Generalized Approaches to Analyses

These generalized best practices can help you start making inferences about your data.

All approaches start with the first basic step: Look at your data.

Data Anomalies

  1. Look for dips or spikes (which we call "anomalies") in a broad view of your data, on the Business Intelligence page, for example.
  2. Focus on one of the anomalies.
  3. Reduce the time range of your view into the anomaly.
  4. Filter the view by each of the various dimensions.
  5. Look for correspondences among the dimensions, such as other dips or spikes that occurred near the same time as your target anomaly.
  6. Plot the anomaly against other dimensions in the same time period.
  7. Ask yourself many questions about the data, not limited to the following:
    • What conclusions can you draw?
    • What correspondences with other dimensions caught your eye?
    • What major public events happened during the same time period, such as sports or political events?
    • What events happened in your business that could have caused the anomalies?

Areas of Interest

This is the method we use in our examples.

  1. Study the list of Ooyala IQ's dimensions in the Overview to Ooyala IQ.
  2. Which areas of your business are you interested in? Engagement? Sales? Device coverage?
  3. Of the predefined dimensions, which most closely match your areas of interest?
  4. Follow one of the examples, but investigate the dimensions you are interested in, rather than those given in the example.

해당 내용이 도움 되었습니까?