Investing In Web Services

Investing in Web Services

Startup, End User and Investor Perspectives

Simeon SimeonovSimeon Simeonov

Polaris Venture PartnersPolaris Venture Partners

[email protected]@polarisventures.com

2

Agenda

► IntroductionsIntroductions► Web services status quoWeb services status quo► Market analysisMarket analysis► Case studiesCase studies► Lessons learnedLessons learned

3

Polaris Venture Partners

► Experienced teamExperienced team 12 partners, 22 investment professionals12 partners, 22 investment professionals

► $2B under management, current $900M fund$2B under management, current $900M fund Highly diversified (2/3 IT, 1/3 life sciences)Highly diversified (2/3 IT, 1/3 life sciences) Allaire, Akamai, Alnylam, NanosysAllaire, Akamai, Alnylam, Nanosys

► Investment focusInvestment focus Great teams with great ideas, regardless of stageGreat teams with great ideas, regardless of stage Most active early stage VC on the East Coast in 2002Most active early stage VC on the East Coast in 2002

► Web services investment experienceWeb services investment experience Evaluated several dozen Web services startupsEvaluated several dozen Web services startups Invested in Invested in Service IntegrityService Integrity

4

Sim Simeonov

► Joined Polaris in 2002, IT focusJoined Polaris in 2002, IT focus Chief Architect + VP Emerging Technologies, MacromediaChief Architect + VP Emerging Technologies, Macromedia Part of founding team + Chief Architect, AllairePart of founding team + Chief Architect, Allaire

► Deep Web services expertiseDeep Web services expertise Web Distributed Data Exchange (WDDX) back in 1998Web Distributed Data Exchange (WDDX) back in 1998 Led Web services team at Macromedia (Axis architects)Led Web services team at Macromedia (Axis architects) Standards work for W3C, OASIS, JCPStandards work for W3C, OASIS, JCP Award-winning book: “Building Web Services”Award-winning book: “Building Web Services” Editorial board member of Web Services JournalEditorial board member of Web Services Journal Regular contributor to XML JournalRegular contributor to XML Journal

► In touch with startups, big vendors and end usersIn touch with startups, big vendors and end users

5

Agenda

► IntroductionsIntroductions► Web services status quoWeb services status quo► Market analysisMarket analysis► Case studiesCase studies► Lessons learnedLessons learned

6

What Are Web Services?

► Many academic definitions are not helpfulMany academic definitions are not helpful Web services enable app-to-app communication in a Web services enable app-to-app communication in a

manner that is simpler, more open and more extensiblemanner that is simpler, more open and more extensible So, what is a Web services startup anyway?So, what is a Web services startup anyway?

► NOT a technology revolutionNOT a technology revolution The next step in the evolution of distributed computingThe next step in the evolution of distributed computing For some advanced uses, it is a step For some advanced uses, it is a step backwardsbackwards!!

► NOT a business model revolutionNOT a business model revolution ……at least not for a long whileat least not for a long while Web services are just a means to an endWeb services are just a means to an end

► Revolution in industry dynamicsRevolution in industry dynamics For the first time: distributed standards evolutionFor the first time: distributed standards evolution

7

Distributed Standards Evolution

► Standards modelsStandards models Authoritarian: COM/DCOMAuthoritarian: COM/DCOM Design-by-consensus: CORBA, IETF-related standardsDesign-by-consensus: CORBA, IETF-related standards Distributed & composableDistributed & composable: Web services: Web services

► Standards bodiesStandards bodies W3C: core specificationsW3C: core specifications OASIS: higher-level specifications (applications + facets)OASIS: higher-level specifications (applications + facets) WS-I: interoperability specifications + use casesWS-I: interoperability specifications + use cases

► The truth about the standards “democracy”The truth about the standards “democracy” IBM + MS rule IBM + MS rule (and that’s OK)(and that’s OK) Startups have little influenceStartups have little influence

8

Web Services Architecture

MessagingMessaging SOAP, WS-AddressingSOAP, WS-Addressing

DescriptionDescription WSDL, WS-PolicyWSDL, WS-Policy

DiscoveryDiscovery UDDI, WS-MetadataExchangeUDDI, WS-MetadataExchange

SecuritySecurity WS-Security, WS-Trust, WS-Security, WS-Trust, WS-SecureConversation, WS-FederationWS-SecureConversation, WS-Federation

