Understanding Software Ecosystems

3,833 views

Published on

Published in: Business

Understanding Software Ecosystems

  1. 1. Visualizing Software Ecosystems Prof. Bala Iyer Twitter: @balaiyer 04/04/14
  2. 2. 2 Agenda  Ecosystem basics  Examples  Visualization Methodology  Lessons  Future directions
  3. 3. 3 Ecosystem  Loose networks – of suppliers, distributors, outsourcing firms, makers of related products or services, technology providers, and a host of other organizations – that affect, and are affected by the creation and delivery of a company’s own offerings. [Marco Iansiti]  An economic community supported by a foundation of interacting organizations and individuals—the organisms of the business world. The economic community produces goods and services of value to customers, who are themselves members of the ecosystem [James Moore]
  4. 4. 4 Creating Awesome Users  Don't make a better [X], make a better [user of X].  Kathy Sierra  Think about how customers experience the product or service
  5. 5. 5 Computer Industry [circa 1980s…] Source: Bill Gates Testimony - link Bill Gates, DOJ testimony 2002.
  6. 6. 6 Computer Industry circa 2002… ony - link Bill Gates, DOJ testimony 2002.
  7. 7. 7 Network -- 2002 Linux Unix Windows Other IBM MSFT SAP Environments companies 1144 Nodes 232 68 630 520
  8. 8. 8 Layered model of capabilities Hardware (consoles) Data Services (safety, entertainment, insurance) Connectivity (broadband, mobile) Operating SysStems Social or Community Layer Gaming Layer Apps
  9. 9. 9 Software Ecosystem  .. ―a set of actors functioning as a unit and interacting with a shared market for software and services, together with the relationships among them. These relationships are frequently underpinned by a common technological platform or market and operate through the exchange of information, resources and artifacts.‖
  10. 10. 10
  11. 11. 11 Networks  Deliver value to customer segments using a portfolio of capabilities some of which are achieved through links or dependencies.  People  Objects  Systems  Organizations
  12. 12. 12 Roles  Platform provider  Component supplier  Systems integrator  Orchestrator
  13. 13. 13 Industry Platform Definition  Technology or set of components (or services) that creates a common foundation,  That brings together multiple parties beyond a single firm (―market sides‖) for a common purpose and generates network effects,  Where the value can increase exponentially with (a) more users and (b) more ―complementary‖ products & services built around the platform Michael Cusumano, Platform Strategy Fundamentals 2014.
  14. 14. 14 Examples  Mobile Payment  Telematics  Cloud services
  15. 15. 15 Example 1
  16. 16. 16
  17. 17. 17 Google Wallet  Phones: Google Wallet will work with the Android & iOS  Cards: Citi MasterCard, a pre-loaded Google Prepaid Card (which can be ―refilled‖ from any source of funds) and gift cards from various participating merchants.  Markets: available in San Francisco and New York  Participating merchants: American Eagle Outfitters, Bloomingdale’s, Champs Sports, The Container Store, Duane Reade, Einstein Bros. Bagels, Foot Locker, Guess, Jamba Juice, Macy’s, Noah’s Bagels, Peet’s Coffee & Tea, RadioShack, Subway, Toys―R‖Us and Walgreens.  Google Wallet relies on the MasterCard PayPass infrastructure it will work ―at more than 124,000 PayPass- enabled merchants nationally and 311,000 globally.‖ 17
  18. 18. 18 Google Wallet  Google Offers linked to Wallet: Offers are directly integrated into Google Wallet. Consumers will be able to send offers they encounter to Wallet with a single click (if you’re signed in). Redemption will be accomplished with a ―SingleTap‖ at the point of sale or via scanning or keying in a code if the POS system doesn’t support SingleTap.  Future uses: Google envisions that ―receipts, boarding passes and tickets will all be seamlessly synced to your Google Wallet.‖ 18
  19. 19. 1919
  20. 20. 2020Link
  21. 21. 2121
  22. 22. 22 M-payment Stakeholders  Platform provider  Network Operators  Banks  Credit Card Issuers  Device Manufacturers  Merchants  Consumers 22
  23. 23. 23 Market  Forrester Research estimates the size of the opportunity will grow to $90 billion by 2017, growing at a 48% rate since 2012.  Worldwide mobile payment transaction values will surpass $171.5 billion ... according to Gartner, Inc.  By 2015, about half of the 863 million phone handsets sold yearly worldwide will contain the chips, according to Frost & Sullivan.  150,000 merchants on Google Wallet  ―only about 7 percent of retail sales occur online, compared to $4 trillion that is still spent in person at stores. "We're going after the 93 percent," Google VP, Osama Bedier. 23
