Upgrade to Pro — share decks privately, control downloads, hide ads and more …

Data Science and the Games industry

Data Science and the Games industry

I gave a series of talks in 2014 on Data Science in various verticals. This is one from the Game industry.

springcoil

July 07, 2014
Tweet

More Decks by springcoil

Other Decks in Technology

Transcript

  1. How can Data Sciences benefit
    How can Data Sciences benefit
    the Game Industry
    the Game Industry?
    ?
    Peadar Coyle
    Data Science Consultant

    View full-size slide

  2. What opportunities are there?
    What opportunities are there?
    I think there are two interesting areas that spring to mind.
    1) Marketing analytics
    2) Data Science experiments for PR and Marketing purposes
    (internal marketing and external marketing).
    Overall the biggest challenge is growth and cultural change.

    View full-size slide

  3. Marketing Analytics
    Marketing Analytics
    1) Uses gamers play data to optimize marketing
    communications across channels. - Customer segmentation
    modelling
    2) Building Personalization Engine Rules for 1:1
    communications with individual gamers. To help reduce customer
    churn.
    3) Predicts gamers likelihood to churn or to respond to
    up-sell offers.

    View full-size slide

  4. Personalization emails
    Personalization emails
    1) Everyone is different.
    2) Novice gamers need different levels of interaction than non-
    novice gamers.
    For example I suspect novice gamers may respond to encouraging
    emails.
    Experienced gamers may need to hear about Downloadable
    Content
    Activision have about 6 million email templates...

    View full-size slide

  5. Customer Churn
    Customer Churn
    One can build a model of customer segmentation
    There are different kinds of churn - customers not using a
    product versus customers not using a product that is inbuilt
    There are (or should be) key metrics such as net play
    time, frequency of play time.
    Also another interesting thing to measure is 'time to first
    downloadable content' - I am sure there are some intriguing
    models here. I hypothesize if you spend money on in game
    content you are more likely to play the game.

    View full-size slide

  6. Why these models?
    Why these models?
    Better models, allows us to make better decisions.
    I am passionate about using agile software development
    methods, with lots of feedback to create 'data products'.
    These can be reports, automated reports, interactive models,
    data visualizations or data journalism.
    A data product to me is simply something which covers up data
    analysis and presents it in a way to allow decisions to be made.
    Fundamentally our goal is to help customers have a better
    gaming experience.

    View full-size slide

  7. Finally: Data Science and PR
    Finally: Data Science and PR

    View full-size slide

  8. Thank You
    Thank You
    Any questions?

    View full-size slide