MyPictureOnline aims to gain enough visibility | My Assignment Tutor

Table of Contents Table of Contents. 2 1       Task 1: A data analytics Engine to manage sales in MyPictureOnline. 2 1.1         Introduction. 2 1.2         Implementing Qualitative and Quantitative analysis. 4 1.3         Applying other analysis techniques for decision making. 6 1.4         Conclusion. 9 2       Task 2: Tableau analysis. 9 2.1         Different types of expense categories. 9 2.2         A comparison of travelling expenses and all other expense categories and the yearly trend in each country  10 2.3         Departments in low, medium & high expense groups in each country. 19 2.4         Prediction of spending patterns in non-travel related expenses for the next financial year. 26 2.5         Conclusion. 26 3       Post scriptum.. 27 4       References. 27 1         Task 1: A data analytics Engine to manage sales in MyPictureOnline 1.1        Introduction MyPictureOnline aims to gain enough visibility into their consolidated sales data from all the outlets in different countries. Making huge business decisions could be nerve-wrecking. Nonetheless, having enough data to analyze, could make the process more comfortable. Data analysis will produce findings to determine which areas need improving or which areas to drop completely. Data analysis could also be used to figure out problems such as loss of revenue. The report discusses different analysis techniques available for MyPictureOnline. 1.1.1        Diagnostic Analytics MyPictureOnline company discovered revenue losses that upon consolidation of sales data.  Diagnostic analytics involves a deep-dive into the dataset in order to find insights such as correlations and answers as to why the volume of sales produced such results. Diagnostic analysis will attempt to produce the reasoning behind the results and offer meaningful information for decision making and for building good predictive models. Figure 1‑1 Diagnostic Analysis Data analysts could find correlations between variables in the data such as time of the year and total sales so as to understand what drive the sales up or down. (Uspenskiy, 2012)Diagnostic analysis involves three main steps: Data discovery – identification of relevant data sourcesDrill down – centering focus on a specific part of the dataData mining – gaining information from massive data to build predictive models. 1.1.2        Data Mining MyPictureOnline collects a colossal amount of data from the 3000 outlets in different counties as well as online purchases. In data mining, such huge available raw data is used to extract valuable information and knowledge. Figure 1‑2 Data Mining The target data obtained from selection is then preprocessed, and transformed. Consequently, data mining algorithms such as classification algorithms (decision trees and K nearest neighbors), detection of anomalies, and clustering are applied  to realize patterns. Based on  the gained patterns and  correlations, interpretations could be made to predict future trends such as the expected volume of sales in a certain period. New patterns could be identified and classified for example specific people of a certain country prefer specific photo production equipment or service. Other meaningful insights from data mining include detecting associations and similarities for instance when a customer requires a certain type of equipment, they will most likely prefer another specific subsequent service or equipment. 1.2        Implementing Qualitative and Quantitative analysis 1.2.1        Quantitative analysis Qualitative analysis involves manipulation of raw data to come up with meaningful information which can be acted up on or to be used in the decision-making process. Figure 1‑3: Qualitative Analysis Process Financial managers at MyPictureOnline want to develop a quantitative model to optimize the annual profits, and determine the Breakeven point of some outlet stores in order to be profitable. The model could be developed using the following steps: Define the Problem: Optimization of annual profits and to determine outlet break even points of some stores.Develop Model: A model based on a controllable/independent variable such as different classes of equipment and services against the total sales gained. Patterns could be identified from the model to determine and optimize which equipment had the maximum sales. Regression can also be used to observe how total sales(depended) behave when advertisements(Independent) are put out.Acquire Input Data: It is important to get the correct datasets that could be manipulated to produce the intended insights. Relevant dataset variables in this scenario will include total sales, average annual profits and  inventory amount.Develop solutions:  Model variables are manipulated to get the desired result. Trial and error and complete enumeration methodology could be used to find the optimal sales needed to break even.Test solution: implement the solution to the stores to see how it works.Analyze solution: Here sensitivity analysis could be used to change model parameter such as increase Tv ads and see how the model changes.  The more sensitive(rapid changes), the less effective the model will be.Implement the solution: solutions could include increasing advertisements, or improving delivery times to reach the required number of sales to make profits. 1.2.2        Qualitative Analysis This is analyzing non-quantifiable data and factors that could affect the decision-making process. A scenario could occur whereby, in a specific country, federal legislation require equipment imports to have 2019 and above as the year of manufacture. Consequently, based on such factors, MyPictureOnline will only export photo equipment inventory to stores in the specified country with the required year of manufacturing. Managers at MyPictureOnline need critical qualitative tools such as entrepreneurial  instincts, understanding consumer needs, and better market penetration strategies. While such tools are important, qualitative analysis works better when joined with quantitative analysis which are numbers showing financial reports, production data and sales reports (Kim et al., 2012). Combining both qualitative and quantitative analysis produces enough information to support decision making. 