analyst( 2-4 Years) Analyst (0-2 Years) Masters Analytics is a high growth field and salaries in analytics have been sky rocketing in the last few years. There is a strong demand for “data-savvy” professionals in the industry today. As per a recent McKinsey research report, the US alone faces a shortage of 140000 to 190000 people with “deep analytical skills” as well as a million and half data-savvy managers and analysts who can deal with big data. India ranked 3rd in the study – in terms of number of people available with analytical skills – behind the US and China. Yet India has emerged as the global hub for analytics. This is because India has a steady supply of English-speaking analytical talent. Captive & Non-captive Analytics Company Captive centers are those that are dedicated to a particular company. For example, those working for HSBC analytics or Dell analytics will work exclusively on the HSBC or Dell business. On the other hand, an analyst working in an analytics service company such as Manhattan, Symphony, Fractal Analytics, Genpact or Marketelligent is likely to work across multiple businesses and clients. The latter work tends to be looked upon as more rewarding since it is more varied in scope. To make up for this, captive centers usually offer slightly higher salaries than non-captives. Human Resource Analytics Entire activity of HR Analytics is governed by the workforce. It offers an accurate estimate of the number of employees required with matching skill requirements to meet organizational objectives. It is a forward looking function as human resource estimates are made well in advance. Analytics in Human resource revolve around different stages of maturity. Maturity 1. Encourage sufficient turnover to minimize layoffs and provide new openings. 2. Control compensation costs. 3. Maintain flexibility and skills of an aging workforce. Growth 1. Plan management succession. 2. Establish formal compensation structures. 3. Maintain employee motivation, and morale. Decline 1. Plan and implement workforce reductions and reallocations Implement 2. Tighter cost control. 3. Implement retraining and career consulting services. Introduction 1. Attract best technical and professional talent. 2. Meet or exceed Attract best labor market rates to attract needed talent. technical and 3. Define future skill requirements and begin professional establishing career ladders. talent. HR Maturity Model “Workforce Planning is the process that provides strategic direction to talent management activities to ensure an organization has the right people in the right place at the right time and at the right price to execute its business strategy“ Why HR Analytics/ Work Force Analytics? External Challenges: Liberalization, privatization and globalization (LPG era) have created huge demand for people in software, finance marketing, and manufacturing fields. Organizational Decisions: Decisions such as expansion, diversification, and relocation leading to demand for people possessing requisite skills Workforce Factors: Such as retirement, resignation, and termination etc creating manpower gaps. Basic Workforce Supply Model Sources of Inflows The Firm Transfers Current Staffing Level Promotions New Recruits Recalls Projected Outflows Employees Out Employees In Promotions Quits Terminations Retirements Deaths Current staffing level – Projected ou tflows this year + Projected inf lows this year = Layoffs Firm’s internal supply for this time next year Work Force Planning Process flow Predictive Analytics as an emerging trend Workforce Predictive Analysis Workforce predictive analysis Step 1: Scan the External Job Market Start developing a workforce strategy by looking outside of your company at what’s going on in work and employment in your community. There are key pieces of workforce data you can gather: External Factors Affecting Labor Supply Unemployment rate Graduation rates Demographic makeup of local market Knowledge of direct competitors When reviewing these areas of the job market, you are looking for opportunities and threats. Is there an especially large number of college grads coming into the job market in your geographic area who can fill entry level office positions cheaply? Or, more specific to our current economic situation, is the job market saturated with highly-skilled employees looking for work who you could be snapping up before the economy improves? Workforce predictive analysis Step 2: Take an Internal Scan Once you’ve taken a look at the workforce available to you, now is your chance to analyze the jobs that need to be done in your organization and the skills of your current employees available to do them. It’s important to take a look at the whole business and get a sense for all of your human resources needs. Here’s a way to organize your research. Start with: Jobs and Skills Required by the Organization Jobs that exist now Number of employees doing each job Importance of each job Characteristics of anticipated jobs Once you’ve summarized your company’s human resource needs, take a look at the actual employees you have available to you to fill those needs. Review your: Organizational Capabilities Detailed information about current employees and their capabilities (knowledge, skills and abilities) Information about special expertise, mobility restrictions and specialized job qualifications Workforce predictive analysis Step 3: Predicting