Big data was very aptly named; there are already more numbers available than anyone could ever use. Only a minute fraction of this data ever generates legitimate insights that drive actionable findings to create business value. Irfan Kamal writes for the Harvard Business Review about what steps need to be taken to derive more and better insights from the data available to us.
What a Sight
Kamal points to how difficult it can be for major businesses to use data to earn even minor gains. Netflix challenged tens of thousands of people to improve its recommendation engine accuracy by 10 percent, and only two teams were able to accomplish it. In order to achieve goals without eliciting the help of a few of your favorite thousand friends, Kamal recommends a four step data-centered process to mining better insights:
- Collect data from blogs, search, ad engagement, etc.
- Connect data to specific segments or individuals to make it more useful.
- Manage data via special techniques, algorithms, and storage solutions.
- Analyze statistics, reporting, and visualization tools to discover value.
Kamal then segues into three major insights that are still being used today to create value. The first is that friends’ interests make ads more relevant. The second is that it is somehow okay for people to hate your TV show (or similar product), as long as lots of people hate it enough to talk about it. And finally, social media works best in combination:
Measuring the actual business impact of social media and cross-media interactions… could have perhaps the most profound impacts of all on making marketing better and more efficient. For example, by exploring panel-based data on brand encounters by socially-engaged customers in the restaurant industry, Ogilvy and ChatThreads found that social media was very effective in driving revenue from this segment. However, this effect was strongest when social media were combined with other channels such as traditional PR and out-of-home media. Exposure to these combinations drove 1.5x to 2x increases in the likelihood of revenue gains. [source]
Your business might not have a lot to do with television, but the point remains that there are smarter connections to be made with the data we have, if we are willing to study what those connections are. If collecting metrics is your homework, and deriving actionable insights from your metrics means doing homework on your homework, so be it.
Read the original article summary in AITS here>>