INFT2150 Business Analysis Essay Is Crucial For Understanding Business Processes.

This is the act of iterative, procedural exploration through a firm/organizations data base with a mentality of analysis in order to achieve a data driven decision. These are collected from the various records in the firm records.

Business analytics is very necessary. This has always enabled the firm to realize the progress after a specified period. It has always enabled businesses to understand the strength points and points of weaknesses. This is very key in planning as it enables investors and business runners to improve on weakness and mature up on various areas. They are able to determine from statistics the quality of products needed and the kind of customers they need to supply most. This is very efficient in business prosperity.

Business analytics has got a wide ecosystem of operation. These includes; descriptive, predictive, prescriptive and the exploratory models. For the descriptive, the business checks into the past records; market record, production record, finance record and operations records. This gives the business a platform to plan for the future as it helps the organization understands the past; this may be a minute ago or years back. This can also be applied in finding the correlation of various variables i.e. the market demand for movies in the market.

Predictive analysis in the ecosystem gives the business insights of the future based on the data. It estimates what is likely to be the future outcome in the organization. Predictive analysis is achieved through various ways; machine learning, data mining, game theory and modeling (Nisbet, et al., 2009). This can be used to determine the success the business is likely to accrue in the future. By this, all departments are able to know their future progress i.e. the kind of behavior the customer will have for a specified quality of goods, the number of credit that will be given out after a specified period (Stubbs, 2011). From history and transactional data, patterns achieved to arrive at the future predictions.

The last stage, prescriptive, helps one understand the business very well. It’s an innovation trigger that usually takes 5-10 years to arrive to its peak. It foresees what will happen in the future, how it will happen and why it will happen. It thereby gives recommendations on how to curb the predictions in order to achieve advantages in the predictions (Davenport, et al., n.d.). To achieve it, any techniques are applied; mathematical sciences, business rule algorithms, machine learning and computational modeling, historical data, social data and many other data sources (Bartlett, 2013). This technique is very key in the success of the business as it determines the scheduling, production, inventory and supply chain design to ensure right products at the right time to the right persons (Das & Vidyashankar, 2006).

To achieve this, data mining is very key. This is the extraction of the necessary data, combination and analysis in order to achieve the necessary ecosystem in the business analytics. First, the data is prepared. The appropriate technique is then applied to ensure a specific pattern/ relation between the data in the data sets (Finlay, 2014). The objective of the data is mainly considered while arriving at this. The algorithm applied will determine the pattern which the data will follow.

The data is then interpreted and assessed to confirm its accuracy, validity and relevance. If this is confirmed right, then the data is used as tool in driving decisions and if it is irrelevant then a larger data is applied (Korn, 2011). This ensures the right data is applied for accurate decision making.

After ensuring the data is accurate and relevant, the necessary models from the conclusions is applied: when the organization producing bulbs realizes that it’s the energy saver bulbs that is in high demand at a specific town, then they will increase the supply of that particular bulb in that area.

However, the data mined to acquire the business analytics may have challenges. Since the data applied are usually big most of the problems are realized;

• it’s very technical managing big data,
• the security of the data may also be a problem,
• inaccurate data may be used that may result to inaccurate predictions,
• the expense of working on the big data may be too much,
• the large data may accrue the problem in maturity as it demands for a lot of time,
• Since the data used is large, it may be difficult to reuse the data.

This may make it hectic implementing the business analytics. If the data required is not well mined and interpreted using the right algorithms then the conclusion is likely to fail the business and the strategy may land into a loss.

Accuracy should therefore be taken seriously in order to ensure right channels and conclusions that will ensure the success of a business.

The achievement of the business analytics is however hindered by various factors. This involves up to leadership involve. The following are challenges that may result from the governance in the organization;

• The governance that does not understand the business itself. Where a leader does not know the main objective of not only that business but the general sector of such particular business then the business is likely to fail. He/ she should thereby consider the kind of competitors and the kind of market to be explored.
• They governance that does not focus on value. The leadership team that are not innovators and do no explore the opportunities within the business and how it can improve and expand the market is likely to fail the business analytics. This is because they will not be willing to see the success and the future of the business.
• The governance that does not speak the right language in the business. Leaders that are not solutions to the changes they want are likely not to implement them. Such have never created room for the opportunities in the business analytics hence making technical for implementations.
• Poor leader worker relations and interdepartmental relations may also be a problem. In an environment where departments are not in good terms, workers are not well spoken and there are various communication problems, the business analytics can be rarely applied as team work to realize mistakes and points to be improved isn’t there.
• The governance where the personalities are not analytical. The leaders should deep and thoughtful. They must be focused and serious individuals that will help the implantation take place. Failure to which the business dynamics may not take place.
• The leaders who are not drivers. With no confidence such cannot be achieved. The leaders that are not willing to risk in such are likely not to implement the business analytics to achieve a success.

When these problems are not sorted out, then the leadership culture in the business may not be positive hence the business analytics not applied. This will automatically lead to the business failure.

Proper governance must then be considered to enhance the business a success.

