Capturing Value from Big Data through Data-Driven Business Models - - PowerPoint PPT Presentation

capturing value from big data through data driven
SMART_READER_LITE
LIVE PREVIEW

Capturing Value from Big Data through Data-Driven Business Models - - PowerPoint PPT Presentation

Capturing Value from Big Data through Data-Driven Business Models Patterns from the Start-up world Philipp Hartman, Dr Mohamed Zaki and Prof Duncan McFarlane Cambridge Service Alliance University of Cambridge Data is the new oil 1 Data


slide-1
SLIDE 1

Capturing Value from Big Data through Data-Driven Business Models

Patterns from the Start-up world

Philipp Hartman, Dr Mohamed Zaki and Prof Duncan McFarlane Cambridge Service Alliance University of Cambridge

slide-2
SLIDE 2

“Data is the new oil”1

1 ¡various ¡authors, ¡e.g. ¡Clive ¡Humby ¡ ¡ ¡

0 ¡ 5000 ¡ 10000 ¡ 15000 ¡ 20000 ¡ 25000 ¡ 30000 ¡ 35000 ¡ 40000 ¡ 45000 ¡ 2005 ¡ 2010 ¡ 2015 ¡ 2020 ¡

Data ¡volume ¡per ¡year ¡(Exabytes)2 ¡

2 ¡IDC's ¡Digital ¡Universe ¡Study, ¡December ¡2012 ¡

56% ¡

¡

Top ¡Priority: ¡ “How ¡to ¡get ¡value ¡ from ¡big ¡data” ¡3 ¡

3 ¡Gartner ¡“Big ¡Data ¡Study” ¡2013 ¡

slide-3
SLIDE 3

Two general areas can be identified where big data creates value

3 ¡

How ¡to ¡get ¡value ¡from ¡Big ¡Data? ¡ OpLmizaLon ¡of ¡ exisLng ¡business1 ¡ New ¡Business ¡ Models1 ¡

Chesbrough, ¡Rosenbloom ¡(2002): ¡ Business ¡model ¡to ¡capture ¡value ¡from ¡ an ¡innovaLon ¡ ¡ Crisculo ¡(2012): ¡ ¡ New ¡technologies ¡oQen ¡first ¡ commercialized ¡through ¡start-­‑up ¡ companies ¡ ¡

1 ¡cf. ¡McKinsey ¡2011, ¡IBM ¡2012, ¡Davenport ¡2006, ¡AT ¡Kearney ¡2013, ¡EMC ¡2012 ¡

slide-4
SLIDE 4

Based on this motivation the research question was developed

4 ¡

What ¡types ¡of ¡business ¡models ¡that ¡rely ¡on ¡data ¡as ¡a ¡key ¡resource ¡(i.e. ¡ data-­‑driven ¡business ¡models) ¡can ¡be ¡found ¡in ¡start ¡up ¡companies? ¡

How ¡to ¡analyse ¡data-­‑ driven ¡business ¡ models? ¡ Sub ¡ quesLons ¡ Data-­‑driven ¡business ¡ model ¡framework ¡ How ¡to ¡idenLfy ¡ pa^erns? ¡ Research ¡ QuesLon ¡ Clustering ¡

slide-5
SLIDE 5

The research was done in five steps

5 ¡ Case ¡studies ¡ Finding ¡ Pa^erns ¡ Data ¡collecLon ¡ & ¡coding ¡ Build ¡the ¡ framework ¡ Literature ¡Review ¡

How ¡to ¡analyse ¡data-­‑ driven ¡business ¡ models? ¡ How ¡to ¡idenLfy ¡ pa^erns? ¡

slide-6
SLIDE 6

The first step was a literature review with three different topics

6 ¡ Literature ¡Review ¡ Big ¡Data ¡ DefiniLon ¡ Value ¡CreaLon ¡ Business ¡Model ¡ DefiniLon ¡ Business ¡Model ¡ Frameworks ¡ Related ¡Work ¡ Data ¡driven ¡business ¡ Models ¡ Cloud ¡business ¡ models ¡ Case ¡studies ¡ Finding ¡ Pa^erns ¡ Data ¡collecLon ¡ & ¡coding ¡ Build ¡the ¡ framework ¡ Literature ¡Review ¡

slide-7
SLIDE 7

The first step was a literature review with three different topics

7 ¡ Literature ¡Review ¡ Big ¡Data ¡ DefiniLon ¡ Value ¡CreaLon ¡ Business ¡Model ¡ DefiniLon ¡ Business ¡Model ¡ Frameworks ¡ Related ¡Work ¡ Data ¡driven ¡business ¡ Models ¡ Cloud ¡business ¡ models ¡ Case ¡studies ¡ Finding ¡ Pa^erns ¡ Data ¡collecLon ¡ & ¡coding ¡ Build ¡the ¡ framework ¡ Literature ¡Review ¡

slide-8
SLIDE 8

Literature review: Business Model

8 ¡ Literature ¡Review ¡ Big ¡Data ¡ DefiniLon ¡ Value ¡CreaLon ¡ Business ¡Model ¡ DefiniLon ¡ Business ¡Model ¡ Frameworks ¡ Related ¡Work ¡ Data ¡driven ¡business ¡ Models ¡ Cloud ¡business ¡ models ¡ Case ¡studies ¡ Finding ¡ Pa^erns ¡ Data ¡collecLon ¡ & ¡coding ¡ Build ¡the ¡ framework ¡ Literature ¡Review ¡

slide-9
SLIDE 9

Business model key components were synthesized from existing frameworks

ExisLng ¡Business ¡Model ¡Frameworks ¡

  • ­‑

Chesbrough ¡& ¡Rosenbloom ¡2002 ¡

  • ­‑

Hedman ¡& ¡Kaling ¡2003 ¡

  • ­‑

Osterwalder ¡2004 ¡

  • ­‑

Morris ¡2005 ¡

  • ­‑

Johnson, ¡Christensen ¡et. ¡al. ¡2008 ¡

  • ­‑

Al-­‑Debei ¡2010 ¡

  • ­‑

Burkhart ¡2012 ¡

Value ¡Capturing Value ¡CreaFon

Key ¡Resources Key ¡AcLviLes Cost ¡structure Revenue ¡Model Customer ¡Segment Value ¡ProposiLon Business ¡Model ¡DefiniLon ¡ Business ¡Model ¡Key ¡Components ¡