Reliable MessagingReliable Messaging WS-Reliable MessagingWS-Reliable Messaging

TransactionsTransactions WS-Coordination, WS-AtomicTransaction, WS-Coordination, WS-AtomicTransaction, WS-BusinessActivity WS-BusinessActivity

Service CompositionService Composition BPEL4WSBPEL4WS

9

Driving Forces

► Supply sideSupply side Platform vendors want app servers on every nodePlatform vendors want app servers on every node Platform vendors going after the $40B integration marketPlatform vendors going after the $40B integration market MS going after the business applications market for SMBsMS going after the business applications market for SMBs Long term: networking vendors will want to play, tooLong term: networking vendors will want to play, too

► Demand sideDemand side Agile business (both revenue and cost focused)Agile business (both revenue and cost focused)

► Internally: rationalized infrastructure and easier integrationInternally: rationalized infrastructure and easier integration► Externally: syndication, outsourcing, multi-step value chainsExternally: syndication, outsourcing, multi-step value chains

Multiple presentation modesMultiple presentation modes► Flexible applications: HTML / smart client / Win GUIFlexible applications: HTML / smart client / Win GUI► Mobile Web: 3G / PDA Mobile Web: 3G / PDA

10

Web Services Now

► Early adopter stage; have not crossed the chasmEarly adopter stage; have not crossed the chasm Lots of experimentation with increasing deploymentsLots of experimentation with increasing deployments Mostly point-to-point internal integrationMostly point-to-point internal integration B2B is still done over private networksB2B is still done over private networks Some CIOs are starting to demand Web servicesSome CIOs are starting to demand Web services

► Focused demandFocused demand Legacy extensionLegacy extension Integration/workflowIntegration/workflow SecuritySecurity Management/monitoringManagement/monitoring

Better luckBetter lucknext year?next year?

11

The Current StatisticsCurrent Adoption

0% 10% 20% 30% 40% 50% 60% 70%

Live

Pilots

Experimentation

Investigation

Importance to End Users and ISVs / ASPs

0% 10% 20% 30% 40% 50%

Very Important

Important

Neutral

Unimportant

VeryUnimportant

End-User

ISV / ASP

Key Benefits

0% 10% 20% 30% 40% 50%

Differentiation

Efficient Business

Increase Revenues

Tighter Partnerships

Time to Market

Reduce Costs

Key Risks

0% 20% 40% 60% 80%

Unclear Biz Benefits

Immature Technology

Scalability

Reliability

Security

Immature Standards

SIIA Study of 779 ProfessionalsSummer 2002

12

Web Services Futures

► MainstreamMainstream Mainly used for integration and syndication (EAI + B2B)Mainly used for integration and syndication (EAI + B2B) Significant focus on SLAs and management/monitoringSignificant focus on SLAs and management/monitoring

► On the edgeOn the edge Service composability and application assemblyService composability and application assembly High-level virtualization (pragmatic utility computing)High-level virtualization (pragmatic utility computing) Focus on semantics: key for more dynamic integrationFocus on semantics: key for more dynamic integration

► Still far outStill far out Fully dynamic e-business (auto-discovery/use of services)Fully dynamic e-business (auto-discovery/use of services) Hindered not by technology but by business issuesHindered not by technology but by business issues

► Trust, liability, business quality of service, …Trust, liability, business quality of service, …

13

Agenda

► IntroductionsIntroductions► Web services status quoWeb services status quo► Market analysisMarket analysis► Case studiesCase studies► Lessons learnedLessons learned

CaveatCaveat

This is just a surface snapshot of This is just a surface snapshot of a rapidly evolving spacea rapidly evolving space

14

Investment StatisticsWeb Services Investments (2000-2003)

(# of investments and total $M)

$-

$100

$200

$300

$400

$500

2000 2001 2002 2003

0

1

2

3

4

5

6

7

8

9

10

► Analysis based on VentureSource dataAnalysis based on VentureSource data 50 startups funded since 2000 working in core Web services50 startups funded since 2000 working in core Web services 5 acquired, 10 out of business, 35 tagging along5 acquired, 10 out of business, 35 tagging along

