GENERAL INSTRUCTIONS TO CANDIDATES

1. Individual Work Only: Any evidence of collusion, contract cheating or unauthorised assistance will be treated as serious academic misconduct.
2. Open-Book Conditions

• You may consult textbooks, lecture notes, academic journal articles, and credible online sources.
  • However, analysis, structure, interpretation and wording must be your own.
  • Copying (including from AI tools) without proper citation is plagiarism.

• Referencing & Academic Integrity

• Use APA referencing style consistently.
  • Include in-text citations and a reference list at the end of your script.
  • You are expected to reference at least: Sharda, R., Delen, D., & Turban, E. (2023). Business Intelligence, Analytics, and Data Science: A Managerial Perspective (12th ed.). Pearson. Include a minimum of 5 additional academic sources across your four answers.

All submissions are subject to plagiarism and authorship checks under the University’s regulations.

• Word Count

• Total suggested word count: 3,500 words (excluding references, tables, figures, and appendices).
  • Each question has an indicative word count; ±10% is acceptable per question.
  • State the word count at the end of each answer.

• Formatting of Written Script

• Typed, 12-point Times New Roman, 1.5 line spacing, normal margins
  • Number all pages and clearly label each answer (eg. Question 1, Question 2, etc.).
  • Save and submit the written report as a WORD document. File name format below
    • Written report: Student ID- Course Code.docx
    • PowerPoint Present: Student ID-Course Code.pptx
    • Video Report: Student ID-Course Code.mp4

• Mandatory Maximum 15-Minute Video Presentation with PowerPoint: This is needed to support academic integrity and assess your ability to communicate complex ideas. There are 2 parts to this

1. PowerPoint Slides (Slide 1 should have student name, ID, course code & case title)

• Prepare a PowerPoint of approximately 8–12 slides summarising your four answers, key BI recommendations, and proposed dashboard/system designs for VRDG.

• Video Presentation

• Record a narrated video presentation based on your slides (Dress officially).
  • Duration: 12–15 minutes.
  • Format: MP4
  • In the Video, you must:

• Appear on camera while screen-sharing slides
  • Introduce yourself (Student ID and course).
  • Explain your main arguments and conclusions in your own words (do not simply read your script).
  • Demonstrate that you understand and can defend your own analysis.

CASE TITLE:

“DATA IN THE DARK: VOLTA RETAIL & DISTRIBUTION GROUP AND THE BUSINESS INTELLIGENCE TRANSFORMATION IMPERATIVE”

1. Company Background

Volta Retail & Distribution Group (VRDG) is a fictional pan-African retail conglomerate founded in 2005 and headquartered in Accra, Ghana. Named after the Volta River, VRDG operates across three business divisions: (1) Volta Supermarkets — a chain of 148 supermarkets and convenience stores in Ghana (102 stores), Nigeria (31

stores), and Côte d’Ivoire (15 stores); (2) Volta Wholesale — a business-to-business distribution arm supplying 6,200 independent retailers, kiosks, and informal market traders; and (3) Volta Connect — a mobile commerce and loyalty platform with 1.1 million registered users, enabling customers to order groceries, earn loyalty points, and access micro-savings products.

VRDG employs 9,400 staff across all three countries and generates annual revenues of GHS 780 million (approximately USD 52 million). The company processes over 480,000 point-of-sale (POS) transactions per day across its store network and approximately 64,000 mobile commerce orders per month through Volta Connect.

Despite its scale, VRDG operates in what its CIO, Dr. Emmanuel Asante-Poku,

describes as a state of ‘data darkness.’ Transactional data from store POS systems, the wholesale ERP, the Volta Connect mobile platform, the HR system, and the finance module are stored in seven separate, incompatible systems with no integration layer.

Executive reporting relies on manually compiled Excel spreadsheets that take finance teams up to 12 days to produce each month. There is no data warehouse, no analytics capability, and no dashboards. The CEO has received no near-real-time operational visibility since the company’s founding.

EXHIBIT 1: VRDG Current IT Systems Landscape

System	Function	Data Volume	Vendor/Technology	Key Gap
Oracle POS (148 stores)	Sales transaction processing	480k transactions/day	Oracle Retail	No central aggregation; each store is siloed

SAP ERP (Wholesale)	Procurement, inventory, invoicing for wholesale division	6,200 customer orders/week	SAP S/4HANA (partial)	Not integrated with retail POS or HR
Volta Connect App	Mobile orders, loyalty programme, customer data	64k orders/month; 1.1m customer profiles	Custom-built (Laravel/MySQL)	No analytics layer; loyalty data not linked to store sales
Sage Payroll & HR	Employee records, payroll, attendance	9,400 employee records	Sage 300 People	No link to store performance or sales KPIs
QuickBooks (Côte d’Ivoire)	Local finance and accounts payable	15-store data	QuickBooks Enterprise	Incompatible with SAP; manual reconciliation
Excel (Finance)	Monthly management reporting	All historical data	Microsoft Excel	12-day report lag; error-prone; no forecasting
WhatsApp & phone logs	Informal supply chain communication with kiosk retailers	Unstructured; unquantified	N/A	No structured capture; invisible to management

• Key Business Intelligence Challenges

VRDG’s Board has approved a GHS 4.2 million investment in a Business Intelligence Transformation Programme over 24 months. The CIO has identified five critical intelligence gaps that the programme must address:

Gap 1 — Inventory & Stockout Visibility: VRDG loses an estimated GHS 18 million per year in sales due to stockouts at store level. Without real-time inventory data, replenishment decisions are made manually by store managers based on experience rather than data. Perishable goods wastage (fresh produce, dairy) is estimated at 11% of category revenue.

Gap 2 — Customer Intelligence: VRDG has 1.1 million loyalty app users but cannot link their digital behaviour to in-store purchasing. Customer segmentation has never been performed. The marketing team cannot identify its most profitable customers, at-risk churners, or the optimal product mix for different customer segments.

Gap 3 — Supply Chain & Supplier Performance: VRDG works with 340 suppliers. Supplier on-time delivery, quality rejection rates, and pricing trends are tracked informally. No supplier performance scorecard exists. The CFO estimates that 22% of procurement spend is inefficient due to lack of data-driven negotiation and demand-based ordering.

Gap 4 — Cross-Country Financial Consolidation: Monthly financial statements for Ghana, Nigeria, and Côte d’Ivoire require 12 working days to consolidate due to incompatible systems and currency conversion processes performed manually in Excel. The CFO cannot see real-time P&L by country, store, or product category.

Gap 5 — HR & Workforce Analytics: With 9,400 staff across three countries, VRDG’s voluntary turnover rate is 28% per year — significantly above the 15–18% retail industry benchmark in West Africa. The HR Director has no predictive attrition model, no employee engagement data, and no link between workforce metrics and store performance outcomes.

EXHIBIT 2: VRDG Business Performance Snapshot (Last Financial Year)

Metric	Ghana	Nigeria	Côte d’Ivoire	Group Total
Revenue (GHS million equivalent)	512	198	70	780
Gross margin	22.4%	19.1%	21.8%	21.5%
Stockout rate (% of SKUs per	14.2%	18.7%	11.4%	15.3%

week)
Perishable wastage (% of category revenue)	10.8%	13.2%	9.1%	11.2%
Customer loyalty app active users	780k	240k	80k	1.1 million
Voluntary staff turnover	24%	34%	26%	28%
Monthly report production time	12 working days	12 working days	12 working days	12 working days

EXHIBIT 3: VRDG BI Maturity Assessment (Conducted by External Consultant)

BI Capability Area	Current Maturity Level (1–5)	Industry Benchmark	Gap
Data integration & warehousing	1 — No integration; fully siloed	3.8	Critical
Reporting & dashboards	1 — Manual Excel; 12-day lag	3.5	Critical
Descriptive analytics	1 — Basic Excel summaries only	3.2	Critical
Predictive analytics	1 — None	2.9	Critical
Prescriptive analytics	1 — None	2.3	Significant
Data governance & quality	1 — No formal governance	3.1	Critical
BI user adoption & literacy	2 — Basic use of Excel by finance	3.0	Significant

EXHIBIT 4: Volta Connect Mobile App – Customer Behaviour Data Sample (Last Quarter)

Customer Segment (Inferred)	% of App Users	Avg. Monthly Spend (GHS)	Avg. Order Frequency	Top Categories	Churn Risk (Self-reported)
High-frequency urban shoppers	18%	GHS 420	8.2x/month	Dairy, fresh produce, snacks	Low (9%)
Weekly family shoppers	31%	GHS 285	3.9x/month	Staples, beverages, household	Medium (22%)
Occasional deal-seekers	27%	GHS 110	1.4x/month	Promotions, impulse items	High (41%)
Dormant users (no order in 90 days)	24%	GHS 0	0x	N/A	Very high (76%)

You are engaged as an MBA 505 student and BI strategy consultant to VRDG’s CIO, Dr. Emmanuel Asante-Poku. Drawing on the exhibits and your knowledge of business intelligence, analytics, MIS, and data systems, you are required to answer ANY FOUR

(4) of the six questions below. Your answers should demonstrate both technical BI knowledge and managerial decision-making judgment appropriate to VRDG’s African retail context.