1.3        Applying other analysis techniques for decision making Other data analysis techniques that could support decision – making in MyPictureOnline include: 1.3.1        Quantitative Analysis Quantitative analysis together with expert judgement produces meaningful insights which are needed to make quality decisions. Quantitative analysis can be used in either optimizing(finding best through maximizing or minimizing) or satisficing(finding something that is good enough) of business processes (Yousuf and Zainal, 2020). If a data analyst at the company is dealing with a deterministic problem, for instance the resources needed to adequately produce all the photo production services and equipment needed, linear programming could be utilized. Linear programming could determine the resources needed while accounting for constraints such as cost and time in order to achieve maximum profitability. If the analyst is developing a stochastic model to determine why money is being lost upon sales consolidation, probabilities can be used. There could be a probability of accounting errors, or poor stock control and management processes leading to holding of stocks, or inadequate access control systems for storage facilities. 1.3.2        Visual analysis Visual analysis involves data visualization which is painting a picture of how data is behaving and communicating which helps support good decision making. Data visualization will show patters, trends, findings and facts that must be internalized in order to make quality business choices. Figure 1‑4 Visual Analysis (Kohlhammer et al., 2011) A data analyst at a big company like MyPictureOnline must use data visualization tools such as Tableau and Oracle to better communicate and help decision makers better understand their sales and inventory data. A chart of sales against all retail stores depicts the stores with poor sales making bases for decision making. 1.3.3        Semantic analysis Semantic analysis draws meaning from text and extracts valuable information from unstructured datasets such as customer feedbacks, emails and suggestions. (Haller, Link and Groß, 2017) Semantic analysis models are made to mimic the human brain and how it processes information and context to form cognitive decisions. Figure 5.Semantic Analysis (MonkeyLearn Blog, 2020) MyPictureOnline could develop semantic analysis systems using machine learning algorithms to better understand customer feedback and formulate ways of better improving the customer experience.  Figure 1‑5 Semantic Analysis. (Sciencedirect.com, 2017) 1.3.4        Heat maps MyPictureOnline is an international company with stores in different geographical countries. The sales and marketing team will need heat maps to better understand the geographical and consumer data and leverage it to improve the customer experience. Heat maps are visual analysis business tools which plots quantitative data on geographical map and uses color intensity to display different values in data.   Figure 1‑6 same data but one is easier to understand (Hotjar, 2020)  Heat maps will display the current state of the company in terms of sales and product use and more importantly they can depict areas of opportunities and intensity areas which communicates where the majority consumers are allocated. Managers at the company could use all the insights from heat maps to justify business decisions. 1.4        Conclusion There are more than enough data analysis techniques for the company to leverage on to design decision support systems and aid in making quality business choices. The analysis techniques discussed can  help managers at MyPictureOnline be aware of the company’s bottom line, discover trouble areas where the company is hurting, how the company is fairing on the market relative to its competitors, and to improve consumer experience and study customer behavior and product interaction, all  based and baked by facts. 2         Task 2: Tableau analysis 2.1        Different types of expense categories From the given data, there are five types of expense categories that the company has been incurring for the years 2014, 2015, 2016, 2017 and 2018. These five expenses are Miscellaneous expenses,Car expenses (parking)Mileage expenses,Air travel expenses andTrain expenses. The screenshot below from Tableau shows a graphical visualization of the five expenses. Figure 2.1 The five different expense categories Rationale and justification: The rationale for this working is pretty simple because it only involves one attribute that is the category and one measure value that is the expense cost all which have been provided by the data set. By adding category on the column shelf and expense cost on the row shelf, the above graph is obtained. Since expense cost is a measure value, Tableau sums all records for the different expenses and obtains the total of each different expense incurred by the company. The sum of the different expenses incurred is show on top of each bar which represents an expense. 2.2        A comparison of travelling expenses and all other expense categories and the yearly trend in each country The screenshot below shows a comparison between travel expenses and other expenses (miscellaneous, air travel) in each country. Figure 2.2 Travel expense vs all other expense categories It is evident that travel expense is more than other expenses in all countries except Germany, Romania and United Kingdom where miscellaneous expenses are more. Rationale and justification: In solving this, the dimension country is dragged into the columns shelf and Expense_cost into the rows shelf. A bar graph is obtained and is colored by category. Before coloring, travelling expense which comprises of parking, train, mileage is grouped. To obtain clearer insights, percentages are calculated downwards for each bar. For the yearly trends in each country, the start_date is added to the column shelf to obtain a line graph which shows how the different expenses vary from country to country in a span of four years. 