Analysis Report Four - Trend Analysis and Value Creation

Executive Summary

This report examines how digital platform characteristics can either support or limit the success of modern business models. The analysis finds that strong network effects, connected networks, reduced multi-homing, and network bridging can help platforms attract users, collect valuable data, and expand into services such as e-commerce, subscriptions, advertising, and online marketplaces. However, platform growth can be weakened when customers use competing services, bypass the platform, or become dissatisfied with excessive restrictions. Apple’s closed ecosystem demonstrates this balance: its high switching costs and connected services strengthen customer loyalty and revenue, but its control over developers and transactions has also created regulatory and competitive concerns. These findings support the main reading’s argument that gaining scale is easier than sustaining a healthy and defensible platform network.

The Northwind analysis builds on the previous study of QUICK-Stop by using time-trend analysis to examine how the customer’s purchasing activity changes by year and month. This information can help an organization identify seasonal patterns, anticipate demand, improve inventory planning, and provide more targeted promotions or subscription-style ordering options. The report recommends that digital organizations use their transaction and customer-behavior data to predict future needs, personalize services, and create stronger relationships across their platforms. At the same time, organizations should balance platform control with customer and partner value so that efforts to protect revenue do not weaken the overall network.

Introduction

Digital platforms have become a major part of how organizations connect with customers, provide services, and earn revenue. In Why Some Platforms Thrive…and Others Don’t, Feng Zhu and Marco Iansiti explain that it may be easy for a digital platform to grow quickly, but it can be much harder to maintain that success. The authors identify five characteristics that can affect a platform’s long-term success: the strength of network effects, network clustering, the risk of disintermediation, vulnerability to multi-homing, and the ability to bridge different networks. Strong network effects can make a platform more valuable as more people use it. However, a platform may struggle when customers can easily use competitors, bypass the platform, or participate only within small local networks.

In The 7 Most Successful Business Models of the Digital Era, Bernard Marr discusses several ways digital organizations create revenue. These include advertising-supported services, e-commerce, freemium services, marketplaces and platforms, subscriptions, aggregator sites, and crowdfunding. Many successful companies use more than one of these models. For example, Amazon sells products directly, allows other sellers to use its marketplace, offers subscriptions through Amazon Prime, and earns advertising revenue. These business models work best when organizations attract users, collect useful data, and continue providing enough value to keep customers on their platforms.

Christopher Mims’s article about Apple shows how a strong platform can create both advantages and problems. Apple’s “walled garden” connects its devices, software, applications, and services. This makes it convenient for customers to stay within the Apple ecosystem, while making it more difficult for them to switch to another company. This helps Apple reduce multi-homing and earn revenue from subscriptions and App Store commissions. However, Apple’s control over developers, payments, and access to customers has also caused complaints, lawsuits, and regulatory concerns. Apple shows how platform control can protect revenue, but too much control can also hurt relationships with the people and businesses that help the platform succeed.

Together, these readings show that successful digital organizations must balance growth, control, customer value, and participation from business partners. They must also use the data created through platform activity to better understand customer needs and identify new opportunities. Building on my previous analysis of QUICK-Stop, this report uses Northwind time-trend analysis to examine how the customer’s purchasing activity changes over time. This analysis shows how an organization can use customer transaction data to plan inventory, prepare for changes in demand, create targeted promotions, and provide more personalized services.

Outside Research

The issues discussed in this week’s readings are also affecting organizations in the current business environment. Digital platforms may grow quickly because they connect large numbers of customers, sellers, developers, or service providers. However, the same network can become difficult to manage when participants use competing platforms, complete transactions outside the platform, or feel that the platform owner has too much control. This supports the main argument from Why Some Platforms Thrive…and Others Don’t: reaching a large scale does not guarantee long-term success because the platform must also maintain a healthy and defensible network.

One major struggle is multi-homing, which occurs when people use more than one platform for the same purpose. Yu et al. studied the behavior of ride-sourcing drivers using actual data from multiple platforms. Their research showed that drivers regularly switched among platforms based on factors that affected their ability to earn revenue and receive rides. This creates a challenge for companies such as Uber and Lyft because drivers and customers can easily compare prices, availability, and wait times. Multi-homing can benefit users by giving them more choices, but it makes it harder for a platform to build loyalty and maintain profitable prices. This research supports the assigned reading’s explanation that ride-sharing platforms are vulnerable because the cost of joining and using an additional platform is relatively low (yu2021multihoming?).