15

XML/WS Product Categories

125 startups, 44 public companies, 169 in total

0

10

20

30

40

50

60

CMS DB Proxies Security SOI SOM Testing Tools

Together Startups Public

16

Web Services Product Focus

125 WS companies: Tools, Security, SOM, SOI, Testing

30

5 2 1

88

0

10

20

30

40

50

60

70

80

90

One Two Three Four Five

17

Virtualize

Web Services Technology Map

Deploy ExposeTestDevelop

Manage Meter

Discover Compose

NewSystems

ExistingApplications

Integrate

MonitorPolicy Mgmt

18

The Holistic View

Implementation

Architecture / Design

Testing

Deployment

IT Operations

Business Impact Analysis

Business Requirements

Tech

Str

ate

gy

Polic

y

Busi

ness

Str

ate

gy

19

Challenges To Adoption

► Business strategyBusiness strategy ““Must have” vs. “good to have”Must have” vs. “good to have” New capability vs. “better, faster, cheaper”New capability vs. “better, faster, cheaper”

► Technology strategyTechnology strategy Top-down vs. bottom-up deploymentTop-down vs. bottom-up deployment Integrated vs. best-of-breed platformsIntegrated vs. best-of-breed platforms Processes, policies and best practicesProcesses, policies and best practices Adoption of industry-specific vocabularies & biz processesAdoption of industry-specific vocabularies & biz processes

► ImplementationImplementation Immaturity of standards, tools and platformsImmaturity of standards, tools and platforms

► OperationsOperations Lack of good tools for IT/business ops monitoring/analysisLack of good tools for IT/business ops monitoring/analysis

20

Agenda

► IntroductionsIntroductions► Web services status quoWeb services status quo► Market analysisMarket analysis► Case studiesCase studies► Lessons learnedLessons learned

21

Case Study #1

► Fortune 50 companyFortune 50 company Mission critical business and consumer servicesMission critical business and consumer services Business strategy demands agility and IT cost reductionBusiness strategy demands agility and IT cost reduction

► Top-down strategy for SOA deploymentTop-down strategy for SOA deployment Abstract services notion (superset of Web services)Abstract services notion (superset of Web services) Thousands of services in heterogeneous technology islandsThousands of services in heterogeneous technology islands Architecture group defines policiesArchitecture group defines policies

► Bottom-up implementationBottom-up implementation Islands responsible for own implementationIslands responsible for own implementation No service bus; services accessed on multiple transportsNo service bus; services accessed on multiple transports Custom SOAP headers for security and other functionsCustom SOAP headers for security and other functions

► OperationsOperations Standard IT management tools but custom monitoring/analysisStandard IT management tools but custom monitoring/analysis

22

Case Study #2

► Division of Fortune 100 companyDivision of Fortune 100 company Diversified financial institution doingDiversified financial institution doing Business strategy demands agility and IT cost reductionBusiness strategy demands agility and IT cost reduction

► Top-down strategy for SOA deploymentTop-down strategy for SOA deployment Web services are just an access mechanismWeb services are just an access mechanism One mega-service exposes dozens of operationsOne mega-service exposes dozens of operations

► Top-down implementationTop-down implementation Infrastructure group implements native language librariesInfrastructure group implements native language libraries SOAP messages travel on enterprise message busSOAP messages travel on enterprise message bus Custom SOAP header for securityCustom SOAP header for security

► OperationsOperations Standard IT management toolsStandard IT management tools

23

Points of Commonality

► Same business driversSame business drivers But very different approachesBut very different approaches

► Strategy is top-downStrategy is top-down Department-level experimentation does not lead to Department-level experimentation does not lead to

company-wide deploymentcompany-wide deployment

► Only one product from a startup is in useOnly one product from a startup is in use Both companies use some open-source WS softwareBoth companies use some open-source WS software

► Great willingness to deploy Web servicesGreat willingness to deploy Web services ……despite standards/tools immaturitydespite standards/tools immaturity