ANSWER ANY FOUR (4) OUT OF SIX (6) QUESTIONS

Each question carries 15 marks. All parts within each chosen question must be answered.

Part a — BI and MIS Architecture Diagnosis                                           [5 marks] Using VRDG’s current IT systems landscape (Exhibit 1), apply the MIS components framework (Transaction Processing Systems, Management Information Systems, Decision Support Systems, and Executive Information Systems) to classify each of VRDG’s seven existing systems. Identify the two most critical system integration failures that are limiting VRDG’s ability to convert operational data into management intelligence, and explain the business cost of each failure using specific Exhibit 1 and 2 data.

Part b — Target BI Architecture Design                                                   [5 marks]

Design a target-state BI architecture for VRDG. Your architecture must include: (i) a data layer (source systems, ETL pipeline, and a centralised data warehouse); (ii) an analytics layer (descriptive, predictive, and prescriptive analytics capabilities); and (iii) a presentation layer (dashboards, reports, and self-service BI tools). Draw or describe the architecture clearly,

labelling each component and explaining its function in VRDG’s context. Identify which existing VRDG systems feed into each layer and what integration approach should be used to connect them.

Part c — Data Strategy and Governance Framework                              [5 marks] VRDG’s data is currently ungoverned, siloed, and inconsistent. Propose a data strategy and governance framework for VRDG’s BI transformation. Your framework must address: data ownership (who is responsible for data quality in each functional area); data standardisation (how VRDG will resolve inconsistencies between Oracle POS, SAP, and QuickBooks data);

and a prioritisation approach for which data integration should be tackled first, with justification based on VRDG’s five identified intelligence gaps.

Part a — Data Warehouse Design                                                               [5 marks]

Design a dimensional data warehouse schema for VRDG’s retail operations using either a star schema or snowflake schema (clearly state and justify your choice). Your schema must include at least one fact table and four dimension tables appropriate for VRDG’s business. For each table, specify the primary and foreign keys and at least four key attributes. Explain how this schema would enable VRDG to answer the following management questions: (i) Which product categories generate the highest gross margin per store per month? (ii) Which Volta Connect loyalty customers have the highest lifetime value?

Part b — ETL Process Design                                                                     [5 marks]

Design an ETL (Extract, Transform, Load) process for integrating VRDG’s three most data-rich source systems: the Oracle POS, the SAP ERP (Wholesale), and the Volta Connect mobile app. For each of the three ETL stages, describe: (i) Extract — what data is extracted from each source system, at what frequency, and using what method (batch vs real-time API);

(ii) Transform — identify at least two specific data quality or transformation challenges VRDG will encounter (e.g., currency conversion between GHS, NGN, and XOF; inconsistent product codes across systems) and how each should be resolved; (iii) Load — specify the loading strategy (full load vs incremental load) and justification.

Part c — Data Quality Assessment & Master Data Management           [5 marks] Based on the Exhibit 1 systems landscape, identify three specific data quality risks VRDG faces as it consolidates its data into a central warehouse (examples: duplicate customer records between Volta Connect and in-store POS, inconsistent product category naming

across countries, missing transactional data from WhatsApp-based informal orders). For each

risk, explain: the business impact if unresolved; the data quality dimension affected (accuracy, completeness, consistency, timeliness, or uniqueness); and a Master Data Management (MDM) approach VRDG should adopt to prevent or resolve the issue.

Part a — Descriptive Analytics: Sales & Inventory Insight                     [5 marks] Apply descriptive analytics to VRDG’s situation. Using Exhibit 2 data, design a descriptive analytics report for the CEO that answers: (i) Where are VRDG’s largest performance gaps across countries, using stockout rate, perishable wastage, and gross margin as the primary

metrics? (ii) What does the cross-country comparison suggest about operational consistency and best-practice sharing? Specify the visualisation types you would use (e.g., heatmap, waterfall chart, bar chart, trend line) for each insight, and explain why each visualisation is appropriate for its specific data type and management audience.

