W11 - Competitive Strategy Lecture Notes PDF

IS4242 INTELLIGENT SYSTEMS & TECHNIQUES L11 – Competitive Strategy Aditya Karanam © Copyright National University of Singapore. All Rights Re...

IS4242 INTELLIGENT SYSTEMS & TECHNIQUES L11 – Competitive Strategy Aditya Karanam © Copyright National University of Singapore. All Rights Reserved.  Announcements ▸ Project Check-in ‣ Make use of all the resources: Tutorials, consultation sessions, etc. ▸ Final Exam Syllabus ‣ Weeks 8, 9, 10 and 11 ‣ Week – 12 is not included IS4242 (Aditya Karanam) 2  In this Class… ▸ How do we pursue a sustainable competitive advantage? ‣ Value chain view ‣ Porter’s Five Forces ▸ Analyzing competition using tweets ‣ Graph Analytics: Louvain Community Detection IS4242 (Aditya Karanam) 3  Sustainable Competitive Advantage ▸ Firms strive for sustainable competitive advantage – financial performance that consistently outperforms their industry peers ▸ The world is so dynamic: New products and new competitors arise overnight, making it impossible to obtain a truly sustainable advantage ‣ New competitors or copycats may cut costs, cut prices, or increase features that may benefit consumers but erode profits industry-wide ▸ This balance is more difficult when competition involves technology IS4242 (Aditya Karanam) 4  Contentious and Confusing Messages! IS4242 (Aditya Karanam) 5  It Isn’t All Despair! ▸ But there are winners – big, consistent winners – empowered through their use of technology ‣ Tech companies are driving disruption in almost all industries ▸ How do they do it? Is it the technology? ‣ If yes, what is so special about technology that drives competitive advantage? ‣ Operational effectiveness? or something else? IS4242 (Aditya Karanam) 6  Technology for Operational Effectiveness ▸ Technology provides operational effectiveness – the same tasks can be performed better than rivals. ▸ The danger in operational effectiveness is “sameness” ‣ Tech can be copied quickly, and followers can be fast. ‣ An innovative feature of a firm was matched by at least one of its three major rivals in one and half months (Gallaugher & Downing, 2000) ▸ Operational effectiveness alone is not sufficient to yield sustainable dominance over the competitors IS4242 (Aditya Karanam) 7  Technology for Strategic Positioning ▸ Use technology for strategic positioning ‣ Performing different activities from those of rivals or the same activities in a different way ▸ While the technology itself is often very easy to replicate ‣ Novel business approaches enabled by technology that are different from rivals can be difficult for others to copy IS4242 (Aditya Karanam) 8  Example: FreshDirect New York City–based grocery firm IS4242 (Aditya Karanam) 9  Example: FreshDirect ▸ The two most pressing problems for grocery shoppers in NYC: ‣ Limited selection and high prices ‣ Both problems are due to the high cost of real estate in New York ▸ FreshDirect uses technology to craft an efficient solution ‣ The firm’s “storefront” is a website offering a fresh goods selection that’s over five times larger than local supermarkets ‣ Same-day and Next-day deliveries are from a vast warehouse IS4242 (Aditya Karanam) 10  Example: FreshDirect ▸ Efficiently manages inventory and reduces wastes ▸ Buys directly from suppliers, eliminating middlemen wherever possible. ▸ Products are offered at prices that can undercut the competition by as much as 35 percent ▸ Got so popular that apartment buildings began to redesign common areas to include secure freezers that can accept FreshDirect deliveries! IS4242 (Aditya Karanam) 11  Example: FreshDirect ▸ More than technology, it’s the collective impact of the firm’s differences when compared to rivals that delivered success. ‣ Delivery business: Difficult to fully copy as this would leave traditional grocery stores to straddle two markets ‣ Warehouses: High entry costs for would-be competitors are high ‣ Customer Data: Further help refine its recommendations ▸ FreshDirect has built an imitation resistant value chain using technology! IS4242 (Aditya Karanam) 12  The Value Chain ▸ Value chain: a set of activities through which a product or service is created and delivered to