Big Data in E-Commerce

Status: final, Type: Report

Wanru Li, fa20-523-329, Edit

Keywords: e-commerce, big data, data analysis

1. Introduction

E-commerce is already changed by big data a lot. As we can see in the lecture slides, the retail store was closed rapidly in the past three years. It can be seen that e-commerce has begun to take shape and has been accepted by customers. E-commerce can make shopping more convenient for customers, and enable companies to better discover current trends and customers' favorite categories for better development.

For customers, they can find the item they want easier than find it in a retail store. Maybe the customer doesn’t know what he wants, doesn’t know his brand, only knows its style, but that’s enough to search for the item on e-commerce. There are also some e-commerce services that offer photo search, which makes shopping easier. Shopping in e-commerce usually keeps a record of the purchase. In this way, you don’t have to go to a retail store to buy some products repeatedly. Instead, you can directly find the products in the record and place orders, which saves a lot of time.

For companies, there are more changes. They can analyze customers preferences and purchasing power based on their browsing data, shopping cart data, and purchasing data. Large enough to predict the future business trend, small enough to better see the customer’s evaluation of the product.

E-commerce companies have access to a lot of data, which makes it easy for them to analyze product trends and customer preferences. As talent says, “Retail websites track the number of clicks per page, the average number of products people add to their shopping carts before checking out, and the average length of time between a homepage visit and a purchase. If customers are signed up for a rewards or subscription program, companies can analyze demographic, age, style, size, and socioeconomic information. Predictive analytics can help companies develop new strategies to prevent shopping cart abandonment, lessen time to purchase, and cater to budding trends. Likewise, e-commerce companies use this data to accurately predict inventory needs with changes in seasonality or the economy” ¹. There is an example of the Lenovo, to enhance the customer experience and stand out from the competition, Lenovo needs to understand customers' needs, preferences, and purchasing behaviors. By collecting data sets from various touchpoints, Lenovo USES real-time predictive analytics to improve customer experience and increase revenue per retail segment by 11 percent ¹.

Meeting customer needs is not just an immediate problem. E-commerce depends on having the right inventory in the future. Big data can help the company to be prepared for the emerging trend, in the slow or potential prosperity and development of the year, or around major activity plan marketing activities. E-commerce companies will compile large data sets. By evaluating the data of a few years ago, electronics retailers can plan accordingly inventory, inventory to predict peak, simplify the overall business operations, and predict demand. E-commerce sites, for example, can do it in the shopping rush hour in social media significantly depreciate sales promotion, to eliminate redundant products. In order to optimize pricing decisions, e-commerce sites can also provide a special discount. Through big data analysis and machine learning, learn when to offer discounts, how long they should last, and what discount prices are offered more accurately (para 8).

E-commerce is bound to dominate the retail market in the future because it can help retailers better analyze and predict future trends, which retailers cannot resist. At the same time, e-commerce provides better ways for customers to shop. With better analysis, retail companies will be able to provide better service to customers, so e-commerce will be more and more accepted and popular in the future.

2. Background Research and Previous Work

As Artur Olechowski wrote, “According to the IDC, the digital universe of data will grow by 61% to reach a smashing 175 zettabytes worldwide by 2025. There’s no denying that a large chunk of the digital world belongs to e-commerce, which takes advantage of customer social media activity, web browser history, geolocation, and data about abandoned online shopping carts. Most e-commerce businesses are able to collect and process data at scale today. Many of them leverage data analytics to understand their customers’ purchasing behaviors, follow the changing market trends, gain insights that allow them to become more proactive, deliver more personalized experiences to customers. The global Big Data in the e-commerce industry is expected to grow at a CAGR of 13.27% between 2019 and 2028. But what exactly is Big Data? And how can e-commerce businesses capture this powerful technology trend to their advantage? In this article, we take a closer look at the key trends in the usage of Big Data technologies by e-commerce companies and offer you some tips to help you get started in this game-changing field” ².