  • ­‑

No ¡universally ¡accepted ¡definiLon ¡

  • f ¡the ¡concept ¡

(Weill, ¡Malone ¡et ¡al. ¡2011) ¡

  • ­‑

Most ¡definiLons ¡refer ¡to ¡ ¡ value ¡creaFon ¡& ¡value ¡capturing ¡ ¡ ¡

slide-10
SLIDE 10

Only a few papers are available in this field

10 ¡ Literature ¡Review ¡ Big ¡Data ¡ DefiniLon ¡ Value ¡CreaLon ¡ Business ¡Model ¡ DefiniLon ¡ Business ¡Model ¡ Frameworks ¡ Related ¡Work ¡ Data ¡driven ¡business ¡ Models ¡ Cloud ¡business ¡ models ¡ Case ¡studies ¡ Finding ¡ Pa^erns ¡ Data ¡collecLon ¡ & ¡coding ¡ Build ¡the ¡ framework ¡ Literature ¡Review ¡

slide-11
SLIDE 11

The review was extend to cloud business models

11 ¡ Literature ¡Review ¡ Big ¡Data ¡ DefiniLon ¡ Value ¡CreaLon ¡ Business ¡Model ¡ DefiniLon ¡ Business ¡Model ¡ Frameworks ¡ Related ¡Work ¡ Data ¡driven ¡business ¡ Models ¡ Cloud ¡business ¡ models ¡ Case ¡studies ¡ Finding ¡ Pa^erns ¡ Data ¡collecLon ¡ & ¡coding ¡ Build ¡the ¡ framework ¡ Literature ¡Review ¡

slide-12
SLIDE 12

The literature review identified several gaps

12 ¡

  • Li^le ¡academic ¡research ¡on ¡big ¡data ¡and ¡value ¡creaLon ¡– ¡mostly ¡

whitepapers ¡

  • Gap ¡in ¡literature: ¡data-­‑driven ¡business ¡models ¡
  • O^o, ¡Aier ¡(2013) ¡interesLng ¡paper ¡but ¡limited ¡to ¡specific ¡

industry ¡> ¡no ¡generalizaLon ¡possible ¡

  • Similar ¡research ¡for ¡cloud ¡business ¡models ¡(cf. ¡Labes, ¡Erek ¡et. ¡Al. ¡

2013) ¡

Case ¡studies ¡ Finding ¡ Pa^erns ¡ Data ¡collecLon ¡ & ¡coding ¡ Build ¡the ¡ framework ¡ Literature ¡Review ¡

slide-13
SLIDE 13

The framework was build from literature starting from the key components

Data-­‑Driven-­‑ Business ¡ Model ¡ Data ¡Sources ¡ Internal ¡ exisLng ¡data ¡ Self-­‑ generated ¡ Data ¡ External ¡ Acquired ¡ Data ¡ Customer ¡ provided ¡ ¡ Free ¡ available ¡ Open ¡Data ¡ Social ¡Media ¡ data ¡ Web ¡Crawled ¡ Data ¡ Key ¡AcLvity ¡ Data ¡ GeneraLon ¡ Crowdsourci ng ¡ Tracking ¡& ¡ Other ¡ Data ¡ AcquisiLon ¡ Processing ¡ AggregaLon ¡ AnalyLcs ¡ descripLve ¡ predicLve ¡ prescripLve ¡ VisualizaLon ¡ DistribuLon ¡ Offering ¡ Data ¡ InformaLon/ Knowledge ¡ Non-­‑Data ¡ Product/ Service ¡ Target ¡ Customer ¡ B2B ¡ B2C ¡ Revenue ¡ Model ¡ Asset ¡Sale ¡ Lending/ RenLng/ Leasing ¡ Licensing ¡ Usage ¡fee ¡ SubscripLon ¡ fee ¡ AdverLsing ¡ Specific ¡cost ¡ advantage ¡

Data-­‑Driven ¡ Business ¡Model ¡ Data ¡Sources ¡ Key ¡AcLvity ¡ Offering ¡ Target ¡Customer ¡ Revenue ¡Model ¡ Specific ¡cost ¡ advantage ¡

Data ¡collecLon ¡ & ¡coding ¡ Case ¡studies ¡ Finding ¡ Pa^erns ¡ Literature ¡Review ¡ Build ¡the ¡ framework ¡

Features ¡for ¡ each ¡dimension ¡ Data-­‑Driven ¡Business ¡Model ¡ Framework ¡ Business ¡Model ¡Key ¡ Components ¡(Dimensions) ¡

Data ¡Sources ¡

Features ¡for ¡ data ¡sources ¡

slide-14
SLIDE 14

Synthesizing the different sources leads to the taxonomy

14 ¡ Data ¡Sources ¡ Internal ¡ exisLng ¡data ¡ Self-­‑generated ¡ Data ¡ External ¡ Acquired ¡Data ¡ Customer ¡ provided ¡ ¡ Free ¡available ¡ Open ¡Data ¡ Social ¡Media ¡data ¡ Web ¡Crawled ¡ Data ¡

slide-15
SLIDE 15

Dimension: Activities

15 ¡

Key ¡AcLvity ¡ Data ¡GeneraLon ¡ Crowdsourcing ¡ Tracking ¡& ¡Other ¡ Data ¡AcquisiLon ¡ Processing ¡ AggregaLon ¡ AnalyLcs ¡ descripLve ¡ predicLve ¡ prescripLve ¡ VisualizaLon ¡ DistribuLon ¡

