ASSESSMENT 4: FINAL PROJECT REPORT AND

PRESENTATION
Abstract
This project uses business analytics and data visualization to study retail sales data and find useful
insights for company leaders. Understanding sales trends, tracking regions and products with
performance issues and forecasting sales supports better planning of both inventory and marketing.
Data was cleaned, analyzed, and put into a dashboard using Microsoft Excel, Python and Power BI.
The main findings are noticeable spikes in sales at different times, differentiated performance of
products, and differences between regions. Learnings from this study can enhance how the business
operates, increase its earnings, and ensure customer demand is met.
Introduction resource and first used Microsoft Excel to

Because the retail market is so competitive clean it and process it preparatory to machine

now, businesses depend more on data learning. At this step, the team examined data

analytics to make better choices and operate and fixed any discordant or missing

more effectively. Because retail businesses information before analysis began.

handle huge amounts of transactional data, Researchers changed the format and the

turning it into useful insights is necessary to structure to help the data work with analytical

compete effectively. The main issue for tools. The clean data was then uploaded to

business retailers is that they’re not able to Power BI, allowing us to examine it further

understand and apply large datasets to and make useful dashboards for making

forecast what customers want, manage decisions.

available items and personalize their Key Tools Used

promotions. It solves this problem with the Researchers mostly relied on Power BI to

help of business analytics and visualization on build the dashboards and present sales data

a full retail sales dataset. The objective is to through visualization. It supported visual and

identify tasks, trends and patterns that aid in changing ways to illustrate trends, patterns

shaping decisions that give the business an and comparisons in product categories and

advantage. locations. Because the system was easy to use

Methodology and had powerful models, researchers did the

analysis and reporting easily and promptly.
In this project, the data consists of records
With the help of DAX and visualization tools
from retail sales showing information about
in Power BI, Researchers performed complex
products, how much was sold, where it was
calculations, analyzed trends and did
sold and the dates of the transactions.
forecasting [1]. With the use of charts, slicers,
Researchers obtained the data from a public
1
and filters, the dashboard made it easy for along the analysis, the analyst assumes that
stakeholders to get a detailed picture of sales the data behaves the same each time and does
and gain useful advice from the data. not account for factors that may influence this
Analysis Process like price changes, products being unavailable
At the outset, null and duplicate entries were or customers’ changing preferences [3].
removed to secure the quality of the data. Thanks to this approach, trends and patterns
Where some data was missing, researchers can be analyzed similarly everywhere,
either deleted it or used sensible estimates so although it does not always capture all the
the results remain accurate. Duplicate records details unique to specific retail situations.
were found by looking at the transaction IDs Limitations
and timestamps and then cleared from the Discounts, competition and changes in the
data. The way dates were stored was changed economy were not studied here, even though
to a uniform method for smoother analysis of these can play a big role in determining sales
time data [2]. As a result, researchers could numbers [4]. There is no information about
study changes over time, different seasons customers’ demographics in the data, so it is
and patterns that repeat. In order to make difficult to segment or analyze their actions.
comparisons easy, similar products and Because of this, the data reveals what actions
regions were grouped before the analysis people take but not the truth behind why they
began. To see how aggregate sales patterns make each purchase.
changed over the years, trend analysis was Analytical Dashboard & Insights
applied and seasonal decomposition was used
to pinpoint the repeating highs and lows. It
helped me to better understand the busiest and
slowest months for revenue. Short-term
unpredictability was reduced and underlying
patterns were discovered using simple
moving averages. The process was iterative Figure 1: The dashboard
(Source: Self-created)
and focused on ensuring accuracy, relevance,
Using Power BI, designed an analytical
and usability of the insights derived from the
dashboard that features interactive filters for
dataset.
date, region and product category. Find the
Assumptions
segmented data on sales trends, how products
The analysis expects the data to be accurate
are doing, analysis by region and sales
and unedited by anyone. It is also assumed
predictions here.
that no area or category is more important
than another in the overall evaluation. All
2
Sales Trends active and involved customers but not many
The sales trends area shows that in November purchases [6]. This might mean that the
and December, sales were much higher due to business has supplier delays, inefficient
people shopping for the holiday season. This transportation or is sometimes out of stock.
ongoing pattern points out when each With this comparison, businesses are able to
business might get particularly busy during identify places where operations can be
the year. The rise in sales seen during the year improved.
points to an increase in customer demand and Forecasting
the wider growth of the market. Through this In this section, simple moving averages and
visualization, stakeholders observe changes linear regression are applied to predict the
over a period and measure how well strategies performance of the company’s future sales.
have performed. The system uses historical sales figures to
Product Performance predict that sales will go up by 8% next
In this section, sales contributions from quarter. Using this approach, businesses can
individual products are compared and see what is likely to happen and use that to
differences are pointed out. Everywhere guide their future decisions. Users can use
Product X is available, it brings in consistent forecast visuals to forecast likely trends, as
revenue, so it deserves first place in the well as rely on confidence intervals to foresee
marketing and should be stocked ahead of likely times of high or low activity.
others. On the other hand, Product Z produces Business Impact
the least amount of total revenue, indicating The dashboard reveals that it has a major
there may be problems with demand, pricing impact on business activities. Paying special
or how it’s being advertised [5]. With the attention to strong demand products during
analysis, stakeholders can compare key November and December can improve the
product metrics, choose the best inventory inventory strategy [6]. If Product Z does not
handling, decide on advertising budget and perform as expected, it may be examined to
address phasing out of products. see if it can be improved or it should be ended
Regional Analysis to lower costs. In addition, focusing on
The regional analysis part sheds light on how marketing and sales in profitable places such
sales are distributed by region. Region A as Region A becomes practical, while
reports the biggest total sales, giving it the attention is given to solving problems in
best performance among the markets. It Region B.
means customers are interested in the product, User Instructions
it is available or the distribution system is Easily understood and used, the dashboard is
effective. This Group, however, has many handy for those unfamiliar with coding. Each
3
analysis sector on the platform has a distinct Recommendations
tab labeled Sales, Product Performance, Management should follow some strategic
Regional Analysis and Forecasting [7]. guidelines motivated by the results on the
Customers can use slicers to look at data for dashboard to boost performance. Stocking up
specific dates, various locations or different on best-selling products becomes necessary in
products, to get insights suitable for them. Q4 to handle the growing demand. Efforts
Additional details and information appear should mainly focus on Regions A and C
when the mouse hovers over the visuals. because they offer the best sales
Ethics and Contemporary Issues opportunities. It would be helpful to analyze
Using data ethically is a key part of good operations in Region B to find out if supply
business intelligence and data analysis. Since chain problems are stopping sales growth,
no personal or confidential information about even though customers are very engaged [9].
customers was included, data privacy was When making product decisions, bundling
kept up-to-date and all data protocols were together or removing products with low
obeyed, recognized ethical practices. No demand can streamline what anyone offered
biased algorithms were used in the process, and raise profits. Real-time data integration
but it’s important not to mistake regional means the dashboards will always display
sales differences for assumptions about information on the latest company activities.
consumer choices. Although automation helps Conclusion
with efficiency, people need to be involved to Analyzing the retail sales data has revealed
properly understand intricate patterns and to useful details that can guide better business
ensure ethical choices [8]. Using predictive decisions, relying on data. The increases in
analytics, resources can be used more demand during the final quarter create
effectively, yet it’s important to check why chances to adjust marketing strategies and
certain regions are doing poorly rather than handle inventory better. High-performing
just ignoring them. By being fair, analysis products must be supported with stronger
prevents business strategies from sidelining investment and a bigger presence, as low-
specific markets or groups of products. By performing items might call for
embedding ethical awareness into the reconsideration of strategy. Regional
analytical process, businesses can promote disparities point to operational challenges that
transparency, fairness, and accountability in need addressing to unlock potential growth.
their data-driven decisions. Overall, the integration of ethical practices
and continued dashboard development
ensures that the insights remain actionable,