The most common and widely used application of big data is in e-commerce. Nowadays, the application of big data in e-commerce is relatively mature. As Artur Olechowski wrote, “As businesses scale up, they also collect an increasing amount of data. They need to get interested in data and its processing; this is just inevitable. That’s why a data-driven e-commerce company should regularly measure and improve upon: shopper analysis, customer service personalization, customer experience, the security of online payment processing, targeted advertising” ².

There are also some disadvantages of the big data, or to say more need to do after getting the data. Artur Olechowski wrote, “Understand the problem of security — Big Data tools gather a lot of data about every single customer who visits your site. This is a lot of sensitive information. If your security is compromised, you could lose your reputation. That’s why before adopting the data technology, make sure to hire a cybersecurity expert to keep all of your data private and secure”² . Security is always a big problem with big data. This is one of the components will be analyzed in my report. He also wrote, “Lack of analytics will become a bigger problem — Big Data is all about gathering information, but to make use of it, your system should also be able to process it. High-quality Big Data solutions can do that and then visualize insights in a simple manner. That’s how you can make this valuable information useful to everyone, from managers to customer service reps” ². The analysis is also an important part of the big data. Only collecting data cannot help e-commerce anything. Security and analytics will be talked about in my report.

3. Choice of Data-sets

QUARTERLY RETAIL E-COMMERCE SALES 2 nd QUARTER 2020:https://www.census.gov/retail/mrts/www/data/pdf/ec_current.pdf

For the dataset, the source website provided in the project requirements will be used, if there needs more information, data and information on the web will be searched for. As a result of recent COVID-19 incidents, many organizations work in a small capacity or have entirely ceased activities. The Census Bureau has tracked and analyzed the response and data quality in this manner.

Monthly Retail Trade from Census will be analyzed. The Census Bureau of the Department of Commerce today reported that the forecast of U.S. retail e-commerce revenue for the second quarter of 2020 adjusted for seasonal fluctuations, but not for price adjustments, was $211.5 billion, a rise of 31.8 per cent (plus or minus 1.2 per cent) from the first quarter of 2020. Total retail revenues were projected at $1,311.0 billion for the second quarter of 2020, a decline of 3.9 percent (plus or minus 0.4 percent) from the first quarter of 2020. The e-commerce forecast for the second quarter of 2020 increased (para1).

Retail e-commerce sales are estimated from the same sample used for the Monthly Retail Trade Survey (MRTS) to estimate preliminary and final U.S. retail sales. Advance U.S. online transactions are calculated from a subsample of the MRTS survey that is not of appropriate magnitude to calculate improvements in retail e-commerce transactions.

A stratified basic random sampling procedure is used to pick approximately 10,800 retailers, except food services, whose transactions are then weighted and benchmarked to reflect the entire universe of over two million retailers. The MRTS sample is focused on probability and represents all employer firms engaged in retail activities as described in the North American Industry Classification System (NAICS). Coverage covers all vendors whether or not they are active in e-commerce. Internet travel agents, financial brokers and distributors, and ticket sales companies are not listed as retail and are not included with either the gross retail or retail e‐commerce sales figures. Non employees are reflected in the projections by benchmarking of previous annual survey estimates that contain non employer revenue based on administrative data. E-commerce revenues are included in the gross monthly sales figures.

The MRTS sample is revised on a continuous basis to account for new retail employees (including those selling over the Internet), company deaths and other shifts in the retail business environment. Firms are asked to report e-commerce revenue on a monthly basis separately. For each month of the year, data for non-responsive sampling units shall be calculated from reacting sampling units falling under the same class of sector and sales size segment or on the basis of the company’s historical results. Responding firms account for approximately 67% of the e-commerce sales estimate and approximately 72% of the U.S. retail sales estimate for any quarter.

Estimates are obtained by summing the weighted sales (either reported or charged) for each month of the quarter. The monthly figures are benchmarked against previous annual survey estimates. Quartal projections are determined summing up the monthly benchmarked figures. The forecast for the last quarter is a provisional forecast. The calculation is also open to revision. Data consumers who make their own projections using data from this study can only apply to the Census Bureau as the source of input data.

