Gartner’s Gareth Herschel talks predictive analytics, social data, and tips for getting ahead of the analytics game.
Q: How have predictive analytics tools become more sophisticated?
A: New models and techniques are always becoming available, but most organizations continue to use the traditional ones, such as decision trees and regression.
Q: Is increased sophistication in analytics simply a function of more data and more processing power, or have the models themselves evolved?
A: In that respect I would say more data and more processing power have been more significant factors.
Q: How can organizations, which may be lagging compared to their competitors’ analytical capabilities, close the gap more quickly?
A: Work with an analytic service provider that already has domain expertise.
Q: Do brands risk losing competitive advantage by outsourcing their analytics expertise?
A: In some respects but not necessarily. For example, a lot depends on how you use the results of the analysis.
Q: Have data quality disciplines kept pace with the ever-increasing supply of data?
A: Quality is always a subjective matter, how good the data needs to be depends upon the purpose it is applied to. I would say that data quality for new categories of data is inferior to that of (some) more established types, but this doesn’t mean it is “bad,” just that it needs to be used with appropriate levels of caution.
Q: There has been a great deal of consolidation among analytics providers in recent years, with many marketing automation vendors being swallowed up by heavy hitters, such as IBM and Teradata. How has that affected the competitive landscape?
A: For every one company acquired by a mega vendor, two or three smaller ones emerge with new capabilities to offer the market. The mega vendors will generally struggle to offer the same targeted functionality as the new vendors, so the suite versus niche dynamic remains in play.
Q: How has industry consolidation affected the affordability of high-end analytical solutions?
A: Consolidation has not impacted affordability as much as the willingness of some organizations to consider options they would not have before. Many organizations prefer to work with well-known providers they feel have the maturity, experience, and support to deliver solutions that they can rely on, particularly for mission-critical solutions or global deployments.
Q: Are brands overweighting the importance of social data in their analytical models because social is the new big thing?
A: I wouldn’t say “overweighting” because the analyst should make sure that doesn’t happen as part of the process. Social data also covers a wide variety of different types of data. I would say that because it is the new big thing, organizations are very keen to experiment with social data in their models to see if there is any value there.