customers ▸ Five primary components: ‣ Inbound logistics: obtaining materials and other inputs into the firm from suppliers ‣ Operations: turning inputs into products or services ‣ Outbound logistics: delivering products or services to consumers, distribution centres, retailers, or other partners ‣ Marketing and sales: customer engagement, pricing, promotion, and transaction ‣ Support: service, maintenance, and customer support IS4242 (Aditya Karanam) 13  Example: Amazon ▸ Inbound Logistics: Amazon sources products from various suppliers globally and manages warehousing efficiently to fulfill orders quickly and accurately ▸ Operations: Uses technology and automation to optimize its operations, from inventory management to order processing ▸ Outbound Logistics: Offers various shipping options, including Prime two-day delivery, same- day delivery, and more. IS4242 (Aditya Karanam) 14  Example: Amazon ▸ Marketing and Sales: Amazon offers advertising services to sellers, allowing them to promote their products within the Amazon ecosystem. ▸ Services: Amazon offers customer services by allowing seamless return of products and money-back policies IS4242 (Aditya Karanam) 15  Porter’s Five Forces ▸ Value chain provides understanding with a focus on the activities of the firm ‣ How about new entrants or substitute products? ▸ Porter’s five forces: Frameworks for examining a firm’s competitive environment by looking at the all players in the industry ▸ Helps in understanding: ‣ Underlying drivers of average industry profitability ‣ How profitability will evolve in the future IS4242 (Aditya Karanam) 16  Analyzing the Industry: Porter’s Five Forces Framework ▸ The five forces of any industry are: 1. Intensity of rivalry among existing competitors 2. The threat of new entrants 3. The threat of substitute goods or services 4. Bargaining power of buyers 5. Bargaining power of suppliers IS4242 (Aditya Karanam) 17  Example - 1: Music Industry ▸ Consider how the rise of the Internet has impacted the five forces for music retailers. ▸ Rivalry among existing competitors: ‣ Traditional music retailers (e.g., Tower, Virgin, etc.) found that customers are buying music online. ‣ Invested in the new channel out of necessity ‣ compete based on the geography of where brick-and-mortar stores are physically located, and online as well ▸ Threat of new entrants: ‣ Low barrier to entry: New firms can sell their music through Amazon ‣ Free from brick-and-mortar stores, Amazon, provides a highly scalable cost structure. ‣ The online buying experience is superior to the in-store experience ‣ Customers can listen to all tracks; selection is seemingly limitless: the long tail phenomenon IS4242 (Aditya Karanam) 18  Example - 1: Music Industry ▸ Substitutes: ‣ Digital tracks purchased online were almost always inferior to their CD counterparts. ‣ The tech-based market changed the listening habits of consumers ‣ The convenience of carrying thousands of songs trumps a slight quality degradation. ‣ iTunes sold more music than any other firm: online or offline. ‣ The Internet has also boosted illegal music “sharing” services ▸ Bargaining power of consumers: ‣ Lots of choice gives high bargaining power to consumers. ‣ They demand lower prices and greater convenience. ▸ Bargaining power of suppliers: ‣ At the start of the Internet revolution, retailers could pressure labels to limit sales through competing channels. ‣ Now, bands have alternative ways that can entirely bypass the traditional music labels ‣ The bargaining power of the music labels and artists—also has increased. IS4242 (Aditya Karanam) 19  Example - 2: Automotive Industry ▸ Consider how the introduction of electric cars has impacted the five forces of the automotive industry ▸ Rivalry among existing competitors: ‣ Intensified Competition: The introduction of electric cars has intensified competition, with both traditional automakers and new entrants vying for market share. ▸ Threat of New Entrants: ‣ High Barrier to Entry: Developing electric vehicles (EVs) requires substantial investment in R&D. Established automakers have an advantage due to their resources and economies of scale. IS4242 (Aditya Karanam) 20  Example - 2: Automotive Industry ▸ Substitute Products: ‣ Hybrid and alternative fuel technologies are substitutes for pure electric vehicles. These substitutes can become more viable and competitive. ▸ Bargaining power of Consumers: ‣ Increasing Options: Buyers can choose from a variety of models and brands, which can increase their bargaining power. ▸ Bargaining power of suppliers: ‣ Batteries are a crucial component of electric vehicles, and suppliers of lithium and battery technology have gained significant leverage. ‣ However, automakers are also investing in battery technology and exploring alternative sources. IS4242 (Aditya Karanam) 21  How to Conduct Such Analysis Efficiently? ▸ In this day and age, the boundaries of the firms have become more fluid; the first and most difficult part is to identify competitors ‣ New innovations, international govt policies, etc., can have a profound impact on the competitive advantage of the firms ▸ In social media, consumers compare firms that serve the same needs or have similar features ‣ Ex: In tweets on stocks, users compare companies that are impacted by the changes (new entrants, govt policy, innovation, etc.) in the industry IS4242 (Aditya Karanam) 22  Analyzing the competition structure from online word-of- mouth © Copyright National University of Singapore. All Rights Reserved. 23  Data ▸ Data from: ‣ StockTwits: microblogging website on stocks ‣ Tweets on S&P 500 companies ‣ Show you how to extract data in the tutorial ▸ Examples: IS4242 (Aditya Karanam) 24  Task ▸ Generate network data of stocks based on their mentions in the tweets ▸ Identify closely connected communities of stocks within the network ‣ These communities represent a set of competitors ▸ Unsupervised task ‣ Graph or Network Analysis: Community Detection IS4242 (Aditya Karanam) 25  Networks: Examples ▸ Sociology ‣ Friends' networks ▸ Medicine ‣ Infectious disease transmission networks ▸ Technology ‣ Computer networks, power grids ▸ Business: Strategic Alliance Networks IS4242 (Aditya Karanam) 26  Network analysis ▸ Processes data structured as graphs ▸ Study properties of the network ‣ Central nodes ‣ Highly connected components ‣ Shape ‣… IS4242 (Aditya Karanam) 27  Graph ▸ Mathematical term for a network ‣ Used interchangeably ‣ Defined as a set of nodes with connecting links called edges ▸ Node: entity of interest (e.g., person, firms, etc) ▸ Edge: relationship (e.g., friendship, comparison, etc.) IS4242 (Aditya Karanam) 28  Types of Graphs ▸ Directed: Has direction (asymmetry) ‣ E.g.: Person A knows Person B, but Person B may not know Person A ‣ E.g.: Twitter ▸ Undirected No direction (symmetry) ‣ E.g.: Person A is a friend of person B ‣ E.g.: Facebook ▸ Edge may be weighted or unweighted ‣ E.g.: the number of interactions between two users in social media IS4242 (Aditya Karanam) 29  Graph Representation ▸ Represented as adjacency matrix or list of pair of nodes and their attributes ▸ Adjacency matrix (A): 𝐴𝑖𝑗 : represents the edge weight between 𝑖 and 𝑗 ▸ Example: 0 1 1 0 0 Node1 Node2 Weight 1 0 0 1 1 A B 1 1 0 0 0 0 A C 1 0 1 0 0 1 … 0 1 0 1 0 IS4242 (Aditya Karanam) 30  Degree ▸ Number of edges connected to a node ▸ Directed graph ‣ In-degree: #incoming edges ‣ Out-degree: #out-going edges ▸ We may associate weights with each edge in the graph IS4242 (Aditya Karanam) 31  Path ▸ A path in a graph is a sequence of nodes connected by edges ▸ The shortest path between nodes A and B: The path between A and B with minimum: ‣ Number of edges (unweighted graph) ‣ Sum of weights of edges in the path (weighted graph) ▸ What is the shortest path between A and E? IS4242 (Aditya Karanam) 32  Connected Components ▸ A graph is connected if for every pair of nodes, there is a path between them ▸ A graph is fully connected if each pair of nodes is connected by an edge ‣ Also called complete graph Fully connected graph IS4242 (Aditya Karanam) 33  Subgraphs and Communities ▸ A subgraph is a subset of nodes and edges of a graph. ‣ The node subset must include all endpoints of the edge subset. ▸ Communities: Subgraphs with non-overlapping sets of nodes with lots of internal connections and few external ones (to the rest of the