This article publishes forecasts optimized for seasonal variation and variations in holiday and trade days, but not for adjustments in rates. As feedback for the X‐13ARIMA‐SEATS programme, we have used the updated figures of quarterly figures of e-commerce revenue for the fourth quarter 1999 up to the present quarter. For revenue, we estimated the quarterly adjusted figures for each year with an additional modified monthly revenue forecast. Seasonal estimate adjustment is an approximation based on current and previous experiences.

The estimates containing sample errors and non-sample errors in this article are based on a survey.

The difference between the prediction and the results of the full population listing under the same sample conditions is the sampling error. This mistake happens when a national poll only tests a sub-set of the total population. Estimated sampling variance measurements are standard errors and variance coefficients, as stated in Table 2 of this article.

4. Search and Analysis

This year the pandemic accelerated growth in ecommerce in the US, with online revenues projected to hit just 2022. The top 10 ecommerce retailers will strengthen their hold on the retail market with our Q3 American retail prediction.

This year, revenues of US eCommerce are projected to hit $794.50 billion, up 32.4% annually. This is even more than the 18.0% predicted in our Q2, since customers are now ignoring shops and opting to buy online in the wake of the pandemic.

In “US Ecommerce Growth Jumps to More than 30%, Accelerating Online Shopping Shift by Nearly 2 Years”, it says, “‘We’ve seen ecommerce accelerate in ways that didn’t seem possible last spring, given the extent of the economic crisis,’ said Andrew Lipsman, eMarketer principal analyst at Insider Intelligence. ‘While much of the shift has been led by essential categories like grocery, there has been surprising strength in discretionary categories like consumer electronics and home furnishings that benefited from pandemic-driven lifestyle needs’” ³.

This year, ecommerce revenues are projected to hit 14.4 percent and 19.2 percent of all US retail spending by 2024. Without purchases of petrol and cars, ecommerce penetration leaps to 20.6% (classes sold almost entirely offline).

In “US Ecommerce Growth Jumps to More than 30%, Accelerating Online Shopping Shift by Nearly 2 Years”, it writes, “‘There will be some lasting impacts from the pandemic that will fundamentally change how people shop,’ said Cindy Liu, eMarketer senior forecasting analyst at Insider Intelligence. ‘For one, many stores, particularly department stores, may close permanently. Secondly, we believe consumer shopping behaviors will permanently change. Many consumers have either shopped online for the first time or shopped in new categories (i.e., groceries). Both the increase in new users and frequency of purchasing will have a lasting impact on retail’” ³.

Online commerce will be so high that this year, at $4,711 trillion, this will more than compensate for the 3,2 percent fall in brick and mortar expenses. Complete US retail revenue will also remain relatively flat.

5. Conclusion

More users are benefiting from the majority of online resources, including eCommerce, as internet penetration and connectivity improve. In everyday life e-commerce has become a mainstream, with fundamental advantages. The e-commerce market is projected to reverse double digit growth in net accounts from anywhere around the world. However, e-commerce can expand enormously as digital payment options are growing in these areas. About 22% of the world’s stores are now online. By 2021, e-Commerce online revenues are projected to hit $5 trillion.

Fru Kerik says, “The most popular eCommerce businesses worldwide are Amazon, Alibaba, eBay, and Walmart. These eCommerce giants have redefined the retail industry irrespective of location. They accumulate revenues that exceed billions of dollars yearly. As internet accessibility increases, these estimates would skyrocket. At the time of this writing, Amazon is present in 58 countries, Alibaba in 15, Walmart in 27, MercadoLibre in 18” ⁴.

E-Commerce firms have also contributed to the rise of e-Commerce through methodological findings. E-Commerce firms follow customer expectations and make important discoveries about the business-to-consumer model. These insights are then incorporated in market models, ensuring smooth future revenue increase globally.

6. References