► No fancy products, pleaseNo fancy products, please No interest in SOI/SOM right now; waiting out marketNo interest in SOI/SOM right now; waiting out market

24

Agenda

► IntroductionsIntroductions► Web services status quoWeb services status quo► Market analysisMarket analysis► Case studiesCase studies► Lessons learnedLessons learned

25

End Users

► Poor spending climatePoor spending climate Vendors must show clear, hard ROI within 3-12 monthsVendors must show clear, hard ROI within 3-12 months

► Moving slowlyMoving slowly Migrating from pilots to deploymentsMigrating from pilots to deployments Business cases are still in developmentBusiness cases are still in development Evaluators/decision-makers/users not always well-definedEvaluators/decision-makers/users not always well-defined

► Leveraging existing approachesLeveraging existing approaches For example: private networks for securityFor example: private networks for security

► Want the worldWant the world Get started quickly and cheaply (no big infrastructure)Get started quickly and cheaply (no big infrastructure) Long-term aligned with platform vendorsLong-term aligned with platform vendors

26

Industry

► Industry dynamics matter more than everIndustry dynamics matter more than ever► Platform vendor pathPlatform vendor path

Significant evolution in platform vendor positionSignificant evolution in platform vendor position How does their trajectory evolve over time?How does their trajectory evolve over time?

► Market structure is still formingMarket structure is still forming Product/solution boundaries are evolvingProduct/solution boundaries are evolving Integrated vs. best-of-breed is undecidedIntegrated vs. best-of-breed is undecided

► Battle of the CloudBattle of the Cloud Two world views are going to collide in the futureTwo world views are going to collide in the future Networking vendors: smart cloud, dumb endpointsNetworking vendors: smart cloud, dumb endpoints Platform vendors: smart endpoints, dumb cloudPlatform vendors: smart endpoints, dumb cloud

27

Startups

► TeamTeam Web services & domain expertiseWeb services & domain expertise Connections with standards bodies + industry leadersConnections with standards bodies + industry leaders

► ProductProduct Solid roadmap: start small, grow bigSolid roadmap: start small, grow big Design for extensibilityDesign for extensibility Make sure it fits in the evolution of the spaceMake sure it fits in the evolution of the space

► CompetitionCompetition Align with platform vendors’ expansion pathAlign with platform vendors’ expansion path A little “better, faster, cheaper” is not enoughA little “better, faster, cheaper” is not enough Need non-trivial innovation to surviveNeed non-trivial innovation to survive Applications: look for outstanding trapped assetsApplications: look for outstanding trapped assets

28

Startups (2)

► Business modelBusiness model Tricky: customers have not fully developed business casesTricky: customers have not fully developed business cases Know who you are selling to: Web services span Know who you are selling to: Web services span

development, LOB, IT operations, network, securitydevelopment, LOB, IT operations, network, security Choose carefully: best-of-breed vs. integrated solutionChoose carefully: best-of-breed vs. integrated solution

► TimingTiming The chasm is big because of market conditions and The chasm is big because of market conditions and

business models that are in flux; control burn carefullybusiness models that are in flux; control burn carefully

► Valuation + choosing the right investorsValuation + choosing the right investors Series A valuations are reasonableSeries A valuations are reasonable Huge diversity in follow-on roundsHuge diversity in follow-on rounds Look for co-investors with deep platform expertiseLook for co-investors with deep platform expertise

29

Q&A

Simeon SimeonovSimeon Simeonov

[email protected]@polarisventures.com

30

SOA Enablement

► Web services “platform” startupsWeb services “platform” startups Key players: Cape Clear, Systinet, TMEKey players: Cape Clear, Systinet, TME Mixed customer tractionMixed customer traction Diverging strategiesDiverging strategies

► RisksRisks Core platforms owned by platform leaders + other Core platforms owned by platform leaders + other

incumbents such as IONA and Novell/SilverStreamincumbents such as IONA and Novell/SilverStream Too late for startupsToo late for startups

► OpportunitiesOpportunities Web services platforms for mobile devicesWeb services platforms for mobile devices New SOA-based development models/toolsNew SOA-based development models/tools