network). Also known as modules, clusters, etc. IS4242 (Aditya Karanam) 34  Communities ▸ Community structure is a prominent feature of networks ▸ A particular network may have multiple communities such that nodes inside a community are densely connected ▸ E.g.: In your Facebook network, you may have different communities – school friends, college friends, friends from your hometown, etc. Facebook Network IS4242 (Aditya Karanam) 35  Modularity ▸ Modularity: quantifies how well a network can be partitioned into groups compared to a random network ▸ Community: sets of nodes with lots of internal connections and few external ones (to the rest of the network). ▸ Modularity (Q) of partitioning S of graph G: 𝑄 ∝෍ [# 𝑒𝑑𝑔𝑒𝑠 𝑤𝑖𝑡ℎ 𝑖𝑛 𝑔𝑟𝑜𝑢𝑝 𝑠 − 𝐸𝑥𝑝𝑒𝑐𝑡𝑒𝑑 # 𝑒𝑑𝑔𝑒𝑠 𝑤𝑖𝑡ℎ 𝑖𝑛 𝑔𝑟𝑜𝑢𝑝 𝑠 ] 𝑠∈𝑆 IS4242 (Aditya Karanam) 36  Modularity ▸ Modularity (Q) of partitioning S of graph G: 𝑄 ∝෍ [# 𝑒𝑑𝑔𝑒𝑠 𝑤𝑖𝑡ℎ 𝑖𝑛 𝑔𝑟𝑜𝑢𝑝 𝑠 − 𝐸𝑥𝑝𝑒𝑐𝑡𝑒𝑑 # 𝑒𝑑𝑔𝑒𝑠 𝑤𝑖𝑡ℎ 𝑖𝑛 𝑔𝑟𝑜𝑢𝑝 𝑠 ] 𝑠∈𝑆 1 𝑘𝑖 𝑘𝑗 𝑄= ෍ ෍ ෍(𝐴𝑖𝑗 − ) 2𝑚 𝑠∈𝑆 2𝑚 𝑖∈𝑠 𝑗∈𝑠 𝑘𝑖 , 𝑘𝑗 : degree of nodes 𝑖 and 𝑗, respectively 𝐴𝑖𝑗 : edge weight between 𝑖 and 𝑗 2𝑚: sum of all the edge weights in the graph 𝑘𝑗 Probability of one edge stub match: 𝑝 𝑖 → 𝑗 = 2𝑚 𝑘𝑖 There are 𝑘𝑖 edge stubs 𝑘𝑖 𝑘𝑗 Expected number of edges between nodes 𝑖 and 𝑗: 2𝑚 𝑘𝑗 Edge stubs IS4242 (Aditya Karanam) 37  Modularity ▸ We can identify communities by maximizing modularity value ▸ Modularity values take range of [-1, 1] ‣ Positive if the number of edges in a group exceeds the expected number ‣ If Q value lies in the range of 0.3 to 0.7, implies significant community structure IS4242 (Aditya Karanam) 38  Community Detection Methods ▸ There are two types of Community Detection methods: ‣ Agglomerative Methods ‣ Each node belongs to its own community, and they are merged iteratively by maximizing an objective function ‣ Ex.: Louvain Community Detection ‣ Divisive Methods: ‣ Edges are removed one by one from a complete graph iteratively by maximizing an objective function ‣ Ex: Girvann-Newman algorithm IS4242 (Aditya Karanam) 39  Louvain Community Detection ▸ Provides hierarchical communities ▸ Widely utilized to study large networks because: ‣ Fast and High modularity output (i.e., “better communities”) ▸ Communities are formed in two phases: ‣ Phase 1: Local moving of nodes ‣ Phase 2: Aggregation of the network IS4242 (Aditya Karanam) 40  Louvain Algorithm ▸ Greedy Algorithm. ▸ Starts by considering each node is in its own community ▸ Each iteration (or pass) has 2 phases: ‣ Phase 1: Modularity is optimized by allowing local changes to node- communities memberships ‣ Phase 2: The identified communities are aggregated into super-nodes to build a new network ‣ Go to Phase 1 ▸ Repeated until no increase in modularity is possible IS4242 (Aditya Karanam) 41  Louvain Algorithm: Illustration IS4242 (Aditya Karanam) 42  Application: Identifying Competition ▸ Network of companies ‣ Nodes: Stocks ‣ Edge: Weighted and undirected ‣ Weight is the probability of two stocks occurring in a tweet calculated based on co-occurrences ‣ Number of times stocks appear together in a single tweet 𝐶o𝑜ccurrence (𝐴,𝐵) ‣ Weight = 𝑂𝑐𝑐𝑢𝑟𝑟𝑒𝑛𝑐𝑒 𝐴 × 𝑂𝑐𝑐𝑢𝑟𝑟𝑒𝑛𝑐𝑒(𝐵) ▸ Run Louvain Community detection using networkx library in Python ‣ Compare the sectors within the community with sectors provided by the companies it their SEC filings IS4242 (Aditya Karanam) 43  Final Result Firm S&P Sector Community (Majority sector) Consumer Discretionary Tesla, Inc. Information Technology (Automobiles) PayPal Financials Information Technology Netflix Communication Services Information Technology Amazon Consumer Discretionary (retail) Information Technology Meta Platforms Communication Services Information Technology Alphabet Inc. Communication Services Information Technology Firms’ sector in their filings is quite different from how consumers view them! Also, represents the fluidity of the boundaries that is not yet captured by the SEC Filings IS4242 (Aditya Karanam) 44  References ▸ Competitive Strategy: ‣ https://open.lib.umn.edu/informationsystems/ ▸ Community Detection: ‣ https://web.stanford.edu/class/cs246/slides/11-graphs1.pdf IS4242 (Aditya Karanam) 45  Thank You © Copyright National University of Singapore. All Rights Reserved.