slide-16
SLIDE 16

Dimension: Offering

16 ¡

Offering ¡ Data ¡ InformaLon/ Knowledge ¡ Non-­‑Data ¡ Product/Service ¡

slide-17
SLIDE 17

Dimension: Revenue Model

17 ¡

Revenue ¡Model ¡ Asset ¡Sale ¡ Lending/RenLng/ Leasing ¡ Licensing ¡ Usage ¡fee ¡ SubscripLon ¡fee ¡ AdverLsing ¡

slide-18
SLIDE 18

Dimension: Target Customer

18 ¡

Target ¡Customer ¡ B2B ¡ B2C ¡

slide-19
SLIDE 19

Data ¡collecLon ¡ & ¡coding ¡

The final framework

19 ¡ Case ¡studies ¡ Finding ¡ Pa^erns ¡ Literature ¡Review ¡ Build ¡the ¡ framework ¡

Data-­‑Driven-­‑ Business ¡Model ¡ Data ¡Sources ¡ Internal ¡ exisLng ¡data ¡ Self-­‑generated ¡ Data ¡ External ¡ Acquired ¡Data ¡ Customer ¡ provided ¡ ¡ Free ¡available ¡ Open ¡Data ¡ Social ¡Media ¡ data ¡ Web ¡Crawled ¡ Data ¡ Key ¡AcLvity ¡ Data ¡GeneraLon ¡ Crowdsourcing ¡ Tracking ¡& ¡Other ¡ Data ¡AcquisiLon ¡ Processing ¡ AggregaLon ¡ AnalyLcs ¡ descripLve ¡ predicLve ¡ prescripLve ¡ VisualizaLon ¡ DistribuLon ¡ Offering ¡ Data ¡ InformaLon/ Knowledge ¡ Non-­‑Data ¡ Product/Service ¡ Target ¡Customer ¡ B2B ¡ B2C ¡ Revenue ¡Model ¡ Asset ¡Sale ¡ Lending/RenLng/ Leasing ¡ Licensing ¡ Usage ¡fee ¡ SubscripLon ¡fee ¡ AdverLsing ¡ Specific ¡cost ¡ advantage ¡
slide-20
SLIDE 20

Data collection and coding

20 ¡ Case ¡studies ¡ Finding ¡ Pa^erns ¡ Build ¡the ¡ framework ¡ Literature ¡Review ¡ Data ¡collecLon ¡ & ¡coding ¡

Data ¡collecLon ¡ Data ¡analysis ¡ Sampling ¡

slide-21
SLIDE 21

The data was generated using public available sources

21 ¡

Tag: ¡“big ¡data” ¡ “big ¡data ¡analyLcs” ¡ 1329 ¡companies ¡ Data ¡collecLon ¡ Company ¡informaLon ¡

  • Company ¡websites ¡
  • Press ¡releases ¡

Public ¡sources ¡

  • Coding ¡of ¡sources ¡

using ¡data ¡driven ¡ business ¡model ¡ framework ¡

  • Nvivo ¡

Data ¡analysis ¡ 299 ¡Sources ¡ ~3 ¡sources/comp ¡ Sampling ¡ 100 ¡Companies ¡ cleaning ¡ Random ¡sample ¡ 100 ¡binary ¡feature ¡ vectors ¡

slide-22
SLIDE 22

Overall Analysis: Data Source

22 ¡

0% ¡ 10% ¡ 20% ¡ 30% ¡ 40% ¡ 50% ¡ 60% ¡ ¡Acquired ¡Data ¡ ¡Customer&Partner-­‑provided ¡Data ¡ ¡Free ¡available ¡ Crowd ¡Sourced ¡ Tracked ¡& ¡Other ¡

Note: ¡Sum ¡> ¡100% ¡as ¡companies ¡might ¡use ¡mulLple ¡data ¡sources ¡

  • >50% ¡of ¡companies ¡

rely ¡on ¡free ¡available ¡ data ¡

  • >50% ¡of ¡companies ¡

use ¡data ¡provided ¡by ¡ customers/partners ¡

slide-23
SLIDE 23

Overall Analysis: Key Activities

23 ¡

0% ¡ 10% ¡ 20% ¡ 30% ¡ 40% ¡ 50% ¡ 60% ¡ 70% ¡ 80% ¡ ¡AggregaLon ¡ ¡AnalyLcs ¡ ¡DescripLve ¡AnalyLcs ¡ ¡PredicLve ¡AnalyLcs ¡ ¡PrescripLve ¡AnalyLcs ¡ ¡Data ¡acquisLon ¡ ¡Data ¡generaLon ¡ ¡Data ¡processing ¡ ¡DistribuLon ¡ ¡VisualizaLon ¡

  • >70% ¡of ¡

companies ¡use ¡ analyLcs ¡ ¡

  • ­‑ ¡mostly ¡

descripLve ¡ ¡

Note: ¡Sum ¡> ¡100% ¡as ¡some ¡companies ¡rely ¡on ¡mulLple ¡revenue ¡models ¡

slide-24
SLIDE 24

Overall Analysis: Revenue Model

24 ¡

0% ¡ 5% ¡ 10% ¡ 15% ¡ 20% ¡ 25% ¡ 30% ¡ 35% ¡ 40% ¡ 45% ¡ 50% ¡ ¡AdverLsing ¡ ¡Asset ¡Sales ¡ ¡Brokerage ¡Fees ¡ ¡Lending ¡RenLng ¡Leasing ¡ ¡Licensing ¡ ¡SubscripLon ¡fee ¡ ¡Usage ¡Fee ¡ ¡No ¡informaLon ¡

