Analytical Decisions: A Great Place to Start

Analytical decisions involve using data and systematic reasoning to evaluate options before choosing a course of action. Organizations that prioritize data-driven decision making reduce guesswork and improve outcomes across departments. Starting with analytical approaches establishes a foundation for consistent, measurable results. The following sections explore how to implement analytical decision frameworks effectively in your operations.

Organizations live in a world overflowing with data. As a result company decisions no longer need to rely solely on the “gut” of the leaders, or opinions of the outspoken.

The goal of this article is to discuss Analytical Decisions: A Great Place to Start. Thoughts will be shared on how you can approach incorporating data-based decision making into your company culture that will actually help you to better compete based on analytics.how executive coaching accelerates leader effectivenessmarketing leadership for scaling teams

At the heart of any company wishing to get better at Analytical Decisions is the DELT2A2 framework which has its origins in the work by Tom H Davenport. The following highlights the key components that companies should address:

It begins with identifying the Data that will be used to provide insights into the areas of opportunity and where the business should be focused. In many instances, data may not exist and the company needs to find ways to gather data. This can then be turned into information to be analyzed, which can then be turned into insights.

It is critical that all departments across the Enterprise are coordinating well to support resources related to analytics (people and tools) are being properly coordinated. Most companies or divisions that choose to compete on analytics have their employees who perform analysis and reporting in a centralized support function to use talent, provide cross-training and backup, and provide for growth opportunities.

Any company choosing to compete on analytics will need senior-level Leadership commitment, without this support the proper culture will not flourish and data-supported decision making will not be adhered to.

The organization must have processes in place to Target the initiatives with the best opportunities so that resources can be focused and prioritized where companies have the highest potential. A governance process must be in place to support all initiatives (where possible) are supported by analytics.

Securing the proper Technology tools to run the analysis needed is foundational to the success of competing on analytics.

Resourcing the right Analyst (depth and breadth), and supporting their continued growth is a cornerstone to a successful analytics implementation. It is critical that a company identify the proper level of analytical skills needed to conduct the types of analysis that are needed. Not every situation requires an individual with a PhD in mathematics.

Finally, the company must assess the various types of Analysis Methods that it should be used to compete in their marketplace.

Analytical Decision-Making

Analytical decision-making is one of four styles of decision making typically used by leaders. The other styles are directive, conceptual, and behavioral. In addition, consultative and consensus may also be used.

Steps to incorporating analytical decisions into your business

Numerous steps are involved to incorporating analytical decision making into your business practices and culture:

• Senior Leader Commitment: The style of your most senior leader will set the tone for the rest of the organization. If you wish to embed analytical decision-making into your culture then make sure your most senior leader believes in the value of analytical decision-making and practices it.
• Understanding your company’s analytical maturity: It is helpful to assess where you are today with your analytical capabilities. Many models exist to assess where you are at. You will find the following common stages of analytical maturity.
  • Descriptive Analytics: This is often considered data on what happened. It is a baseline capability that includes reports
  • Diagnostic Analytics: This is when you take raw data and use various analytical methods to answer “Why did something happen”. Tableau built a business from being able to drill-down into your data, mine massive amounts of information, and identifying correlations.
  • Predictive Analytics: Prescriptive Analytics builds on diagnostics to make predictions of what may happen based on existing data factors
  • Prescriptive Analytics: Prescriptive Analytics builds upon predictive to evaluate alternatives and identify what is the best that can happen.
  • Autonomous Analytics: Machine-learning can be incorporated when large amounts of data are available and self-learning algorithms can be applied to provide more insight or direction
• Determine which analytical techniques to use: The depth of analytical techniques you use will be highly dependent on your business and the skills of your personnel. The following techniques should at least be understood and reviewed by your senior leaders to determine which you may wish to use. These are outlined and explained further in the great book “Competing on Analytics: The New Science of Winning” by Thomas H. Davenport and Jeanne G. Harris.
  • Internal Process Evaluation
    • Activity-based costing (ABC). Allocating costs amongst activities by using models that incorporate activities, materials, resources, and product-offering components and then optimization based on cost and prediction of capacity needs.
    • Bayesian inference (e.g., to predict revenues). A numerical estimate of the degree of belief in a hypothesis before and after evidence has been observed.
    • Combinatorial optimization (e.g., for optimizing a product portfolio).
    • Constraint analysis (e.g., for product configuration). The use of one or more constraint satisfaction algorithms to specify the set of feasible solutions. Constraints are programmed in rules or procedures that produce solutions to particular configuration and design problems using one or more constraint satisfaction algorithms.
    • Experimental design: For website analysis
    • Future-value analysis: The decomposition of market capitalization into current value and future value, or expectations of future growth.
    • Genetic algorithms: Used for decryption/ code-breaking or product engineering/ design such as scheduling satellite communications, optimally loading cargo containers, and optimizing delivery routes.
    • Monte Carlo simulation: Used in project valuation using a computerized technique to assess the probability of certain outcomes or risks over multiple trials and comparing the outcome with predefined probability distributions.
    • Multiple regression analysis: Used to determine how non-financial factors affect financial performance.
    • Neural network analysis: Useful in predicting needed factory maintenance in which systems are initially “trained,” or fed large amounts of data and rules about data relationships.
    • Simulation: Used in pharmaceutical “in silico” research by manipulation of parameters using mathematics and/ or rule bases to model how different values would generate a result.
    • Textual analysis: Assesses text (such as transcribed call center discussions or textual data from a survey or social media) for customer sentiment. Yield analysis: Using means, median, standard deviation, etc. to understand yield volume and quality.

