liu.seSearch for publications in DiVA
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • oxford
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Överlevnadsanalys inomen streamingtjänst: En jämförelse i risk mellan abonnemangstyper
Linköping University, Department of Computer and Information Science, Statistics.
Linköping University, Department of Computer and Information Science, Statistics.
2016 (Swedish)Independent thesis Basic level (degree of Bachelor), 10 credits / 15 HE creditsStudent thesisAlternative title
Survival analysis for a streaming company : Risk comparison between subscription types (English)
Abstract [sv]

Marknaden för streamingtjänster växer i snabb takt. Därför eftersöks sätt för att på ett korrekt sätt ta reda på om det finns bakomliggande faktorer till att en abonnent avslutar sitt abonnemang vid en viss tidpunkt.

Syftet med uppsatsen ligger i att analysera när en abonnent avslutar sitt abonnemang. Samtliga startpunkter för abonnemangen är kända men inte alla abonnemang har vid studiens slut avslutats. Med anledning av detta väljs överlevnadsanalys då denna har möjlighet att hantera censurering och även jämföra risken att avsluta sitt abonnemang med hjälp av förklarande variabler. Då det finns abonnenter som har avslutat sina konton mer än en gång kommer överlevnadsanalyser som tar hänsyn till multipla händelser att användas.

Resultaten från studien visar att ökad tittning leder till ökad risk för churn för samtliga abonnemangstyper som undersökts i denna studie. Kwong-Kay Wong (2011) påvisar att en abonnent som bytt abonnemang till ett mer passande abonnemang har en lägre risk för churn. Då de flesta abonnenter som efter churn återkommit, kommer tillbaka till samma abonnemangstyp kan inte samma slutsatser dras. För abonnemangstyp 4 ökar risken för churn om abonnenten har abonnemangstypen som sitt andra eller tredje abonnemang, jämfört med om abonnemanget varit abonnentens första abonnemang i ordningen. Abonnemangstyp 4 har en ökad risk för churn jämfört med abonnemangstyp 1 givet att abonnenterna har tittat lika mycket. Resultaten säkerställs genom användning av LASSO.

Abstract [en]

The market for streaming services is growing rapidly. With a growing number of subscribers, companies are now looking for efficient ways to find which underlying effects that has effect on when a subscriber will churn. The purpose for this paper is to with the use of viewings per month, type of package and number of packages in the time period see if the risk for churn differs between the groups.

The data contains information about all started subscription for four different types of subscription-based-packages for a single streaming company. Information about all subscribers starting dates are known. Due to the fact that not every subscriber has churned at the end of the study, the data is partially censored. We use Cox proportional hazard model to handle this problem. Multiple failure models are used because some subscribers have multiple subscriptions during the study.

The study shows that with increased activity on the site, the risk for churn increases for all four different types of packages. Kwong-Kay Wong’s (2011) study shows that a subscriber who has changed its plan to a more optimal one has a lower risk of churn. In their study the switch of plan was to a known better plan. In our study we see that most subscribers who return after churn comes back to the same type of package that they left, and the results will therefore be interpreted differently. A subscriber who has his second or third subscription in package 4 has an increased risk of churn if they have the same amount of activity on the website compared to if it would have been the subscribers first subscription. We also find that package 4 has an increased risk of churn compared to the risk of package 1, given that the two subscribers have the same amount of activity.

Place, publisher, year, edition, pages
2016. , 38 p.
Keyword [sv]
Överlevnadsanalys streamingtjänst
National Category
Probability Theory and Statistics
Identifiers
URN: urn:nbn:se:liu:diva-129787ISRN: LIU-IDA/STAT-G--16/006—SEOAI: oai:DiVA.org:liu-129787DiVA: diva2:943770
Subject / course
Statistics
Presentation
2016-06-07, 11:00 (Swedish)
Supervisors
Examiners
Available from: 2016-06-29 Created: 2016-06-28 Last updated: 2016-06-29Bibliographically approved

Open Access in DiVA

fulltext(1513 kB)70 downloads
File information
File name FULLTEXT01.pdfFile size 1513 kBChecksum SHA-512
eae5ff356952e01e4d0b87d59e238e89aebca9014cf6bdb3daba67a8dde208f00751b1f05d65c52bd2e822aeda2c6413f3ebbfc2598f9737ed14ce58986d73b9
Type fulltextMimetype application/pdf

Search in DiVA

By author/editor
Eriksson, JosefWretström, Jonathan
By organisation
Statistics
Probability Theory and Statistics

Search outside of DiVA

GoogleGoogle Scholar
Total: 70 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

urn-nbn

Altmetric score

urn-nbn
Total: 346 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • oxford
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf