Probably one of the biggest mistakes publishers can make on converting trials is relying on experience and intuition. The art of implementing and managing trials doesn’t match the science of new data-driven techniques. Using predictive analytics to qualify trial users and focus on those that are most likely to convert can double conversion rates. In a 2012 study, the Aberdeen Group published a finding that companies using predictive analytics have a 73% sales lift versus companies that did not. In a previous note on trials, loyalty was described as a good measure for predicting conversions. This post expands on the loyalty measure to demonstrate how to implement trials scoring that increases conversions.
Trial users can be measured and scored any number of ways such as number of active days, range of active days, highest activity, total activity, and more. Predictive analytics increases trial conversion rates, because it quantifies the predictive effectiveness of a measure for conversion allowing a publisher to implement accurate trial scoring rules.
A cumulative gains chart illustrates the assessment and selection of timing in trial scoring based on effectiveness of predicting a conversion. A gains chart visualizes “lift,” predictive performance, as a ratio between using a specific predictive measure in trial scoring versus randomly attempting to convert trials (i.e., the baseline). If a particular measure is positively correlated, its lift curve will be above the baseline or negatively correlated if below the baseline. The delta between the baseline and the lift curve is the “lift.”
The second gains chart illustrates why high usage of the trial on the first day is a poor predictor. The lift is actually negative in every analysis Scout Research has performed. Factors influencing this may be that many trial users sign up before they need to use it while on the flip-side, many abusers make heavy use of the trial as soon as they sign up.
The third gains chart illustrates why threshold of active days is a strong predictor. The gains chart for threshold on active days typically shows a lift of 100-200 percent. There seems to be a tipping point at which trial users have seen enough value to consider moving to a full subscription.
In addition to threshold, interval period is often a strong predictor. Many users are not ad-hoc in their utilization. They have specific days of the week and times of day for completing tasks – and it shows up in a trial. Finally, active last day of trial is a strong indicator. The gain versus random calling is typically several hundred percent.
All of these measures are easily evaluated and assessed for predictive performance. Once quantified, they can be integrated into trial scoring to increase conversion rates and revenue.
Publishers should use predictive analytics to develop trial scoring rules. These scoring rules can constantly prioritize trials in their likeliness to convert which increases close rates and sales productivity. These same predictive analytics are useful in design of trial parameters such as length and access limits. In the age of big data, trial scoring and conversions become more science and less art.