Workforce Demand Now that you know what you have available to you both inside and outside of your organization, it’s time to take a guess at what human resources you’ll need down the line. And, before you head into this next step, keep in mind that workforce planning and forecasting is an art, not a science. Your best approach is to be as informed as possible and include others in the workforce planning process. You will start by gathering some basic supply and demand information. First, gather information about the possible demand for replacement of positions. Possible Causes of Employee Replacement Retirement trends Turnover rate Internal fill rates Once you’ve gained a sense of which employees might be moving on or up in your organization over the coming years, you should then gather information about the possible demand for new or evolved positions. To research potential causes of new workforce demands in your company, go straight to the top. Meet with your leaders and let them know that your goal is to understand the business and understand the business’s objectives. This information will help you know what sort of new positions may be needed in the coming years. Workforce predictive analysis Step 4: Predicting Workforce Supply At this point, you have a good idea of what resources are available to you and what needs may come up. Now you can drill down to the details of a plan for hiring for the coming years. Remember that workforce forecasting should be done for the short term, intermediate and long term. Here are the practical steps: Set a Realistic Timetable That Includes Fill time for open positions. Economic conditions and availability of labor. Needed ramp time. Take an Assessment of Internal Workforce Supply Who’s in line to step up? What are the training needs? Are there succession plans in place? Adjusting to Unexpected Changes in Workforce Supply and Demand Once you have completed the steps above for forecasting your workforce supply and demand, you cannot sit back and just hope that everything goes smoothly in the roll out of the plan. You next need to know how to deal with unexpected changes to your workforce plans and how to keep consistent, company-wide interest in following through with this HR plan. In our future post, I will discuss dealing effectively with HR surplus, how to maintain support for your HR workforce plan and give you a list of things to keep in mind so that you achieve success with forecasting workforce supply and demand no matter what changes come. Classification of Predictive analysis: Trend Analysis – a quantitative approach to forecast future personnel needs based on extrapolating information from historical changes. Delphi Technique –forecasts and judgments of a group of experts are solicited and summarized to determine the future of employment. Impact Analysis –trends are analyzed by a panel of experts who then predict the probability of future events. Scenario Planning –creating future scenarios that differ radically from those created by extrapolation of present trends Commonly used workforce forecasting methods can generally be classified into two categories based on the utilization of past data. Methods that do not rely on past data include employer’s survey international comparisons and labor market signal analysis. Work Force planning data analysis methods Work Force Surplus/Deficit Internal workforce supply exceeds the organization’s requirement or demand for personnel Demand exceeds the current resources available in the organization's workforce Determine NET HR Requirements External supply requirements = replacement + change supply components Change supply = hiring to increase (or decrease) the overall staffing level Replacement supply = hiring to replace all normal losses External supply = current workforce # x (replacement % per year + change % per year Example: Surplus = 1000 x (4 % per year (0.04) + 2 % per year (0.02) 40 + 20 = 60 Need for 60 employees Deficit = 1000 x (4% per year (0.04) + -7% per year (-0.03) 40 + -70 = -30 Need to reduce by 30 employees Trend Analysis Historical relationship between a business index (e.g., sales, contracts, units sold etc.) and the number of employees required to achieve that index (labour demand) Main idea of the method: a forecast is calculated by inserting a time value into the regression equation. The regression equation is determined from the time-serieas data using the “least squares method” The general equation for a trend line : F=a+bt Where: • F = forecast, t –= time value, a = y intercept, b = slope of the line. Regression Analysis Y = dependent variable (HR demand) A = constant (Y intercept) B = slope of linear relationship between X and Y X = independent variable (e.g. level of sales) Regression Prediction Y = A + BX Y = dependent variable (HR demand) A = constant (Y intercept) B = slope of linear relationship between X and Y X = independent variable (e.g. level of sales) Example data: X Sales ($ Millions) 2.0 3.5 4.5 6.0 7.0 Y # of Marketing Personnel 20 32 42 55 66 Calculate XY, X2, average X and average Y X Y XY Sales ($ Millions) # of Employees X2 2 25 50 4 2.5 28 70 6.25 3.5 30 105 12.25 5 6.5 38 54 190 25 351 42.25 766 89.75 19.5 175 Average X = 19.5/5 = 3.9 Average Y = 175/5 = 35 N = 5 𝑋𝑌 − 𝑁 𝑋 (𝑌) 𝐵= (𝑋2 ) − 𝑁(𝑋)2 Calculate the value of B (slope of the linear relationship between X and Y) XY X2 50 4 70 6.25 105 12.25 190 25 351 42.25 766 