Saturday, May 18, 2013

How to Embed Analytics into Organizational Culture

As information is life blood in modern organization, thereof, how to build up a digital business with culture of analytics is strategy shift to compete for the future. A truly analytical organization not only measures the right things at the right time to facilitate quality decision making but also measures for the sake of improving the business productivity and optimizing customer services.

1. The Roadblocks of Doing Analytics

Analytics is the right way to get ahead, though the road to creating an analytic culture in a company can often be a long and arduous journey and hit a few roadblocks.

    • The biggest bottleneck is adoption. Many organizations and executives are ill-prepared or reticent to act on information, too many organizations target analytic output at "interesting" questions rather than truly "important" crunchy questions that matter to the business. Interesting questions tend to be hindsight ("what happened?") oriented, while crunchy questions tend more toward insights ("Why did it happen?") and foresight ("What will happen if...?"). 
    • "Technical complexity": The problem with traditional advanced analytics solutions is that they are not accessible by decision makers. The highly technical nature of traditional solutions, time delay due to SDLC of custom developments/ customizations/ enhancements, the performance limiting amount of data needed for accuracy, inability to share results to stakeholders etc. causes the decision makers not to have the answers that they want and more importantly when they want it. 
    • Run into "multiple versions of the truth" issue. Lots of data and reports floating around companies and in many cases, each developed to support a specific business unit or individual agenda and not necessarily linked to what drives or could improve a business. For example, two different business units will come up with two very different conclusions based on the same data, depending on how it is analyzed and presented.  
    • The other major problem is that analytics often reveals inconvenient or unpalatable "truths" to the business side of the organization. In an open, progressive culture that is welcomed. However, most cultures are not, so analytics is often used as the proverbial lamppost to support the corporate status quo, rather than being a searchlight to illuminate the way forward. Analytics types tend, by nature, to be quiet, observant types, and a-political, so the function is often perceived as passive, rather than dominant and crusading within organizations.
    • The study indicates that the biggest obstacles to successful business analytic deployments are not data related, but in fact are due to skills and corporate culture. Strong analytic based businesses outperform others hands down, so it is not an issue of not wanting to be successful in this space. Rather, many executives just don't know where to begin and how to get this pushed into practice with so many other competing priorities.
    • Complexity, data quality, intend to “boil the ocean”, are all roadblocks to successful analytics. There's a need for a change management component in each and every business analytics initiative. Fact-based decision making is an activity that does require concrete buy-in from all stakeholders, data quality is key to overcome the Garbage in Garbage out situation, relevant and authorized access to the data is paramount. 
    • Talent Gap: last but not least, it’s lack of cross-boundary talent, the visionary executive who can integrate the collective thoughts, and hard data together to manage the strategy setting and decision making. more often than not, there's a huge disconnect between senior management saying they value analytics and doing what it takes to actually embed analytical thinking into one's organization. 

2.    The Principles in Shaping Culture of Analytics

Success in analytics is more about the people than it is about the tools. The point of analytics is not just to run the number, create new algorithms, etc., but to provide new and unique insights that drive decisions and/or strategies. Make sure the business decision is the first thing discussed. After that then create the analytic solution/methodology/process, work closely with customers to demonstrate how different types of analytics can be employed to provide insights into different types of business problems.
  • “The what, the why and the how" of how analytics helps to inform strategy. Reporting is about the "what". Analytics is about the "why". And you should care about the "why" because it will tell you tomorrow’s "what".  Analytics is not about getting to a point certain, but rather it needs to be used as a tool for reducing uncertainty in business. To make analytics pervasive within an organization, it needs to be promoted as a way of reducing, not eliminating, or explaining uncertainty in both strategic decision executives make and the daily decisions the staff make.  
  • Analytics must be easy to understand, easy to access and rely on validated corporate data as to avoiding the Fantasy Forecasting and Modeling (FFM) syndrome.

  • Let the data do the talking Analytics must drive employee decisions in the field without disclosing too much data, but "just enough" to help staff improve by small steps. Analytics technology exists that can be used to facilitate better communication with business leaders and reduce the aforementioned political problems. This leads to more fact based decisions being embraced and implemented by the business.

  • Simplicity because the people that are going to make the decisions based on the, "why", do not have time to understand the process behind delivering the "why". They just need to know what happened and what caused it.

  • Accountability is needed because far too often people in Analytics simply provide their portion of the work and walk away. Once the numerical part is over so is their part in the effort. Senior leadership gets jaded when this happens. They get all of these promises of closure once root cause is identified but then the ball is dropped because there is no follow through on the remediation of key business problems

  • Have really effective graphics and "presentation ready" results that show well and can be shared at the senior levels. Succinct and insightful commentary also helps get visibility and support. The snazzier and more professional it looks the more interest it gets. They do get excited when present ROIs, efficiency gains, market share, ability to select the most profitable customers, managing risk, and identifying new profitable niches

  • Analytics can redefine the box in such a way as to encourage and reward creative problem solving versus the same old, tired methods that produce consistent results but fail to reach potential if only because the new definition of success has raised the bar. A report is used to answer questions that are known in advance. Analytics are about asking and answering the next question(s), generally in the form of 'why' or even better, 'if/than'.

  • Measures are fundamental for anyone building a system and for such a few more measures, more metrics more angles are better to find and to fix problems. Leadership should demonstrate that they are leading; managing and operating through measurements and that will influence a lot to the rest of the levels in the organization. 