Part b — Predictive Analytics: Customer Churn & Demand Forecasting [5 marks] Design two predictive analytics models for VRDG: (i) A customer churn prediction model using Exhibit 4 loyalty app data. Identify the target variable, at least four predictor variables available in Exhibit 4, the most appropriate modelling technique (e.g., logistic regression,

decision tree, random forest), and how the output would be used by VRDG’s marketing team to intervene before a customer churns. (ii) A demand forecasting model for perishable goods inventory. Identify the input variables, the forecasting method most appropriate for VRDG’s data environment (given that historical sales data is siloed in Oracle POS), and how the model output would directly reduce the 11.2% perishable wastage rate identified in Exhibit 2.

Part c — Prescriptive Analytics: Inventory Optimisation Decision        [5 marks]

VRDG loses an estimated GHS 18 million per year in stockout-related lost sales. Using

prescriptive analytics logic, design an automated replenishment decision model for VRDG’s

supermarket network. Your model must specify: (i) the decision variables (what the model decides, e.g., order quantity per SKU per store); (ii) the objective function (what the model optimises, e.g., minimise total cost of stockouts + holding costs + ordering costs); (iii) the key constraints (e.g., supplier lead times, warehouse capacity, store-level shelf space); and (iv) how the model’s outputs would be presented to store managers and supply chain teams in a format they can act on without requiring data science expertise.

Part a — CEO Executive Dashboard [5 marks]

Design an Executive Dashboard for VRDG’s CEO that provides a single-screen view of group performance. Your design must specify: (i) exactly six KPIs to display, drawn from Exhibits 2 and the five intelligence gaps described in the case, with current values, targets, and RAG (Red/Amber/Green) status indicators; (ii) the visualisation type for each KPI (e.g., gauge, bullet chart, sparkline, map, trend line) with justification for each choice; (iii) the data refresh frequency and source system for each KPI; and (iv) the dashboard layout logic — explain how you have prioritised and arranged the six KPIs on the screen to reflect strategic importance and ease of executive scanning.

Part b — Store Manager Operational Dashboard                                    [5 marks]

Design a daily operational dashboard for VRDG store managers. Unlike the CEO dashboard which shows group-level strategy, this dashboard must support daily operational decisions at the individual store level. Specify: (i) five operational metrics relevant to a store manager

(e.g., today’s sales vs target, current stockout count by category, perishable items expiring within 48 hours, staff attendance, customer complaint count); (ii) the appropriate chart or display type for each metric and why it supports fast decision-making for a store manager with limited data literacy; (iii) one alert/trigger rule that the dashboard should fire

automatically when a threshold is breached (e.g., if stockout rate exceeds 15%, alert store manager and procurement team simultaneously).

Part c — Data Visualisation Principles & Tool Selection                         [5 marks] VRDG’s CIO must select a BI and dashboard tool for deployment across the organisation. Evaluate three BI visualisation tools — Microsoft Power BI, Tableau, and a locally accessible alternative such as Google Looker Studio — against four criteria relevant to VRDG’s African operating context: (i) cost and licensing model; (ii) offline/low-bandwidth capability (relevant for stores in rural Nigeria and Côte d’Ivoire); (iii) integration with Oracle POS and SAP; and (iv) ease of use for non-technical store managers and finance staff.

Recommend one tool with clear justification and identify two key data visualisation principles (e.g., pre-attentive attributes, Tufte’s data-ink ratio, the IBCS notation standards) that VRDG’s dashboard designers should follow.

Part a — Decision Support System Design for Supply Chain                  [5 marks] Design a Decision Support System (DSS) for VRDG’s supply chain and procurement function. Your DSS design must specify: (i) the decision domain it supports (which decisions, at what level — strategic, tactical, or operational); (ii) the three core DSS

components (database management, model management, and user interface/dialogue) and what each component contains in VRDG’s context; (iii) at least two specific decision

scenarios where the DSS would produce a structured recommendation for VRDG’s procurement team (e.g., supplier selection when lead time risk is elevated; order quantity optimisation ahead of a promotional period); and (iv) how the DSS differs from the Executive Information System (EIS) that VRDG’s CEO would use for strategic oversight.