  • Majority ¡of ¡

revenue ¡models ¡ based ¡on ¡ subscripLon ¡and/

  • r ¡usage ¡fee ¡
  • No ¡informaLon ¡

about ¡the ¡ revenue ¡model ¡ as ¡many ¡ companies ¡are ¡in ¡ an ¡early ¡stage ¡

Note: ¡Sum ¡> ¡100% ¡as ¡some ¡companies ¡rely ¡on ¡mulLple ¡revenue ¡models ¡

slide-25
SLIDE 25

Overall Analysis: Target Customer

25 ¡

70% ¡ 17% ¡ 13% ¡ B2B ¡ B2C ¡ both ¡

  • There ¡seems ¡to ¡be ¡a ¡

noteworthy ¡ predominance ¡of ¡ B2B ¡business ¡ models ¡

  • But ¡no ¡reference ¡

data ¡found ¡

slide-26
SLIDE 26

BM patterns were identified using a clustering approach

26 ¡

Ketchen, ¡David ¡J.; ¡Shook, ¡Christopher ¡L. ¡(1996): ¡The ¡ApplicaLon ¡of ¡Cluster ¡Analysis ¡in ¡Strategic ¡Managment ¡Reserach: ¡An ¡Analysis ¡and ¡

  • CriLque. ¡In: ¡Strat. ¡Mgmt. ¡J. ¡17 ¡(6). ¡

¡ Han, ¡Jiawei; ¡Kamber, ¡Micheline ¡(2011): ¡Data ¡mining. ¡Concepts ¡and ¡techniques. ¡ ¡ Mooi, ¡Erik; ¡Sarstedt, ¡Marko ¡(2011): ¡Cluster ¡Analysis. ¡In: ¡A ¡Concise ¡Guide ¡to ¡Market ¡Research. ¡S. ¡237-­‑284. ¡ ¡ ¡ Miligan, ¡Glenn ¡W. ¡(1996): ¡Clustering ¡ValidaLon: ¡Results ¡and ¡ImplicaLons ¡for ¡Applied ¡Analyses. ¡In ¡Phipps ¡Arabie, ¡Lawrence ¡J. ¡Hubert, ¡ Geert ¡de ¡Soete ¡(Eds.): ¡Clustering ¡and ¡classificaLon. ¡pp. ¡341–376. ¡ ¡

Case ¡studies ¡ Data ¡collecLon ¡ & ¡coding ¡ Build ¡the ¡ framework ¡ Literature ¡Review ¡ Finding ¡ Pa^erns ¡

  • 2. ¡Clustering ¡

method ¡

  • 1. ¡Clustering ¡

Variables ¡

  • 3. ¡Number ¡of ¡

Clusters ¡

  • 4. ¡Validate ¡& ¡

Interpret ¡C. ¡

slide-27
SLIDE 27

7 Business Model Cluster were identified

27 ¡

¡ ¡ Cluster ¡ 1 ¡ 2 ¡ 3 ¡ 4 ¡ 5 ¡ 6 ¡ 7 ¡ Data ¡Source ¡ Acquired ¡Data ¡ 0 ¡ 0 ¡ 1 ¡ 0 ¡ 0 ¡ 0 ¡ 0 ¡ Customer-­‑provided ¡Data ¡ 0 ¡ 1 ¡ 1 ¡ 0 ¡ 0 ¡ 1 ¡ 1 ¡ Free ¡available ¡ 1 ¡ 0 ¡ 1 ¡ 0 ¡ 1 ¡ 0 ¡ 1 ¡ CrowdSourced ¡ 0 ¡ 0 ¡ 0 ¡ 0 ¡ 0 ¡ 0 ¡ 0 ¡ Tracked, ¡Generated ¡& ¡other ¡ 0 ¡ 0 ¡ 0 ¡ 1 ¡ 0 ¡ 0 ¡ 0 ¡ Key ¡AcLvity ¡ AggregaLon ¡ 1 ¡ 0 ¡ 0 ¡ 0 ¡ 0 ¡ 1 ¡ 1 ¡ AnalyLcs ¡ 0 ¡ 1 ¡ 1 ¡ 1 ¡ 1 ¡ 0 ¡ 1 ¡ Data ¡acquisLon ¡ 0 ¡ 0 ¡ 1 ¡ 0 ¡ 0 ¡ 0 ¡ 0 ¡ Data ¡generaLon ¡ 0 ¡ 0 ¡ 0 ¡ 1 ¡ 0 ¡ 0 ¡ 1 ¡ Number ¡of ¡companies ¡ 17 ¡ 28 ¡ 5 ¡ 16 ¡ 14 ¡ 6 ¡ 14 ¡ Type ¡ A ¡ B ¡

  • ­‑ ¡

C ¡ D ¡ E ¡ F ¡

slide-28
SLIDE 28

6 significant Business Model types were identified

28 ¡

Type ¡B: ¡“AnalyLcs-­‑as-­‑a-­‑Service” ¡ Type ¡C: ¡“Data ¡generaLon ¡& ¡AnalyLcs” ¡ Type ¡D: ¡“Free ¡Data ¡Knowledge ¡Discovery” ¡ Type ¡A: ¡“Free ¡Data ¡Collector ¡& ¡Aggregator” ¡ Type ¡E: ¡ ¡“Data ¡AggregaLon-­‑as-­‑a-­‑Service” ¡ Type ¡F: ¡“MulL-­‑Source ¡data ¡mashup ¡and ¡analysis” ¡

slide-29
SLIDE 29

The 6 BM types are characterised by the key activities and key data sources

29 ¡

Type ¡F ¡ Type ¡A ¡ Type ¡D ¡ Type ¡E ¡ Type ¡B ¡ Type ¡C ¡ AggregaLon ¡ AnalyLcs ¡ Data ¡generaLon ¡ Free ¡ ¡available ¡ Customer ¡ provided ¡ Tracked ¡& ¡ generated ¡ Key ¡acFvity ¡ Key ¡Data ¡Source ¡

slide-30
SLIDE 30

Type D: “Free Data Knowledge Discovery”

  • 1. ¡DealAngel ¡
  • 2. ¡Gild ¡
  • 3. ¡Insightpool ¡
  • 4. ¡Juristat ¡
  • 5. ¡Market ¡Prophit ¡
  • 6. ¡MixRank ¡
  • 7. ¡Numberfire ¡
  • 8. ¡Olery ¡
  • 9. ¡PeerIndex ¡
  • 10. ¡PolyGraph ¡
  • 11. ¡Review ¡Signal ¡
  • 12. ¡Tellagence ¡
  • 13. ¡traackr ¡
  • 14. ¡Trendspo^r ¡
  • ­‑

Free ¡available ¡

  • ­‑

Social ¡Media ¡

  • ­‑

Open ¡Data ¡

  • ­‑

Web ¡Crawled ¡

B2B ¡ B2C ¡

Key ¡AcLviLes ¡ Revenue ¡Model ¡ Key ¡Data ¡Source ¡

  • ­‑

AnalyLcs ¡ Target ¡Customer ¡

0 ¡ 5 ¡ 10 ¡ 15 ¡ DescripLve ¡ PredicLve ¡ PrescripLve ¡ 0 ¡ 2 ¡ 4 ¡ 6 ¡ 8 ¡ SubscripLon ¡ Usage ¡Fee ¡ AdverLsing ¡ Brokearge ¡Fees ¡ No ¡InformaLon ¡

Companies ¡

slide-31
SLIDE 31

Type D: Examples

31 ¡

“Using ¡patent-­‑pending ¡technology, ¡Gild ¡ evaluates ¡the ¡work ¡of ¡millions ¡of ¡ developers ¡so ¡companies ¡using ¡Gild’s ¡ talent ¡acquisiLon ¡tools ¡know ¡who’s ¡good ¡ and ¡can ¡target ¡the ¡right ¡candidates.” ¡ ¡

  • Key ¡Data: ¡Free ¡available ¡websites ¡

(GitHub, ¡Google ¡Codes) ¡

  • Key ¡AcLviLes: ¡AnalyLcs ¡
  • Revenue ¡Model: ¡Monthly ¡subscripLon ¡
  • Target ¡Customer: ¡B2B ¡

¡ “ ¡Our ¡goal ¡is ¡to ¡provide ¡the ¡most ¡ accurate ¡and ¡honest ¡reviews ¡possible ¡by ¡ using ¡the ¡data ¡consumers ¡create. ¡We ¡ listen ¡to ¡the ¡conversaLons, ¡analyze ¡them ¡ and ¡visualize ¡them ¡for ¡consumers.” ¡ ¡

  • Key ¡Data: ¡Twi^er ¡
  • Key ¡AcLviLes: ¡AnalyLcs ¡
  • Revenue ¡Model: ¡AdverLsing ¡
  • Target ¡Customer: ¡B2B ¡(B2C) ¡

¡

slide-32
SLIDE 32

Finding ¡ Pa^erns ¡

The cases studies will be validated the framework and the clustering

32 ¡ Data ¡collecLon ¡ & ¡coding ¡ Build ¡the ¡ framework ¡ Literature ¡Review ¡ Case ¡studies ¡

4 ¡case ¡studies ¡with ¡ companies ¡from ¡the ¡ sample ¡such ¡as ¡ ¡ Purpose: ¡

  • 1. Validate ¡framework ¡& ¡

clusters ¡

  • 2. Illustrate ¡business ¡

model ¡types ¡through ¡ examples ¡

  • 3. IdenLfy ¡specific ¡

challenges ¡ ¡

slide-33
SLIDE 33

Summary

33 ¡

  • ­‑

Gap ¡in ¡literature ¡idenLfied ¡

  • ­‑

RQ: ¡What ¡types ¡of ¡business ¡models ¡that ¡rely ¡on ¡data ¡as ¡a ¡key ¡resource ¡ (i.e. ¡data-­‑driven ¡business ¡models) ¡can ¡be ¡found ¡in ¡start ¡up ¡companies? ¡

  • ­‑

5 ¡(7) ¡different ¡business ¡model ¡pa^erns ¡idenLfied ¡

  • ­‑

Data-­‑driven ¡business ¡model ¡framework ¡created ¡to ¡enable ¡analysis ¡

  • ­‑

Pa^ern ¡idenLficaLon ¡through ¡clustering ¡

  • ­‑

ValidaLon ¡through ¡Case-­‑Studies ¡

slide-34
SLIDE 34

Limitations & Outlook

34 ¡

LimitaLons ¡

  • Only ¡100 ¡samples ¡
  • Only ¡start ¡up ¡companies ¡

¡

  • Bias ¡of ¡data ¡source ¡(AngelList) ¡
  • StaLsLcal ¡significance ¡of ¡

clustering ¡result ¡

  • Only ¡public ¡available ¡sources ¡

used ¡

  • No ¡statement ¡about ¡success ¡of ¡

a ¡parLcular ¡business ¡model ¡ Outlook/Next ¡Steps ¡

  • 1. Improve ¡validity ¡of ¡findings ¡
  • 1. Increase ¡sample ¡size ¡to ¡test ¡

clusters ¡

  • 2. More ¡Case-­‑studies ¡to ¡

illustrate/validate ¡clusters ¡

  • 2. Include ¡established ¡organizaFons ¡
  • 3. Develop ¡methodology ¡to ¡judge ¡