  24. 24. 2424
  25. 25. 25 Platforms 25
  26. 26. 26 Method used for Ecosystem generation  List of platform players created from virtual currency platforms, experts, and  Created a list of partners from news feeds and websites  Visualized using Pajek
  27. 27. 27 Mobile Payment -- complementors -- platforms -- partnerships
  28. 28. 28 Analysis  237 companies, 25 platforms  Many approaches to M-payment  Diversity is good  Many fragmented platforms  Carriers and card issuers are working across platforms  Still evolving ecosystem  Many experiments in emerging economies
  29. 29. 29 M-payment Platform Wars  A few platforms may emerge  Unique value proposition  Too expensive to carry multiple systems  Direct (developers) and Indirect (end users) Network Effects  Business models and subsidies  Emergence of standards like OpenSocial 29
  30. 30. 30 Product vs. Platform  Winners in a platform market generally have the ―best‖ platform, not the ―best‖ product!  Best products? Hard to define & while good, usually not enough to win or dominate in a platform market  Best platforms? = (1) Open access & interfaces (but not too open). (2) Modular architecture (easy to build on, extend). (3) Most compelling complements (usually result of most vibrant ecosystem). Michael Cusumano, Platform Strategy Fundamentals 2014.
  31. 31. 31 Example #2: Cloud computing
  32. 32. 32 Methodology  Start with a focal set of firms (55 – 631)  Determined dependencies  Identified core  Capture definitions of cloud from the web (~70)  Read descriptions  Run through a tag cloud analysis  Identified capabilities
  33. 33. 33 Infrastructure Platform Applications Collaboration Services Cloud Ecosystem (partnerships)
  34. 34. 34 Core players
  35. 35. 35 Example #3: Telematics
  36. 36. 36
  37. 37. 37 Telematics Ecosystem
  38. 38. 38 January 2007 Mashups
  39. 39. 39 2002 Highest High Medium Low Higher Clustering coefficient Central firms: Between 1990 and 2002, these firms account for 35 to 69% of the industry sale
  40. 40. 40 Illustrative list of software firms in the cluster • NEC • Computer Associates • RealNetworks • CheckFree Corp. • Yahoo! Link strength High (5 or more links) Medium (3-4 links) Low (1-2 links) The size of each node is proportional to the number of alliances
  41. 41. 41 Platform moves • Open sourcing • Cutting off air supply • Entrants with network effects • Creating and defending IP • Acquisitions • Creating exclusive links • Complementor incentives and subsidies • Envelopment (absorb & bundle) • Facebook with social layer and FB credits
  42. 42. 42 Coring & Tipping Strategic Option Possible Technology Actions Possible Business Actions CORING (How create a new platform when none existed before) • Solve “system” problem • Facilitate add-ons • Keep key IP closed or “open but not open” • Strong interdependencies -- platform & complements • Solve “business” problem • Create complementor incentives to innovate • Protect revenue & profit • Raise switching or multi- homing costs TIPPING (How win a platform battle when multiple platforms compete) • Develop compelling features • Absorb & bundle from adjacent markets (―envelopment‖) • Complementor incentives • Coalitions of also-rans • Pricing or other subsidies to attract users/complementors Source: Gawer and Cusumano SMR 2008
  43. 43. 43 Findings  Chart your ecosystem  Ford (telematics, healthcare, entertainment, ..)  Consider shifts in network  Own and third-party moves  Portfolio of proactive and reactive connections  Ecosystem moves
  44. 44. 44 Methodology: EcoSysNetworks™  Determine industry structure or stack  Identify companies and attributes  Get dependency information and attributes  Enter information into database  Determine semantics for firms (size, shape, color)  Determine semantics for links (thickness, color)  Create input file for visualization  Visualize and interpret
  45. 45. 45 Stacks Video Games SmartOS Content providers Software developers Software publishers Platform provider Retailer Consumer Network operator Handset manufacturer Mobile OS provider Content providers and aggregators Application developers
  46. 46. 46 Identify companies and attributes  Sources  Competitive list from industry publications  News feeds  User generated input (bookmarks or brainstorming)  Inputs  Name  Revenue  Business type (stacks)  Platform provider?
  47. 47. 47 Get alliance information and attributes  Sources  Company websites  News feeds  Inputs  Alliance type (technical, marketing, strategic or financial)  Single or multiple alliances