The aim of this section is to provide descriptive analytics by developing a taxonomy of descriptive analytics. Descriptive analytics comprises of mainly measures of central tendency as well as measures of dispersion. Measures of central tendency include; mean, median and mode while measures of dispersion include; standard deviation, range and coefficient of variation. These measures can be represented as shown in the following diagram;

Technique: the mean also known as the average

Purpose: it measures the mid-point (around which all other values cluster) of a set of values

Functionality: obtained by adding up all the numbers together then dividing by the count of the numbers.

Assumptions: we assume that the given set of numbers follow a normal distribution

Method of validation: the mean of another sample is plotted, the numbers are then plotted in a histogram to check for the normality distribution

Sample use case: we find the average monthly spending on non-healthy foods for all the participants. This is obtained as follows;

Technique: the median

Purpose: it measures the mid-point (middle) of a set of data that has outliers.

Functionality: obtained by arranging values in either ascending or descending order and obtaining the middle value. However for even numbers, it is the value of

Assumptions: we assume that the given set of data has outliers

Method of validation: plotting a boxplot and checking whether there are values outside the plot area

Sample use case: we find the median for the monthly spending on non-healthy foods for all the participants. This is obtained as follows;

Technique: the mode

Purpose: it measures the most frequent value in a set of values.

Functionality: obtained by counting the frequency of a given value and obtaining the value with the most frequent counts.  

Assumptions: we assume that the frequency of the values is more than once

Method of validation: count the number of times a given set of values appears

Sample use case: we find the mode for the monthly spending on non-healthy foods for all the participants. This is obtained as follows;

Mode=431

Technique: the standard deviation (SD)

Purpose: it measures how spread out a given dataset (distribution) is

Functionality: begin by obtaining the mean. The mean is subtracted from the individual values then the result is squared. The sum of the squared results is divided by total counts minus 1. What is obtained is the variance. To get the standard deviation, we obtain the square root of the variance.

Assumptions: we assume that the given set of numbers follow a normal distribution

Method of validation: the mean of another sample is plotted, the numbers are then plotted in a histogram to check for the normality distribution

Sample use case: we find the standard deviation of the monthly spending on non-healthy foods for all the participants. This is obtained as follows;

Technique: the coefficient of variation (CV)

Purpose: it measures the relative dispersion of an event or data

Functionality: obtained by dividing the standard deviation by the mean; expressed as a percentage.

Assumptions: we assume that the given set of numbers follow a normal distribution

Method of validation: the mean of another sample is plotted, the numbers are then plotted in a histogram to check for the normality distribution

Sample use case: we find the coefficient of variation of the monthly spending on non-healthy foods for all the participants. This is obtained as follows;

Technique: the range

Purpose: it measures the range of the data.

Functionality: obtained by subtracting the minimum value from the maximum value.

Assumptions: we assume that the given set of data has outliers

Method of validation: plotting a boxplot and checking whether there are values outside the plot area

Sample use case: we find the range for the monthly spending on non-healthy foods for all the participants. This is obtained as follows;

In this section, we present a regression model where we develop a linear regression model using the data on immunity provided. The data was captured to evaluate the propensity scores of employees from a company to contract a flu during a winter season. The data comprises of three variables given in the table below;

Table 1: Description of the variables

Table 2 below presents the correlation matrix between the three variables

Table 2: Correlation matrix (Pearson)

Results shows that there is a weak negative relationship between propensity score and non-healthy food (r = -0.348). However, there is a strong negative relationship between propensity score and percent in office. The negative relationship implies that an increase in the percent in office results to a decrease in the propensity score.

A linear regression model was developed using both R and Excel. The R scripts are given in the appendix A1.

First we look at how fit the model is fit. Looking at the p-value of the F-statistics we see that the value is less than 5% level of significance (p-value = 0.0001).

The R-Squared (R²) is 0.742; which means that 74.2% of the variation in the dependent variable (propensity score) is explained by the two independent variables in the model.

The next table below presents the regression coefficients.

From the table, the following regression equation model can be fitted out of it;

The only significant independent variable in the model is the percent in office (p-value < 0.05)

This study sought to apply analytic techniques to show the benefit of descriptive as well as predictive analytics in a business environment. Results showed that there is a significant strong negative relationship between propensity score and percent in office. Regression model constructed showed that percent in office significantly predicts the propensity score.

Bartlett, R., 2013. A Practitioner’s Guide To Business Analytics: Using Data Analysis Tools to Improve Your Organization’s Decision Making and Strategy.

Das, K. & Vidyashankar, G. S., 2006. Competitive Advantage in Retail Through Analytics: Developing Insights, Creating Value. Information Management.

Davenport, T. H., Jeanne, G. H., David , D. L. W. & Alvin , J. L., n.d. Data to Knowledge to Results: Building an Analytic Capability. California Management Review, 43(2), p. 117–138.

Finlay, S., 2014. Predictive Analytics, Data Mining and Big Data. Myths, Misconceptions and Methods.

Korn, S., 2011. The Opportunity for Predictive Analytics in Finance.

Nisbet, R., Elder, J. & Miner, G., 2009. Handbook of Statistical Analysis & Data Mining Applications.

Stubbs, E., 2011. The Value of Business Analytics.