Another serious issue is disintermediation, where customers and service providers meet through a platform but later work together directly. Gu studied this problem using data from a large online freelance marketplace. The research used the temporary blocking of Skype in mainland China as a natural experiment and found that limiting access to an outside communication method reduced disintermediation by approximately 18%. This finding shows that communication technology can make it easier for buyers and sellers to bypass the platform, which may cause the platform to lose fees and other revenue. At the same time, placing too many restrictions on communication may make a marketplace inconvenient and cause users to choose a competitor. Organizations must therefore provide enough continuing value—such as secure payments, dispute resolution, reviews, insurance, or project-management tools—to give users a reason to remain on the platform (gu2024technology?).

Platform owners also struggle with how much control they should have over the businesses that depend on them. Wen and Zhu examined how Android app developers responded when Google threatened to enter the same markets as their applications. They found that affected developers reduced innovation and increased prices after the threat of Google’s entry became stronger. Once Google actually entered, those responses became even greater. Some developers shifted their innovation efforts toward different or new applications instead. This research relates closely to the Apple article because both Google and Apple operate platforms while also offering products and services that can compete with their own developers. Although this control may help the platform owner increase revenue or improve its services, it can also discourage outside developers from investing in the platform (wen2019platform?).

A similar problem can be seen with Amazon’s marketplace. Zhu and Liu studied Amazon’s entry into product areas that were originally served by third-party sellers. They found that Amazon was more likely to enter successful product areas with strong sales and positive reviews. Amazon’s entry increased customer demand and reduced shipping costs, which created value for customers. However, affected third-party sellers became less likely to continue pursuing growth on the platform. This demonstrates how a platform can use the sales and customer data generated by its network to identify profitable opportunities, while possibly weakening the sellers that helped create that information. If sellers believe the platform will use their data to compete against them, they may reduce their investments, leave the platform, or begin selling through several competing marketplaces (zhu2018complementors?).

Together, these studies show that platform organizations must balance value creation with fairness and trust. Multi-homing and disintermediation make it difficult to retain users, while too much platform control can damage relationships with sellers and developers. Data analytics can help organizations recognize customer needs, improve recommendations, and identify demand trends, but the information should be used in ways that strengthen the entire network rather than only benefiting the platform owner. The Northwind time-trend analysis of QUICK-Stop provides a smaller example of this process. By examining how the customer’s purchases change over time, an organization can prepare inventory, create timely promotions, and offer personalized ordering options. Modern platforms likely possess even more detailed data about searches, product views, transactions, reviews, and repeat purchases, giving them greater opportunities to create value.

Data Visualizations

Visualization One: QUICK-Stop’s Yearly Purchasing Trend

SELECT SUM(CAST(od.Quantity AS INTEGER)) AS total_quantity,
       strftime("%Y", Orders.OrderDate) AS year
FROM Orders
INNER JOIN "Order Details" od
ON Orders.OrderID = od.OrderID
WHERE Orders.CustomerID = "QUICK"
      AND year != "NA"
GROUP BY year
ORDER BY year

ggplot(data = myquery,
       aes(x = year, y = total_quantity, group = 1)) +
  geom_line() +
  labs(
    title = "QUICK-Stop Purchasing Trend by Year",
    x = "Year",
    y = "Total Quantity Purchased"
  )

Visualization Two: QUICK-Stop’s Monthly Purchasing Trend in 2013

SELECT SUM(CAST(od.Quantity AS INTEGER)) AS total_quantity,
       strftime("%m", Orders.OrderDate) AS month,
       strftime("%Y", Orders.OrderDate) AS year
FROM Orders
INNER JOIN "Order Details" od
ON Orders.OrderID = od.OrderID
WHERE Orders.CustomerID = "QUICK"
      AND year != "NA"
      AND year = "2013"
GROUP BY month
ORDER BY month

ggplot(data = myquery1,
       aes(x = month,
           y = total_quantity,
           group = 1)) +
  geom_line() +
  labs(
    title = "QUICK-Stop Monthly Purchasing Trend in 2013",
    x = "Month",
    y = "Total Quantity Purchased"
  )