(financial) ¡performance ¡of ¡ different ¡business ¡models ¡ ¡

slide-35
SLIDE 35

Appendix

35 ¡

slide-36
SLIDE 36

Literature

36 ¡

Al-­‑Debei, ¡Mutaz ¡M.; ¡Avison, ¡David ¡(2010): ¡Developing ¡a ¡unified ¡framework ¡of ¡the ¡business ¡ model ¡concept. ¡In ¡Eur ¡J ¡Inf ¡Syst ¡19 ¡(3), ¡pp. ¡359–376. ¡ ¡ Burkhart, ¡Thomas; ¡Wolter, ¡Stephan; ¡Schief, ¡Markus; ¡Krumeich, ¡Julian; ¡Di ¡ValenLn, ¡ ChrisLna; ¡Werth, ¡Dirk ¡et ¡al. ¡(2012): ¡A ¡comprehensive ¡approach ¡towards ¡the ¡structural ¡ descripLon ¡of ¡business ¡models. ¡In ¡: ¡Proceedings ¡of ¡the ¡InternaLonal ¡Conference ¡on ¡ Management ¡of ¡Emergent ¡Digital ¡EcoSystems. ¡New ¡York, ¡NY, ¡USA: ¡ACM ¡(MEDES ¡’12), ¡pp. ¡ 88–102. ¡ ¡ ¡ Chesbrough, ¡H.; ¡Rosenbloom, ¡R. ¡(2002): ¡The ¡role ¡of ¡the ¡business ¡model ¡in ¡capturing ¡value ¡ from ¡innovaLon: ¡evidence ¡from ¡Xerox ¡CorporaLon's ¡technology ¡spin-­‑off ¡companies. ¡In ¡ Industrial ¡and ¡Corporate ¡Change ¡11 ¡(3), ¡pp. ¡529–555. ¡ ¡ Criscuolo, ¡Paola; ¡Nicolaou, ¡Nicos; ¡Salter, ¡Ammon ¡(2012): ¡The ¡elixir ¡(or ¡burden) ¡of ¡youth? ¡ Exploring ¡differences ¡in ¡innovaLon ¡between ¡start-­‑ups ¡and ¡established ¡firms. ¡In ¡Research ¡ Policy ¡41 ¡(2), ¡pp. ¡319–333. ¡ ¡ ¡

slide-37
SLIDE 37

Literature

37 ¡

Davenport, ¡Thomas ¡H.; ¡Harris, ¡Jeanne ¡G. ¡(2007): ¡CompeLng ¡on ¡analyLcs. ¡The ¡new ¡science ¡

  • f ¡winning. ¡Boston, ¡Mass: ¡Harvard ¡Business ¡School ¡Press. ¡

¡ Everi^, ¡Brian; ¡Landau, ¡Sabine; ¡Leese, ¡Morven ¡(2011): ¡Cluster ¡analysis. ¡5th ¡ed. ¡London, ¡New ¡ York: ¡Arnold; ¡Oxford ¡University ¡Press. ¡ ¡ Hagen, ¡ChrisLan; ¡Khan, ¡Khalid; ¡Ciobo, ¡Marco; ¡Miller, ¡Jason; ¡Wall, ¡Dan; ¡Evans, ¡Hugo; ¡ Yadava, ¡Ajay ¡(2013): ¡Big ¡Data ¡and ¡the ¡CreaLve ¡DestrucLon ¡of ¡Today's ¡Business ¡Models. ¡

  • ATKearney. ¡

¡ Han, ¡Jiawei; ¡Kamber, ¡Micheline; ¡Pei, ¡Jian ¡(2011): ¡Data ¡Mining. ¡Concepts ¡and ¡Techniques. ¡ 3rd ¡ed. ¡Burlington: ¡Elsevier ¡Science. ¡ ¡ Hedman, ¡Jonas; ¡Kalling, ¡Thomas ¡(2003): ¡The ¡business ¡model ¡concept: ¡theoreLcal ¡ underpinnings ¡and ¡empirical ¡illustraLons. ¡In ¡European ¡Journal ¡of ¡Informa=on ¡Systems ¡12 ¡ (1), ¡pp. ¡49–59. ¡ ¡ ¡ ¡

slide-38
SLIDE 38

Literature

38 ¡

Johnson, ¡Mark ¡W.; ¡Christensen, ¡Clayton ¡M.; ¡Kagermann, ¡Henning ¡(2008): ¡ReinvenLng ¡ your ¡business ¡model. ¡In ¡Harvard ¡Business ¡Review ¡86 ¡(12), ¡pp. ¡57–68. ¡ ¡ Labes, ¡SLne; ¡Erek, ¡Koray; ¡Zarnekow, ¡Ruediger ¡(2013): ¡Common ¡Pa^erns ¡of ¡Cloud ¡Business ¡

  • Models. ¡In ¡: ¡AMCIS ¡2013 ¡Proceedings. ¡

¡ Manyika, ¡James; ¡Chui, ¡Michael; ¡Brown, ¡Brad; ¡Bughin, ¡Jacques; ¡Dobbs, ¡Richard; ¡Roxburgh, ¡ Charles; ¡Hung ¡Byres, ¡Angela ¡(2011): ¡Big ¡data: ¡The ¡next ¡fronLer ¡for ¡innovaLon, ¡ compeLLon, ¡and ¡producLvity. ¡McKinsey ¡Global ¡InsLtute. ¡ ¡ Morris, ¡Michael; ¡Schindehu^e, ¡Minet; ¡Allen, ¡Jeffrey ¡(2005): ¡The ¡entrepreneur's ¡business ¡ model: ¡toward ¡a ¡unified ¡perspecLve. ¡In ¡Special ¡Sec=on: ¡The ¡Nonprofit ¡Marke=ng ¡ Landscape ¡58 ¡(6), ¡pp. ¡726–735. ¡ ¡ ¡ Negash, ¡Solomon ¡(2004): ¡Business ¡Intelligence. ¡In ¡Communica=ons ¡of ¡the ¡Associa=on ¡for ¡ Informa=on ¡Systems ¡13, ¡pp. ¡177–195. ¡ ¡