2.2.1        Belgium Figure 2.3 Yearly trend in expenses in Belgium In Belgium, travel expenses are high in all years except 2014 when miscellaneous expenses were more by 3,039. (31,303-28,244). There is a sharp increase of miscellaneous expenses from 2016 to 2017 as shown by the graph. Air travel expenses have been lower than all other expenses and have decreased from 2016 to 2017. 2.2.2        Germany Figure 2.4 Yearly trend in expenses in Germany In Germany, miscellaneous expenses are more than travel expenses. In 2016, miscellaneous expenses were at their lowest while travelling expenses were at their highest. Air travel is lowest but has been steadily increasing over the years. 2.2.3        Netherlands Figure 2.5 Yearly trend in expenses in Netherlands Travel expenses in Netherlands are more than other expenses except in 2014. Air travel is lowest but fluctuates every year from high to low. 2.2.4        Poland Figure 2.6 Yearly trend in expenses in Poland In Poland, the first two years that is 2014 and 2015, travel expenses are higher than all other expenses and then for 2016 and 2017 miscellaneous expenses are more than all other expense. Air travel is the lowest and has been steadily increasing from 2015 to 2017. 2.2.5        Romania Figure 2.7 Yearly trend in expenses in Romania Travel expenses are more in Romania except in 2016 when there was an abrupt increase in miscellaneous expenses. Air travel expenses are low and fluctuate yearly. 2.2.6        United Kingdom Figure 2.8 Yearly trend in expenses in The United Kingdom In the United Kingdom, miscellaneous expenses are more than all other expenses except in the year 2016 where travel expenses were more. Air travel is lowest but fluctuates yearly. 2.2.7        United States Figure 2.9 Yearly trend in expenses in The United States Miscellaneous expenses are more than all other expenses in the United states. At the beginning, travel expenses were highest but have reduced drastically. Air travel is lowest and fluctuates yearly. 2.3        Departments in low, medium & high expense groups in each country 2.3.1        Belgium Figure 2.10 Departments in low, medium and high expense groups in Belgium Low – financeMedium – printingHigh – HR Rationale and justification: The two dimensions used are country and departments and are dragged into the column shelf. The expense cost which is the measure value is dragged into the rows shelf. The departments are ranked in order of expenditure. Countries are filtered to produce results for each country separately. 2.3.2        Germany Figure 2.11 Departments in low, medium and high expense groups in Germany. Low – financeMedium – printingHigh – photography 2.3.3        Netherlands Figure 2.12 Departments in low, medium and high expense groups in Netherlands Low – RetailMedium – DispatchHigh – Marketing 2.3.4        Poland Figure 2.13 Departments in low, medium and high expense groups in Poland Low – RetailMedium – HRHigh – Marketing 2.3.5        Romania Figure 2.14 Departments in low, medium and high expense groups in Romania Low – PrintingMedium – MarketingHigh – Dispatch 2.3.6        United Kingdom Figure 2.15 Departments in low, medium and high expense groups in United Kingdom Low – HRMedium – DispatchHigh – Photography 2.3.7        United States Figure 2.16 Departments in low, medium and high expense groups in United States Low – FinanceMedium – PhotographyHigh – Printing 2.4        Prediction of spending patterns in non-travel related expenses for the next financial year Miscellaneous category is the only non-travel related expense. The screenshot below shows how much MyPictureOnline spent on miscellaneous expenses for the past four years. Figure 2.17 MyPictureOnline miscellaneous expenditure 2.5        Conclusion The screenshot above shows how the company’s miscellaneous expenses have been on the rise for all countries except Romania. Based on the current trend of rising miscellaneous expenditure, it is predictable that the company will also spend more money in covering miscellaneous expenses for the next financial year. From the above analysis, it is evident that the company spends more money in miscellaneous expenses than on travelling expenses. MyPictureOnline needs to find a way to bring down its miscellaneous expenses because the company is losing a huge amount of money in this category.    3         Post scriptum Word count:  2369 4         References Haller, A., Link, M. and Groß, T. (2017) ‘The Term “Non-financial Information”–A Semantic Analysis of a Key Feature of Current and Future Corporate Reporting’, Accounting in Europe, 14(3), pp. 407–429. doi: 10.1080/17449480.2017.1374548. Hotjar. (2020). What Are Heat Maps? Guide to Heatmaps/How to Use Them | Hotjar. [online] Available at: https://www.hotjar.com/heatmaps/ [Accessed 13 Mar. 2021]. ‌Kim, H. M. et al. (2012) ‘A qualitative analysis of information dissemination through twitter in a digital library’, Proceedings of the ACM/IEEE Joint Conference on Digital Libraries, pp. 339–340. doi: 10.1145/2232817.2232880. Kohlhammer, J., Keim, D., Pohl, M., Santucci, G. and Andrienko, G. (2011). Solving Problems with Visual Analytics. Procedia Computer Science, [online] 7, pp.117–120. Available at: https://www.sciencedirect.com/science/article/pii/S1877050911007009 [Accessed 13 Mar. 2021].‌ MonkeyLearn Blog. (2020). Semantic Analysis: What Is It & How Does It Work? [online] Available at: https://monkeylearn.com/blog/semantic-analysis/ [Accessed 13 Mar. 2021]. Sciencedirect.com. (2017). Semantic Analysis – an overview | ScienceDirect Topics. [online] Available at: https://www.sciencedirect.com/topics/computer-science/semantic-analysis [Accessed 13 Mar. 2021]. Uspenskiy, V. (2012) ‘Diagnostic system based on the information analysis of Electrocardiograph’, 2012 Mediterranean Conference on Embedded Computing, MECO 2012, pp. 74–76. Yousuf, H. and Zainal, A. Y. (2020) ‘Quantitative approach in enhancing decision making through big data as an advanced technology’, Advances in Science, Technology and Engineering Systems, 5(5), pp. 109–116. doi: 10.25046/aj050515.