3.The Practices & Factors in Managing Analytics Program

The concept of "Think big, start small" is critical to advancing analytics in most organization, small successes can gain experience in advancing to the next levels of analytics (forecasting, segmentation analysis, predictive and optimization). The long term plan then has credibility because of those small successes and there is trust that the employees will individually benefit which leads to the company benefiting, always understand the key success factors and prepare for the next best practice: capture the latest analytics trends such as Social BI, Operations Analytics or Real-Time Analytics, Customer Loyalty Programs, In-Memory Analytics, etc. These topics will give food for thought to management and help get some momentum and enthusiasm behind effort.

  • People & Culture:  It is about the company culture and whether or not your people are "curious". Do you have employees who are willing to get past the ‘what’ part of numbers to wondering about the why? Is senior leadership in full support of adopting the changes in the culture necessary to get past the report/dashboard stage? Now comes the hard part, moving the organization to next levels so that predictive analytics, forecasting and optimization can be incorporated successfully. This is where training, education, and people will be critical. And the culture has to be right and open to accepting how analytics may change fundamental views about clients, processes, what is a profitable niche, etc.     
  • Process Selection: the key is to understand the business processes. In order to build out an Analytical Capability. Which business processes is the natural first beach head for the service? Marketing? Supply Chain? Operations? Customer Service?  Finance?  IT? Which would be the "last" door you would knock on?  Asking pertinent ‘WHY’ questions could be a good starting point, the real answer to "Why the business is losing customers?" can only be provided by the customers themselves. One way of starting out is conducting targeted surveys to build a "gut feel" and get a sense of direction. Having received inputs for building hypotheses, use data to validate or disprove them wherever possible, as the early part of the exercise is subjective with no statistical justification. And blind exploratory analysis could be equally futile. So always verify the numbers against the subjective information and vice-versa to ensure consistency. And, "Why?" question may be an intermediate to a "How?" question. and business eventually want to find out "How to retain customers?", "How to increase sales?" or "How to enhance profits by retaining profitable customers?"…etc. 
  • Move farther down from Visibility to Understanding to Guidance, you gain more and more added value. That's what you need to impart into your culture. 1) That Visibility is the first step in the process but delivers the least added value (but without visibility you're running around in the dark!). 2) Understanding offers more added values because it helps you understand the business problem better, so you can better create a solution. 3) And Guidance offers the most added values because it's objective and exact. Numbers never lie, only the people using them. So proper Guidance coupled with deep Understanding, both based on clear Visibility... those will deliver the most value to your company/group/team/projects. 
  • Analytic talent has 3 "C"s: curiosity, creativity and concentration. Discover which of your employees are curious, which ones are starting to see things that were not evident before and who will be the ones to advance organizationThe focus has been in identifying individuals who have the curiosity gene in them rather than looking at the specific business processes and who will quickly see benefits for them. 
  • Consider starting a Center of Analytic Excellence (CAE) that is tailored to that particular business' culture. Set out with strategies for: Organization, Architecture, Implementation, Governance, Training & Evangelism. The key activities that needs to be in place for effective implementation and utilization of Analytics are
    * Consistent support and direction from the leadership.
    * Business driving the analytics projects and initiatives.
    * Investment in training and sharing of information within the company.
    * Skill sets (both analytics, business)
    * Infrastructure, in terms of hardware and software.
    * Ready availability of sand box to explore and examine.
    * An integrated, clean and trustworthy data.
    * A data driven process that is consistent across the board and helps determine
    the value acquired through implementing and using Analytics.
    * Dedicated resources that focus on Analytics exclusively and be a main contact for business to drive their initiatives.
    * A governance structure that could help identifies who does what in the company so that there is accountability and responsibility.
    * Governance for security and data that could be implemented, maintained and followed 
  • Frameworks from  "Competing on Analytics" by  Thomas Davenport
    Framework 1: Key Questions Addressed by Analytics

    Information about the Past - Reporting
    Information about the Present - Alerts
    Information about the Future - Extrapolation
    Insight about the Past - Modeling, Experimental Design
    Insight about the Present - Recommendations
    Insight about the Future - Prediction, optimization, simulation

    Framework 2: Ladder of Analytical Applications
    Level 1: Get Data in Order (which would facilitate accurate, timely reporting)
    Level 2: Key Targets and Segments (which allows for customer profiling)
    Level 3: Differentiation Action (begin to think about how to treat segments differently)
    Level 4: Predictive Action (recommend differentiated treatments to segments based on some type of behavioral/attitudinal analysis)
    Level 5: Institutional Action (prediction and differentiated action embedded in repeatable processes...would also include a robust measurement strategy that creates a feedback loop to data collection, reporting, refined segmentation, etc.)
    Level 6: Optimization (recommend modifications to strategy, given a finite set of resources, based on learning from predictions, measurements, etc)  
Embedding analytics into organization culture is vital step in building the “learning organization, with competencies in data analysis, good information management systems which communicate the results of analysis, support and understanding from the top about the value of data analysis so that the effort is supported and promoted. People who work in the various data departments have excellent communication skills to link their work to the organization’s objectives and outputs, a governance structure which has integrating mechanisms so that information is shared, all departments are valued. So analytics culture becomes organizational DNA in competing for the future.


Post a Comment

Twitter Delicious Facebook Digg Stumbleupon Favorites More