Part b — Big Data Technologies for Volta Connect [5 marks]

VRDG’s Volta Connect mobile platform generates a growing volume of customer interaction data: 64,000 orders per month, app click-stream data, loyalty point transactions, customer feedback ratings, and social media-linked behaviour. Evaluate whether VRDG’s current MySQL-based Volta Connect database is adequate for this data volume and variety. Using the 5Vs of Big Data (Volume, Velocity, Variety, Veracity, Value), assess each dimension

against VRDG’s current data situation. Recommend one Big Data technology stack appropriate for VRDG’s scale and African infrastructure context (e.g., Apache Hadoop,

Apache Spark, Google BigQuery, or Amazon Redshift), justifying your recommendation

against at least three of the five Big Data dimensions and VRDG’s infrastructure constraints (e.g., internet reliability in Nigeria and Côte d’Ivoire).

Part c — Cloud BI Strategy & Real-Time Reporting                               [5 marks] VRDG’s CIO is evaluating whether to deploy the BI transformation programme on a cloud-based infrastructure (SaaS/PaaS) or an on-premise solution. Evaluate the cloud vs on-premise decision for VRDG against five criteria: (i) total cost of ownership (capex vs opex model);

• data sovereignty and regulatory compliance across Ghana, Nigeria, and Côte d’Ivoire;
• internet connectivity reliability as a constraint for real-time cloud-based reporting in lower-connectivity markets; (iv) scalability as VRDG’s data volumes grow; and (v)

integration complexity with VRDG’s existing Oracle and SAP systems. Recommend a deployment model (full cloud, hybrid, or on-premise) with justification, and propose one approach for achieving near-real-time financial consolidation reporting that eliminates the current 12-working-day monthly reporting lag.

Part a — BI for HR Analytics and Workforce Management [5 marks]

VRDG’s voluntary staff turnover rate of 28% (well above the 15–18% West African retail benchmark) costs the company an estimated GHS 6,500 per departing employee in recruitment and retraining costs — a total annual cost of approximately GHS 17 million.

Design a People Analytics BI solution for VRDG’s HR Director. Your solution must specify:

(i) the data sources to be integrated into the HR analytics model (from Exhibit 1 and the case context, including Sage HR, Oracle POS performance data, and management observation records); (ii) a predictive attrition model — identify at least five predictor variables that would be available in VRDG’s context (e.g., tenure, department, store performance quartile, manager NPS, overtime hours logged); (iii) how the model outputs would be translated into a weekly HR dashboard metric that a non-technical HR Business Partner can use to trigger targeted retention interventions.

Part b — BI Implementation Roadmap & Risk Management                 [5 marks] VRDG’s Board has approved GHS 4.2 million and 24 months for the BI transformation programme. Design a phased BI implementation roadmap for VRDG. Your roadmap must include three phases: Phase 1 (Months 1–6): Foundation — specify which data integrations and data warehouse components to build first and why, based on VRDG’s highest-priority intelligence gaps; Phase 2 (Months 7–15): Analytics & Dashboards — specify which analytics capabilities and dashboards to deploy and to which user groups; Phase 3 (Months 16–24): Optimisation & Scale — specify predictive/prescriptive analytics and cross-country BI expansion. For each phase, identify the primary risk and propose a mitigation strategy.

Identify one BI implementation failure mode that is particularly prevalent in African enterprise contexts (e.g., data literacy gaps, infrastructure instability, change resistance from middle management) and propose a specific countermeasure for VRDG.

Part c — Change Management & BI User Adoption                                [5 marks]

BI systems frequently fail not because of technical deficiencies but because of poor user adoption, organisational resistance, and leadership disengagement. VRDG has 9,400 staff, most of whom have never used a BI tool; store managers are accustomed to phone calls and WhatsApp messages as their primary information channel. Design a change management and user adoption plan for VRDG’s BI transformation. Your plan must address: (i) stakeholder mapping — identify at least four stakeholder groups (e.g., store managers, finance team, CEO/EXCO, IT team) with their current engagement level and adoption risk; (ii) a training

and capability-building programme that accounts for VRDG’s varying digital literacy levels across Ghana, Nigeria, and Côte d’Ivoire; (iii) an adoption measurement framework —

specify three metrics that VRDG’s CIO should track to monitor whether the BI tools are being used effectively after go-live (e.g., daily active dashboard users as % of licensed users, number of manual Excel reports replaced, decision cycle time reduction).

Note: The written examination, marked out of 100%, shall be converted to a 30% weighting toward the final course grade. An additional 30% will be allocated to the mandatory PowerPoint & video presentation, bringing the total examination contribution to 60% of the overall course assessment.