4
fair, and aligned with long-term business 6. Neba, C., Gerard, S.F., Nsuh, G.,
goals, supporting continuous improvement Amouda, P., Neba, A., Webnda, F.,
Orelaja, A. and Sylla, N.A., 2024.
and competitive advantage in a dynamic retail
Advancing Retail Predictions:
environment. Integrating Diverse Machine Learning
Reference Models for Accurate Walmart Sales
Forecasting. Asian Journal of
1. Sutiteannarong, A. and Tai, P.D., Probability and Statistics.
2023. Analysis on retail sales data 7. Trinh, N.T., Chalumeau, M.,
(Doctoral dissertation, Thammasat Bruckner, T.A., Levy, C., Bessou, A.,
University). Milic, D., Cohen, R., Lemaitre, M.
2. Wang, C.H. and Gu, Y.W., 2022. and Cohen, J.F., 2021. Monitoring
Sales forecasting, market analysis, and outpatient antibiotic utilization using
performance assessment for US retail reimbursement and retail sales data: a
firms: a business analytics population-based comparison in
perspective. Applied sciences, 12(17), France, 2012–17. Journal of
p.8480. Antimicrobial Chemotherapy, 76(9),
3. Qidi, J., 2021, March. Research on pp.2446-2452.
Influencing Factors of Retail Sales in 8. Trinh, N.T., Chalumeau, M.,
E-Commerce Market. In 2021 2nd Bruckner, T.A., Levy, C., Bessou, A.,
International Conference on E- Milic, D., Cohen, R., Lemaitre, M.
Commerce and Internet Technology and Cohen, J.F., 2021. Monitoring
(ECIT) (pp. 16-19). IEEE. outpatient antibiotic utilization using
4. Fatema, B., 2024. Exploring the link reimbursement and retail sales data: a
between offline and online retail sales: population-based comparison in
insights from 15 years of retail sales France, 2012–17. Journal of
data in Great Britain. Management Antimicrobial Chemotherapy, 76(9),
Decision. pp.2446-2452.
5. Zubair, M., Waleed, A., Rehman, A., 9. MacKillop, J., Cooper, A. and
Ahmad, F., Islam, M. and Javed, S., Costello, J., 2021. National retail sales
2024, October. Machine Learning of alcohol and cannabis during the
Insights into Retail Sales Prediction: COVID-19 pandemic in Canada.
A Comparative Analysis of JAMA network open, 4(11),
Algorithms. In 2024 Horizons of pp.e2133076-e2133076.
Information Technology and
Engineering (HITE) (pp. 1-7). IEEE.

5
10.

6
References