Predictive Anaytics Business Applications
As the quantity of irrelevant information has exploded online, so too has the market for the delivery of targeted offers and information. Predictive analytics, a branch of data mining concerned with predicting future probabilities and trends, applies a filter to users’ online interactions with the aim of delivering more value from a sea of irrelevance.
Predictive Analytics / Analysis for Recruiting
- There are many recruiting sites on the web which promise to be able to match candidates with job requirements in unique and increasingly accurate ways.
- Predictive analytics is at the core of their business model, as it automates the process of making these matches.
- When a recruiter posts a job description, a predictive algorithm runs through candidates and calculates compatibility. The technology is, in many cases, embedded in search applications. The most accurate and efficient of these analytics will deliver the most value and see the greatest adoption over time. Those recruiting and talent acquisition sites that allow businesses to leverage the existing networks of their current and former employees are the best positioned to monetize their users’ employment data in new ways. Businesses can get value from these existing networks without the time and resource commitment it takes to build their own.
Predictive Analytics / Analysis for Sentiment Analysis
- As sites like Twitter and Facebook gain value to the business world, many companies have started analysing and establishing collective online sentiment and also identifying individuals with influence and authority.
- Companies including Klout, ViralHeat and Radian6 all scan blogs and other social media channels with predictive models to determine if the content surrounding a brand or person is negative, positive or neutral.
Predictive Analytics / Analysis of Market Fluctuations
- Day traders, retail investors and analysts are cruising around on Twitter and Facebook.
- New models are cropping up to predict stock fluctuations based on Twitter posts. Similar to sentiment analysis, these companies are able to look at the total number of tweets, as well as positive and negative comments to predict whether a stock price will go up or down. These types of companies will become a hot commodity as investors begin to rely on the wisdom of crowds.
Predictive Analytics / Analysis for Recommendation Engines
- The more active you are online, the more effectively predictive analytics can work to deliver targeted and relevant offers. E.g., Facebook offers are no longer random and are therefore increasingly effective.
- Leveraging the existing data from your previous activity to predict what will happen in the future is becoming, rightly, more prevalent and valuable to social networks that can sell this promise to businesses and intermediaries.
Predictive Analytics / Analysis for Location-Based Marketing
- As social networks add in more location-aware features like Facebook Places and whole new businesses are built on the promise of geo-location including SCVNGR and ShopKick, predictive analytics deliver insights into where groups and individuals will be and when.
- Social networks can run their existing location data through predictive models to provide companies with future insights into where to allocate their marketing and advertising budgets for the biggest returns.