• Analysis of Supply Chains
  • Capacity planning: Optimizing the capacity of a supply chain or its elements. Identifying and eliminating bottlenecks.
  • Combinatorics: A sophisticated mathematical technique optimizes components in a supply chain.
  • Demand-supply matching: Determining the intersections of demand and supply curves to optimize inventory and minimize overstocks and stockouts.
  • Location analysis: Optimization of locations for stores, distribution centers, manufacturing plants, and so on.
  • Modeling: Creating models to simulate, explore contingencies, and optimize supply chains.
  • Routing: Finding the best path for a delivery vehicle around a set of locations.
  • Scheduling: Creating detailed schedules for the flow of resources and work through a process.
  • Simulation: Supply chain simulations model variation in supply flows, resources, warehouses, and various types of constraints. They allow for both optimization and visualization of complex supply chains.
• Analysis of Marketing
  • CHAID: A statistical technique used to segment customers on the basis of multiple alternative variables.
  • Conjoint analysis: A conjoint analysis is used to determine which factors:price, quality, dealer location, and so on:are most important to customers who are purchasing a new car.
  • Econometric modeling: Modeling to gain insight into complex market trends and the variables that affect market demand, supply, and costs.
  • Lifetime value analysis: Assessing the profitability of an individual customer (or a class of customers) over a lifetime of transactions.
  • Market experiments: Using direct mail, changes in a website (known as A/ B tests), promotions. And other techniques, marketers test variables to determine what customers respond to most in a given offering.
  • Multiple regression analysis: The most common statistical technique. For predicting the value of a dependent variable (such as…. Sales) in relation to one or more independent variables (such as the number of salespeople, the temperature. Or the day of the month).
  • Price optimization: Determining the price elasticity, or the response (changes in demand) of the buyer to increases or decreases in the product price.
  • Search engine optimization (SEO): Statistical methods and activities designed to improve a website’s ranking in search engines such as Google.
  • Support vector machine (SVM): This machine learning method uses training data to classify cases into one category or another. It is often used for customer segmentation and churn analysis.
  • Time-series experiments: These experimental designs follow a particular population for successive points in time. And are used to determine whether a condition that applied at a certain point led to a change in the variables under study.
  • Uplift modeling: A predictive modeling technique that directly assesses the incremental impact of promotional activity on a customer’s behavior.