89.75 N = 5, Average X = 3.9, Average Y = 35 𝐵= 766−5 3.9 35 766−683 89.75−5 3.9 89.75−76.05 , 𝐵= 2 , 𝐵 = 6.09 Calculate A (constant or intercept) Average X = 3.9, Average Y = 35 B = 6.09 IF Y = A + BX , Then A=Y ̅- BX ̅, AND A = 35 – (6.09) (3.9) A = 11.23 Determine the regression prediction equation Y = A + BX, A = 11.23, B = 6.09, Y = 11.23 + (6.09) (X) Calculate predicted HR demand (Y) by inserting values for X Predict the HR demand for personnel at $8 million and $10 million. Y = A + BX A = 11.23 B = 6.09 X = Sales ($M) Y= 11.23 + (6.09) (X) $8 million $10 million Y = A + BX Y = A + BX A = 11.23 B = 6.09 X=8 A = 11.23 B = 6.09 X = 10 Y = 11.23 + (6.09)(8) Y = 11.23 + (6.09) (10) Y = 59.99 60 Staff required Y = 72.18 72/73 Staff required Steps for Delphi Technique • Define and refine the issue or question • Identify the experts, terms, and time horizon • Orient the experts • Issue the first-round questionnaire • Issue the first-round summary and second-round questionnaire • Continue issuing questionnaire Steps for the Nominal Group Technique • Define and Refine the Issue or Question and the Relevant Time Horizon • Select the Experts • Issue the HR Demand Statement to the Experts • Apply Expert Knowledge, State Assumptions, and Prepare an Estimate • Meet Face-to-Face • Discuss the Demand Estimates and Assumptions • Vote Secretly to Determine the Expert Demand Assessment Scenario Planning Provides a framework to consider what the future holds • Identifies new risks and challenges in the future - helps to prepare for the unexpected • Enables you to expand your vision and think of alternative futures • Most useful when there is limited clarity on the future • Helps to balance intuition, judgment and fact • Recognizes that forecasting is notoriously difficult Where has it been used? Private sector companies, eg, Shell, Body Shop, • NHS (National Health Service) How do you do it? • There is no set approach - depends on organization, situation etc… One approach to Scenario Planning • Identify issues and forces that will shape the service and/or care pathway over a defined period of time • Identify those that will have most impact • Identify those that give the most uncertainty • Identify 2 issues that cause most worry from this list and map against each other, taking a positive and negative view of each one • Consider if there may be any wild cards coming into play • Then consider what it will look like in each of the scenarios you have created – Identify what this will mean for the service design – Identify what this will mean for the workforce – Identify what this will mean for service users – Identify what this mean for partners along the delivery chain • Are any common issues identified? • What are the risk factors that you need to mitigate against? MARKOV ANALYSIS-1 • Produces a series of matrices that detail the various patterns of movement to and from the various jobs in the organization. • Determines the likelihood that an individual will display movement behaviors. TO: A TRANSITION MATRIX FROM: TOP MID TOP .80 .02 MID .10 .76 .04 .06 .78 .01 .15 .01 .84 .15 LOW LOW ASSY EXIT .18 SKILL ASSY SKILLED .10 .05 .88 .07 ------------------------------------------ MARKOV ANALYSIS – 2 (Captures effects of internal transfers) (Start = 3500) FROM/ TO: TOP 100 MID 200 LOW 600 SKILL 600 TOP .80 .10 A TRANSITION MATRIX MID LOW SKILLED .02 ASSY EXIT .18 .76 .04 .10 .06 .78 .01 .15 .01 .84 .15 ASSY 2000 .05 .88 .07 --------------------------------------------------------END YR WITH: NEED RECRUITS ? NEED LAYOFFS ? 100 0 KEEP STABLE 100 190 482 10 118 610 240* (10)* 200 600 1760 600 [358 left] 368 tot (10) tot 2000 = 3500 Tot MARKOV ANALYSIS – 3 (Anticipates Changes in Employment Levels) Employment needs are changing. We need a 10% increase in skilled workers (660), and a 15% decrease in assembly workers (1700) by year’s end. ------------------------------------------------------(Start = 3500) A TRANSITION MATRIX FROM/ TO: TOP MID LOW SKILLED ASSY EXIT TOP 100 .80 .02 .18 MID 200 .10 .76 .04 .10 LOW 600 .06 .78 .01 .15 SKILL 600 .01 .84 .15 ASSY 2000 .05 .88 .07 --------------------------------------------------------END YR WITH: 100 NEED RECRUITS ? NEED LAYOFFS ? 0 NEW LEVELS 100 190 482 610 10 118 50* 1760 [358 left] (60)* 200 600 600 1700 = 3260 tot TARGET STORES STAFFING FORECAST MODEL Y = 8 + .0011(X1) + .00004(X2) + .02(X3) Y = Number of employees needed to staff the store X1 = Square feet of sales space X2 = Population of metropolitan area X3 = Projected annual disposable income in millions of dollars Y = 8 + .0011(50,000sq ft) + .00004(150,000popul) + .00000002($850 million) Y = 8 + 55 + 6 + 17 Y = 86 employees needed at this store VACANCY ANALYSIS Historic departures used to project turnover LEVEL # EMPL TURN % Expected Vacancies TOP MGMT 100 20 % 20 80 MID MGMT 200 24 % 48 152 LOW MGMT 600 22 % 132 468 SKILLED W 600 16% 96 504 ASSY WKRS 2000 12 % 240 1760 TOTALS 3500 536 2964 AVERAGE TURNOVER PERCENTAGE = 536 / 3500 = Expected to Remain .1531 References http://www.gartner.com/hc/images/215650_0001.gif IIM Bangalore programme “Business Analytics and Intelligence Course”: Prof. Dinesh Kumar http://www.analytics-magazine.org/ http://sloanreview.mit.edu/the-magazine/2010-fall/52115/10-insights-a-firstlook-at-the-new-intelligent-enterprise-survey/ www.senga.ca http://www.payscale.com/hr/resources/hr-planningwebinars01?src=blog Email: ramesh@samjna.com