  48. 48. 48 Enter information into database  The database has two tables  Firms  Relationships Visualization semantics can also be stored in the database
  49. 49. 49 Semantics for firms  Size  Currently we use revenue  Shape  Platform players are denoted as diamonds, rest as circles  Color  Based on stack layer Access Light Magenta Advertiser Green Content Lavender Hardware Tan Default Red Operator Melon Services Blue Investors Yellow
  50. 50. 50 Semantics for links  Thickness  Based on repeat links  Color  Based on relationship type  Technical  Marketing  Financial  Strategic
  51. 51. 51 Pajek  Download: http://pajek.imfm.si/doku.php?id=down load  Reference manual
  52. 52. 52 Create input file for visualization  We use Pajek for the visualizations  A sample Pajek input file has two parts  Vertices  A vertex record contains vertex number, name, color, shape, size and the time it appears on the visualization.  Edges  An edge record contains from, to, thickness, color and the time it appears in the visualization
  53. 53. 53 Data entry  Open Notepad  Enter data  Save as .net file  Use Excel  Save as Tab delimited DOS file  Replace tabs with spaces using Notepad
  54. 54. 54 *Vertices 8 1 "Google" ellipse ic Green x_fact 3 y_fact 4 2 ―Zynga" triangle ic Lavender x_fact 2 y_fact 2 3 ―Facebook" box ic Lavender x_fact 2 y_fact 2 4 "Accel" diamond ic Yellow x_fact 3 y_fact 3 5 "Accenture" circle ic Blue x_fact 2 y_fact 2 6 "IBM" box ic Orange bc Black 7 "MSFT" circle ic Cyan bc Magenta 8 "Intel" cross ic Purple bc Pink x_fact 3 y_fact 3 *Edges 1 2 1 c Tan 2 3 3 c Green 3 4 3 c Yellow 4 5 1 c Lavender 5 4 1 c Blue 5 6 2 c Gray 1 7 2 c Red 2 8 2 c Purple 3 5 2 c Brown 4 1 2 c Black Data file
  55. 55. 55
  56. 56. 56 Visualization steps  Launch program  Load data into Pajek  Use the draw option  Chose Layout/Energy/Fructerman- Reingold/2D
  57. 57. 57 Ecosystem Health
  58. 58. 58 Network Measures
  59. 59. 59 Degree centrality  The degree centrality of a node is defined as the total number of connections the node has
  60. 60. 60 Density  Density is the number of connections a node has, divided by the number of possible connections
  61. 61. 61 Betweenness  It is important to find which units lie on the shortest paths among pairs of other units. Such units have control over the flow of information in the network. Idea of betweenness centrality measures: unit is central, if it lies on several shortest paths among other pairs of units.
  62. 62. 62 Closeness centrality  Closeness is preferred in network analysis to mean shortest-path length, as it gives higher values to more central vertices, and so is usually positively associated with other measures such as degree. Source: wikipedia
  63. 63. 63 Clustering Coefficient  A node’s clustering coefficient can be defined as the proportion of alters that are themselves directly connected
  64. 64. 64 Pajek measures  Network/Create Vector/Centrality/Degree/, while other two centralities can be found under Degree  Network reduction  Net/Transform/Reduction/Degree  Core players  Net/Partitions/Core/Degree  Operations/Extract Network/Partitions  Draw-partition
  65. 65. 65 Exporting pictures  Export/2D/bitmap  Save file  Import into PowerPoint
  66. 66. 66 What did we find in the software sector (1990-2001)?  Small worlds  Degree – 3.1 to 2.8  Density – decreased from 0.005 to 0.001  CC – 0.21 to 0.28  Structure of the network and firm performance
  67. 67. 67 Platform Leadership Process  Create & communicate a vision of platform evolution  Build consensus & coalitions among a small group of influential firms for new initiatives  Identify and target “system” bottlenecks  Distribute tools & enabling technologies to help partners develop complements fitting the vision  Highlight business opportunities and help leading firms act as “rabbits” to lead the market  Facilitate multi-firm initiatives to promote standards, platform interfaces, and complementary innovations Michael Cusumano, Platform Strategy Fundamentals 2014.
  68. 68. 68 Data sources  Website  Lexis/Nexis  Sort search results by subject (alliances and partnerships)  SDC Platinum  Programmableweb.com  http://www.programmableweb.com/neoapi.xml  http://www.programmableweb.com/neomatrix.xml  Appdata  www.compete.com ; www.socialmention.com; www.google.com/trends
  69. 69. 69 Future directions  Decision Environments  automate data collection  Visualize  collaborate  Shared repositories  Strong theory and cases  Better understanding of ecosystem risks  API-based networks  Communities to share findings

×