slide-39
SLIDE 39

Literature

39 ¡

Osterwalder, ¡Alexander ¡(2004): ¡The ¡Business ¡Model ¡Ontology. ¡A ¡ProposiLon ¡in ¡Design ¡ Science ¡Research. ¡These. ¡Ecole ¡des ¡Hautes ¡Etudes ¡Commerciales ¡de ¡l’Université ¡de ¡ Lausanne, ¡Lausanne. ¡ ¡ O^o, ¡Boris; ¡Aier, ¡Stephan ¡(2013): ¡Business ¡Models ¡in ¡the ¡Data ¡Economy: ¡A ¡Case ¡Study ¡ from ¡the ¡Business ¡Partner ¡Data ¡Domain. ¡With ¡assistance ¡of ¡Rainer ¡Alt, ¡Bogdan ¡Franczyk. ¡ In ¡: ¡Proceedings ¡of ¡the ¡11th ¡InternaLonal ¡Conference ¡on ¡WirtschaQsinformaLk ¡(WI ¡2013) ¡: ¡ The ¡11th ¡InternaLonal ¡Conference ¡on ¡WirtschaQsinformaLk ¡(WI ¡2013), ¡vol. ¡1. ¡Leipzig, ¡pp. ¡ 475–489. ¡ ¡ ¡ Pham, ¡D. ¡T.; ¡Dimov, ¡S. ¡S.; ¡Nguyen, ¡C. ¡D. ¡(2005): ¡SelecLon ¡of ¡K ¡in ¡K-­‑means ¡clustering. ¡In ¡ Proceedings ¡of ¡the ¡Ins=tu=on ¡of ¡Mechanical ¡Engineers, ¡Part ¡C: ¡Journal ¡of ¡Mechanical ¡ Engineering ¡Science ¡219 ¡(1), ¡pp. ¡103–119. ¡ ¡ Pham, ¡D. ¡T.; ¡Afify, ¡A. ¡A. ¡(2007): ¡Clustering ¡techniques ¡and ¡their ¡applicaLons ¡in ¡

  • engineering. ¡In ¡Proceedings ¡of ¡the ¡Ins=tu=on ¡of ¡Mechanical ¡Engineers, ¡Part ¡C: ¡Journal ¡of ¡

Mechanical ¡Engineering ¡Science ¡221 ¡(11), ¡pp. ¡1445–1459. ¡ ¡ ¡

slide-40
SLIDE 40

Literature

40 ¡

Rousseeuw, ¡Peter ¡J. ¡(1987): ¡Silhoue^es: ¡A ¡graphical ¡aid ¡to ¡the ¡interpretaLon ¡and ¡ validaLon ¡of ¡cluster ¡analysis. ¡In ¡Journal ¡of ¡Computa=onal ¡and ¡Applied ¡Mathema=cs ¡20 ¡(0), ¡

  • pp. ¡53–65. ¡ ¡

¡ Schroeck, ¡Michael; ¡Shockley, ¡Rebecca; ¡Smart, ¡Janet; ¡Romero-­‑Morales, ¡Dolores; ¡Tufano, ¡ Peter ¡(2012): ¡AnalyLcs: ¡The ¡real-­‑world ¡use ¡of ¡big ¡data. ¡How ¡innovaLve ¡enterprises ¡extract ¡ value ¡from ¡uncertain ¡data. ¡IBM ¡InsLtute ¡for ¡Business ¡Value; ¡Saïd ¡Business ¡School ¡at ¡the ¡ University ¡of ¡Oxford. ¡ ¡ Singh, ¡Ranjit; ¡Singh, ¡Kawaljeet ¡(2010): ¡A ¡DescripLve ¡ClassificaLon ¡of ¡Causes ¡of ¡Data ¡ Quality ¡Problems ¡in ¡Data ¡Warehousing. ¡In ¡IJCSI ¡Interna=onal ¡Journal ¡of ¡Computer ¡Science ¡I ¡ 7 ¡(3), ¡pp. ¡41–50. ¡ ¡ Weill, ¡Peter; ¡Malone, ¡Thomas ¡W.; ¡Apel, ¡Thomas ¡G. ¡(2011): ¡The ¡business ¡models ¡investors ¡

  • prefer. ¡In ¡MIT ¡Sloan ¡Management ¡Review ¡52 ¡(4), ¡p. ¡17. ¡

¡ ¡ ¡ ¡

slide-41
SLIDE 41

The Clustering Process

41 ¡

Variables ¡relevant ¡to ¡ determine ¡clustering ¡ (Miligan ¡1996) ¡ #Variables ¡has ¡to ¡ match ¡#samples ¡ (Mooi ¡2011) ¡ ¡ ~ ¡2m ¡samples ¡for ¡m ¡ variables: ¡ ¡ 6-­‑7 ¡variables ¡ Avoid ¡high ¡correlaLon ¡ between ¡variables ¡ (<0.9) ¡(Mooi ¡2011) ¡ 2 ¡Dimensions: ¡ ¡“Data ¡source” ¡& ¡ ¡ “Key ¡AcLvity” ¡ ¡ ¡ 9 ¡variables ¡

  • max. ¡correlaLon: ¡

0,5 ¡

  • 2. ¡Clustering ¡

method ¡

  • 3. ¡Number ¡of ¡

Clusters ¡

  • 4. ¡Validate ¡& ¡

Interpret ¡C. ¡

  • 1. ¡Clustering ¡

Variables ¡

slide-42
SLIDE 42

The Clustering Process

42 ¡

ParLLoning ¡ Hierarchical ¡ Density-­‑based ¡ Grid-­‑based ¡ Clustering ¡ Method ¡ (Han ¡2011) ¡ Proximity ¡ Measure ¡

  • 4. ¡Validate ¡& ¡

Interpret ¡C. ¡

  • 1. ¡Clustering ¡

Variables ¡

  • 3. ¡Number ¡of ¡

Clusters ¡

  • 2. ¡Clustering ¡

method ¡ K-­‑Medoids ¡ Include ¡neg. ¡match ¡ Exclude ¡neg. ¡match ¡ Euclidean ¡ Distance ¡

slide-43
SLIDE 43

There is no “one right solution” for the number of clusters

43 ¡

large ¡to ¡reflect ¡specific ¡ differences ¡ k ¡<< ¡n ¡

  • 1. Use ¡a-­‑priori ¡knowledge ¡to ¡determine ¡number ¡of ¡clusters ¡
  • 2. Visual ¡approaches ¡
  • 3. Rule ¡of ¡thumb ¡(Han ¡2011): ¡ ¡
  • 4. “Elbow” ¡method ¡
  • 5. StaLsLcal ¡methods ¡