31

Virtualization & QoS

► QoS: security, reliability, transactions, etc.QoS: security, reliability, transactions, etc.► Four modelsFour models

VAN (Grand Central): low capex but cumbersomeVAN (Grand Central): low capex but cumbersome Appliance (Datapower, Forum): transparent but what scopeAppliance (Datapower, Forum): transparent but what scope SW Layer (TME): flexible but touches endpointsSW Layer (TME): flexible but touches endpoints Integrated solution (Blue Titan): powerful but intrusiveIntegrated solution (Blue Titan): powerful but intrusive

► RisksRisks Crowded space; market timing; platform competitionCrowded space; market timing; platform competition

► OpportunitiesOpportunities Appliance company acquisitions by network vendorsAppliance company acquisitions by network vendors Solutions delivered via grid technologySolutions delivered via grid technology

32

Service-Oriented Integration

► Spans EAI, B2B and BPA/MSpans EAI, B2B and BPA/M Three key problemsThree key problems

► Legacy extension (exposing systems as Web services)Legacy extension (exposing systems as Web services)► Dynamic service discovery (repositories, negotiation, etc.)Dynamic service discovery (repositories, negotiation, etc.)► Business process automation/managementBusiness process automation/management

Key players: Cape Clear, Nimble, XAwareKey players: Cape Clear, Nimble, XAware

► RisksRisks Many established vendors (Vitria, webMethods) at high endMany established vendors (Vitria, webMethods) at high end Platform vendors have offerings in this space as wellPlatform vendors have offerings in this space as well

► OpportunitiesOpportunities New models for SOI (declarative, semantics-based, etc.)New models for SOI (declarative, semantics-based, etc.)

33

Application Assembly

► Building an entire application from Web servicesBuilding an entire application from Web services Different from SOI: not just back end but also presentationDifferent from SOI: not just back end but also presentation Enables flexible value chain & truly customizable SWEnables flexible value chain & truly customizable SW See WSIA & WSRP initiatives at OASISSee WSIA & WSRP initiatives at OASIS Key players: mainly portal vendors (Plumtree, BEA, …)Key players: mainly portal vendors (Plumtree, BEA, …)

► RisksRisks Too complex a problem; if you build it will they come?Too complex a problem; if you build it will they come? What is the value? Platform vendors own programming What is the value? Platform vendors own programming

model; packaged app vendors own the functionalitymodel; packaged app vendors own the functionality

► OpportunitiesOpportunities Flexible software-as-service solutions (platform vs. apps)Flexible software-as-service solutions (platform vs. apps)

34

Testing in SOA

► Interesting market, though immature right nowInteresting market, though immature right now Traditionally has created standalone companiesTraditionally has created standalone companies

► Current approaches are often too simplisticCurrent approaches are often too simplistic Have yet to see a sustainable roadmapHave yet to see a sustainable roadmap Key players: Empirix, MindReefKey players: Empirix, MindReef

► RisksRisks Basic testing becomes a feature of development toolsBasic testing becomes a feature of development tools

► OpportunitiesOpportunities Complex diagnostics for multi-point service deploymentsComplex diagnostics for multi-point service deployments Linking development and deployment (testing to mgmt)Linking development and deployment (testing to mgmt) Meta-data driven testing can significantly help automationMeta-data driven testing can significantly help automation

35

Service-Oriented Management

► Two key areasTwo key areas Monitoring: operations + SLA + LOB visibility, analysisMonitoring: operations + SLA + LOB visibility, analysis Control: provisioning, version mgmt, policy enforcementControl: provisioning, version mgmt, policy enforcement Key players: Amberpoint, Confluent + many othersKey players: Amberpoint, Confluent + many others

► RisksRisks Established mgmt vendors are moving to service mgmtEstablished mgmt vendors are moving to service mgmt Integrated approaches: powerful but intrusiveIntegrated approaches: powerful but intrusive Best-of-breed approaches: flexible but have dependenciesBest-of-breed approaches: flexible but have dependencies

► OpportunitiesOpportunities Significantly lower opex in non-trivial environmentsSignificantly lower opex in non-trivial environments Provide value-add analysis to business usersProvide value-add analysis to business users