• Determining Analytical Team Structure: There are a variety of ways to structure your team. Will you have the team reporting to finance, will analysts reside within departments, will the analysis be a function within your IT department, or will analytics be a stand-alone group? Regardless of the structure, you choose it will be important to work to you communicate the role that analysts play in the organization and how their input affects decisions.
• Hiring the Analytical Team: Since analysts can be very technical in their skill sets it is encouraged that you fully identify the different types of analysis that you feel will be employed. And determine the specific skills and competencies required to fulfill these types of analysis. Depending upon the existing capabilities within your organization you may need to consult with exterior experts to work to you recruit for the proper skills and competencies.
• Securing the Tools: The types of applications and methodologies you use will ultimately be driven by the types of analytical techniques you use and the skills of the team you hire. A listing of resources and options is to exhaustive to share here. Your analytics team and outside guidance can guide the selection of the tools you will need.
• Books: Reading about analytics can often be dry and academic. However, it is helpful for senior leaders to have a baseline understanding and appreciation for concepts and usage. The following list of books offer some insightful reading:
  • Competing on Analytics: The New Science of Winning by Thomas H. Davenport and Jeanne G. Harris
  • Naked Statistics: Stripping the Dread from the Data by Charles Wheelan
  • The Next America: Boomers, Millennials, and the Looming Generational Showdown by Paul Taylor
  • Astroball: The New Way to Win It All by Ben Reitler
  • Keeping Up with the Quants: Your Guide to “Understanding and Using Analytics by Thomas H. Davenport
  • Linear Regression and Correlation: A Beginner’s Guide by Scott Hartshorn
  • Freakonomics: A Rogue Economist Explores the Hidden Side of Everything by Steve D. Levity and Stephen J. Dubner
  • Big Data: A Revolution That Will Transform How Organizations Live, Work, and Think by Viktor Mayer-Schonberger and Kenneth Cukier
  • The Triple Package: What Really Determines Success by Amy Chua and Jen Rubenfeld
  • Future Smart: Managing the Game-Changing Trends That Will Transform Your World by James Canton
  • The Connection Algorithm: Take Risks, Defy the Status Quo, and Live Your Passions by Jess Warren Tevelow
  • Generation iY: Secrets to Connecting With Today’s Teens &. Young Adults in the Digital Age by Tim Elmore
  • Y-Size Your Business: How Gen Y Employees Can Save You Money and Grow Your Business by Jason Ryan Dorsey
  • Customer Data Integration: Reaching a Single version of the Truth by Jill Dyche and Evan Levey
  • The Little SAS Book by Lora D. Deheiche and Susan J. Slaughter
  • Analytics: Data, Science, Data Analysis and Predictive Analytics for Business by Daniel Covington
  • R for Everyone: Advanced Analytics and Graphics by Jared P. Lander

As computers become even more powerful, as data continues to proliferate. And as automation continues to advance it will become even more critical for companies to incorporate analytic decisions into their critical initiatives and day-to-day operations.

Frequently Asked Questions

What is the DELT-squared-A-squared framework?

The DELT2A2 framework (Davenport-origin) defines seven pillars for competing on analytics: Data, Enterprise coordination, Leadership commitment, Targeting high-value initiatives, Technology tools, Analyst talent, and Analysis Methods. Missing any single pillar undermines the entire analytical capability.

What are the four decision-making styles leaders use?

Leaders default to one of four styles: Analytical (data-driven systematic evaluation), Directive (quick decisions based on experience), Conceptual (big-picture creative thinking), and Behavioral (consensus-building through team input). Most leaders over-index on directive or gut-based approaches, leaving measurable value on the table.

What are the stages of analytical maturity?

Companies progress sequentially through Descriptive analytics (understanding what happened), Diagnostic analytics (understanding why it happened), and then higher tiers of predictive and prescriptive capability. Skipping stages creates capability gaps that undermine data-driven culture.

Should analytics teams be centralized or distributed?

Top-performing companies centralize analyst talent while distributing decision authority. This means housing analytical capability in a shared team that serves the entire organization while allowing business units to make their own data-informed decisions. Centralizing analysts ensures consistency; distributing decisions ensures speed.

How should organizations start with analytical decision making?

Start by assessing current decision-making practices honestly, identifying high-value decisions where data could improve outcomes, and building foundational descriptive analytics capability. Progress through maturity stages sequentially rather than attempting advanced analytics without the data infrastructure and talent to support them.