Application to PHD Program in Business Analytics Abstract Business entities have mastered the art of gathering information on almost every aspect that affects them. This data includes some of the most valuable aspects of their operations such as their customer numbers, market share, brand visibility, employees, and existing and potential competitors among others (nGenera, 2008). However, the existence of this plethora of information has only had a negative impact on balance sheets of most of these businesses due to the enormous amounts that are spent annually on data collection efforts. Research has established that organizations are losing on the opportunities that this data could have on their business improvement strategies due to lack of tools, methodologies and talent for utilizing the information at hand (nGenera, 2008).
Enterprise resource planning (ERP) systems therefore avail information but organizations need to employ business analytics to improve business performance through fact-based decision-making (Indian Institute of Management, Bangalore, 2014). Introduction Business analytics is a branch of business intelligence, which utilizes a set of proven techniques, and processes to analyze data for the purpose of improving performance through fact- based decision making (Indian Institute of Management, Bangalore, 2014).
Independent estimates project that global data will amount to 7.9 zettabytes by 2015. If the available data is analytically looked into, businesses will be able to improve operations and offer innovative services (NTT DATA, 2012). Financial institutions have successfully been able to employ business analytics to save millions of dollars lost through credit fraud. Retailers on the other hand use it to determine the best locations for their stores and appropriate stocking methods. On the other hand, pharmaceutical firms have been able to use it in the prompt distribution of life-saving drugs into the market (Davis, 2011).
This research paper will focus on the importance of business analytics to determine market segmentation strategies, the effectiveness of marketing channels and Paid Search on Sales and Margin. The research methodology employed and the resulting formulae are also presented. This conclusion discusses how the research findings enabled fact-based decision on various marketing strategies. Importance of Business Analytics (BA) The combined use of approaches, procedures and tools can be used to enable businesses gain valuable information, analyze it and predict outcomes in product planning, sourcing, manufacture and delivery.
This helps businesses to improve their inventory turnover and reduce the cost of sales (Trkman, 2010). For example, purchasing records data may be obtained and evaluated using analytics software to determine products that are of interest to consumers in specific demographic areas. This provides a reliable analysis of customer purchase history that leads to the supply of the right kind of products to the appropriate market segment (Elias, 2011). Business analytics is also beneficial in the determination of effective marketing strategies.
It provides the requisite intelligence to structure and mount marketing campaigns that produce results by analyzing the channels that have high product visibility and corresponding sales numbers (Elias, 2011). Analytics can be used to predict market trends of products and services through monthly and annual performance reports. This enables businesses to determine the return on marketing investments and to prioritize on the channels where to spend their marketing efforts in future (Trkman, 2010). Lastly, BA enables marketers to determine the effectiveness of their marketing channels through the determination of the contribution of each channel to the overall profits or losses.
This enhances decision making by ensuring that funds are spent on channels that produce maximum return on investments (Kaushik, 2014). This is because there is a wide range of channels to choose from for their brand campaigns in both electronic and print forms. To make a sound decision however calls for one to make a factual decision based on available evidence (Davis, 2011). Research Methodology This study will describe and explore the efficacy of business analytic approaches in determining market segmentation strategies.
It also assesses the rationale for marketing expenditure on different channels based on their contribution to the sales bottom line and the incremental impact of Paid Search on Sales and Margin. The study relies on secondary sources derived from the history of customer purchase data that are collected by the business to set up the research design. This approach has been selected because the sales data is readily available and it is directly relevant to the current study. Purchase Cadence I carried out this study by looking into the customers’ purchase history over a period covering the last 3 financial years.
From the records available, for a customer to be considered as a buyer, they ought to have made at least one purchase over the period under study. The information provided a basis for the calculation of the laps mean, laps standard deviation and recency. The Laps mean is the average time length between one purchase and the next, the Laps standard deviation is how closely the customer follows his or her purchase pattern while Recency is time length between most recent purchase and the time when the study is being carried out (now-which is relative).
From the three, I come up with a formula to get alert scores derived by laps mean, laps standard deviation, and recency. These alert scores were then used to segment customers based on frequency of purchasing which we classified into normal, frequent, most frequent, and abnormal. Offers were strictly supplied to the middle ranked customers ignoring those customers whose purchasing trends were, most frequent, abnormal and least frequent. Multi Channel Attribution Under this approach, the business is interested to know the channel from which customers made their purchase.
Based on that, I proceeded to give credits to the corresponding channel from which the purchase was made to understand the effectiveness of the business’ marketing expenditure. The historical approaches were biased in that businesses awarded full credits were to the last-touched channel ignoring the impact of earlier channels. To correct this, I pulled the customer’s data to track channels and touches over the last 45 days. For a customer to be considered as a buyer they ought to have made at least a purchase.
Then I used regression modeling to analyze the impact of channel interactions, so we can allocate the credits based on percentage attributions. I used linear regression to get the coefficient of each channel in order to determine the channel that made the most significant contribution to the business’ revenue. This was followed by the standardization of the coefficient (в=coefficient times standard deviation of Revenue divided by standard deviation of channel), in order to reduce the bias. To get the percentage attribution of each channel, I used в times mean of Revenue divided by standard deviation of Revenue.
Incrementality Test for Paid Searches This project aims to measure the incremental impact of Paid Search on Sales and Margin. I therefore created a Treatment and Control group based on the sales performance of the pre period compared to the one under study. I measured key differences between treatment and control through propensity score model and developed weights base on the model score. This was followed by balancing the two groups, in order to reduce bias.
I proceeded to serve the Paid Search advertisements to the treatment and kept the control group in the dark then measured sales and margin difference between two groups. Lastly, I calculated incremental impact by the weights model score. Conclusion Purchase cadence enables the business have an understanding of middle level purchasers preferences that solves the dilemma of having to decide on where, how and when to market their brand to them specifically. Data analytics assisted market segmentation enables the business to know to package their branding activities for this kind of purchasers.
Multi channel attribution on the other hand enables the business under study to be fully aware of how to structure their future marketing campaigns to improve their performance. Linear regression provides a quantifiable basis for making a choice on marketing channels, which the business needs, to concentrate on and those to be dropped. Lastly, Incrementality Test for Paid Searches makes it possible for the business to compute the Incrementality of adding or removing marketing strategies. References Davis, J. (2011, February 15). 8 essentials of business analytics: Find out what business analytics can do for you –and how to get started.
Retrieved from http: //www. sas. com/baexchange Indian Institute of Management, Bangalore. (June, 2014) Business Analytics and Intelligence. Retrieved from http: //www. iimb. ernet. in. SAS Institute. (2008). Business Analytics: Six Questions to Ask About Information and Competition. Retrieved from https: //www. sas. com Trkman, P., Mc Cormack, K., de Oliveira, M., & Ladeira, M.B. ( 2010). The impact of business analytics on supply chain performance. Decision Support Systems. PP 1-10. Retrieved from http: //www. eliseiver. com/ 10.1016/j. dss. 2010.03.007.