𝑙 ¡~√⁠​𝑜/2 ¡→𝑙 ¡~ ¡7 ¡ k? ¡

  • 2. ¡Clustering ¡

method ¡

  • 4. ¡Validate ¡& ¡

Interpret ¡C. ¡

  • 1. ¡Clustering ¡

Variables ¡

  • 3. ¡Number ¡of ¡

Clusters ¡ Several ¡different ¡approaches ¡(Pham ¡2005, ¡Mooi ¡2011, ¡Han ¡2011, ¡Everi^ ¡et. ¡al. ¡2011): ¡

slide-44
SLIDE 44

“Elbow” method

44 ¡

“Elbow ¡Method” ¡(cf. ¡Ketchen ¡1993, ¡Mooi ¡2011): ¡ ¡

  • 1. Hierarchical ¡clustering ¡first ¡
  • 2. Plot ¡agglomeraLon ¡coefficient ¡against ¡number ¡of ¡clusters ¡
  • 3. Search ¡for ¡“elbows” ¡
  • 2. ¡Clustering ¡

method ¡

  • 4. ¡Validate ¡& ¡

Interpret ¡C. ¡

  • 1. ¡Clustering ¡

Variables ¡

  • 3. ¡Number ¡of ¡

Clusters ¡

slide-45
SLIDE 45

“Elbow” method

45 ¡

0.000 ¡ 0.500 ¡ 1.000 ¡ 1.500 ¡ 2.000 ¡ 2.500 ¡ 2 ¡ 4 ¡ 6 ¡ 8 ¡ 10 ¡12 ¡14 ¡16 ¡18 ¡20 ¡22 ¡24 ¡26 ¡28 ¡30 ¡32 ¡34 ¡36 ¡38 ¡40 ¡42 ¡44 ¡46 ¡48 ¡50 ¡52 ¡54 ¡56 ¡58 ¡60 ¡62 ¡64 ¡66 ¡68 ¡70 ¡72 ¡74 ¡76 ¡78 ¡80 ¡82 ¡84 ¡86 ¡88 ¡90 ¡92 ¡94 ¡96 ¡98 ¡

Clustering ¡Coefficient ¡(distance) ¡

<29 ¡ 7 ¡ 16 ¡

  • 2. ¡Clustering ¡

method ¡

  • 4. ¡Validate ¡& ¡

Interpret ¡C. ¡

  • 1. ¡Clustering ¡

Variables ¡

  • 3. ¡Number ¡of ¡

Clusters ¡

Number ¡of ¡cluster ¡k ¡

slide-46
SLIDE 46

Statistical Measure: Silhouette

46 ¡

0 ¡ 0.05 ¡ 0.1 ¡ 0.15 ¡ 0.2 ¡ 0.25 ¡ 0.3 ¡ 0.35 ¡ 0.4 ¡ 0.45 ¡ 2 ¡ 3 ¡ 4 ¡ 5 ¡ 6 ¡ 7 ¡ 8 ¡ 9 ¡ 10 ¡ 11 ¡ 12 ¡ 13 ¡ 14 ¡ 15 ¡ 16 ¡ 17 ¡ 18 ¡ 19 ¡ 20 ¡

Silhoue^e ¡Coefficient ¡

  • 2. ¡Clustering ¡

method ¡

  • 4. ¡Validate ¡& ¡

Interpret ¡C. ¡

  • 1. ¡Clustering ¡

Variables ¡

  • 3. ¡Number ¡of ¡

Clusters ¡ For ¡datum ¡i: ¡ ¡ Compares ¡distance ¡ within ¡its ¡cluster ¡to ¡ distance ¡to ¡nearest ¡ neigbouring ¡cluster ¡ ¡ −1≤𝑡(𝑗)≤1 ¡ Silhoue^e ¡Coefficient ¡s(i) ¡

Number ¡of ¡cluster ¡k ¡ Rousseeuw, ¡Peter ¡J. ¡(1987): ¡Silhoue^es: ¡A ¡graphical ¡aid ¡to ¡the ¡interpretaLon ¡and ¡validaLon ¡of ¡ cluster ¡analysis. ¡In ¡Journal ¡of ¡Computa=onal ¡and ¡Applied ¡Mathema=cs ¡20 ¡(0). ¡

slide-47
SLIDE 47

The Clustering Process

47 ¡

0.335 ¡

  • ­‑1 ¡
  • ­‑0.5 ¡

0 ¡ 0.5 ¡ 1 ¡

Silhoue^e ¡Value ¡

  • ­‑0.40 ¡ ¡
  • ­‑0.20 ¡ ¡

¡-­‑ ¡ ¡ ¡ ¡ ¡0.20 ¡ ¡ ¡0.40 ¡ ¡ ¡0.60 ¡ ¡ ¡0.80 ¡ ¡ ¡1.00 ¡ ¡ 1 ¡ 6 ¡ 11 ¡ 16 ¡ 21 ¡ 26 ¡ 31 ¡ 36 ¡ 41 ¡ 46 ¡ 51 ¡ 56 ¡ 61 ¡ 66 ¡ 71 ¡ 76 ¡ 81 ¡ 86 ¡ 91 ¡ 96 ¡

Silhoue^e ¡

  • 2. ¡Clustering ¡

method ¡

  • 1. ¡Clustering ¡

Variables ¡

  • 3. ¡Number ¡of ¡

Clusters ¡

  • 4. ¡Validate ¡& ¡

Interpret ¡C. ¡ good ¡ no ¡cluster ¡