Visualization Three: QUICK-Stop’s Monthly Purchases by Product Category in 2013

SELECT Categories.CategoryName,
       SUM(CAST(od.Quantity AS INTEGER)) AS total_quantity,
       strftime("%m", Orders.OrderDate) AS month,
       strftime("%Y", Orders.OrderDate) AS year
FROM Orders
INNER JOIN "Order Details" od
ON Orders.OrderID = od.OrderID
INNER JOIN Products
ON od.ProductID = Products.ProductID
INNER JOIN Categories
ON Products.CategoryID = Categories.CategoryID
WHERE Orders.CustomerID = "QUICK"
      AND year != "NA"
      AND year = "2013"
GROUP BY month, Categories.CategoryName
ORDER BY month

ggplot(data = myquery2,
       aes(x = month,
           y = total_quantity,
           group = CategoryName,
           color = CategoryName)) +
  geom_line() +
  labs(
    title = "QUICK-Stop Monthly Purchases by Category in 2013",
    x = "Month",
    y = "Total Quantity Purchased",
    color = "Product Category"
  )

Recommendations for Industry

Based on the research and the Northwind time-trend analysis, organizations should use customer data to identify changes in demand before they occur. By reviewing purchasing activity by year and then drilling down by month and product category, managers can recognize seasonal patterns and prepare inventory, staffing, shipping, and promotions ahead of time. For a customer such as QUICK-Stop, this could include increasing inventory before its busiest purchasing periods, sending promotions at the right time, or recommending products based on its previous ordering patterns.

Organizations should also use their data to provide customers with reasons to remain on the platform. Subscription services, loyalty rewards, automatic reordering, personalized recommendations, and faster delivery options can make the platform more valuable and reduce multi-homing. However, these services should offer a clear benefit rather than simply making it difficult for customers to leave. The Apple example shows that high switching costs can protect revenue, but excessive control can also create dissatisfaction among customers, developers, and regulators.

Executives should also focus on maintaining trust with the sellers, developers, service providers, and other businesses that contribute to the platform. Organizations may be tempted to use platform data only to increase their own profits or compete directly with their partners. Over time, this could cause participants to reduce their investment, use competing platforms, or leave the network. Clear policies, fair fees, secure payments, and transparent use of data can help protect these relationships.

Finally, organizations should look for opportunities to connect different services and business models. A marketplace could combine e-commerce with subscriptions, advertising, payment services, or automatic ordering. This type of network bridging can create additional value because information collected in one area can improve services in another. Executives should begin with a specific customer need shown in the data and then select the business model that best addresses that need.

Overall, the strongest recommendation is for organizations to use time-trend analysis as an ongoing decision-making tool rather than only reviewing past performance. Platforms already collect large amounts of customer and transaction data. When this information is used responsibly, it can help organizations predict demand, improve customer experiences, strengthen relationships within the network, and create new sources of value.

References

(article?){gu2024technology, title = {Technology and Disintermediation in Online Marketplaces}, author = {Gu, Grace}, journal = {Management Science}, year = {2024}, volume = {70}, number = {11}, pages = {7868–7891}, doi = {10.1287/mnsc.2021.02736} }

(article?){wen2019platform, title = {Threat of Platform-Owner Entry and Complementor Responses: Evidence from the Mobile App Market}, author = {Wen, Wen and Zhu, Feng}, journal = {Strategic Management Journal}, year = {2019}, volume = {40}, number = {9}, pages = {1336–1367}, doi = {10.1002/smj.3031} }

(article?){yu2021multihoming, title = {Exploring Multi-Homing Behavior of Ride-Sourcing Drivers via Real-World Multiple Platforms Data}, author = {Yu, Jingru and Mo, Dong and Xie, Ningke and Hu, Simon and Chen, Xiqun Michael}, journal = {Transportation Research Part F: Traffic Psychology and Behaviour}, year = {2021}, volume = {80}, pages = {61–78}, doi = {10.1016/j.trf.2021.03.017} }

(article?){zhu2018complementors, title = {Competing with Complementors: An Empirical Look at {Amazon.com}}, author = {Zhu, Feng and Liu, Qihong}, journal = {Strategic Management Journal}, year = {2018}, volume = {39}, number = {10}, pages = {2618–2642}, doi = {10.1002/smj.2932} }