Marketeers have been using Predictive Analytics long before the advent of Marketing Automation Platforms. However the knowledge gained and decisions were limited by the manual scale of processes that followed. Now that we are using always on campaigns and large multi-track nurture programs, we can track individuals from 1st touch through to customer and beyond, with all the activity in between. However the decisions we make are still often based on intuition - "what worked last time?" - and response metrics. A company that can properly apply the full range of predictive tools to their marketing operations will see benefits beyond simply segmenting prospects.
Predictive Analytics can be used to determine the exact point that a record should become Marketing Qualified. Instead of simply analysing a set of previous marketing qualified records to produce a static scoring model, allow the model to learn on the fly, selecting prospects that are most likely to convert but also selecting prospects which provide the most value to learn from. This way leads you are passing to sales are exactly the ones that are most likely to convert, and as behaviour and key target markets change, your lead selection adapts on the fly too.
However, this is not the only benefit we can gain. Analysing the resulting models can give insight into the success of your marketing activities. A true picture of ROI can be developed where the effect on likelihood to convert from each campaign can be reviewed in the context of your entire marketing environment.
Likelihood to convert is not the only result for which it is beneficial to predict. We can also predict likely lifetime customer value and the time taken to convert. Combining all of this information enables the marketeer to look at the current state of individuals in all stages of the marketing and sales funnel and predict how many are going to convert, when that might be and how much revenue is expected. Combined, this gives a complete picture of the financial contribution of Marketing to the business going forward.
There are some caveats that should be introduced before you jump in. Whatever people may say, computers are still dumb. To be ready for predictive analytics all of your marketing activities need to be organised, standardised and with clear a naming convention. Also keep the end predictive goal in mind. This is not something that can just 'plug and play' - it needs to be a core consideration between multiple stakeholders when devising both your marketing and data strategies.
With predictive analytics, marketers use data science-based techniques to segment databases, determine which behaviors and attributes of a company or individual are statistically predictive of desired outcomes, and then identify a pool of prospective customers that match those characteristics. "You're leaving the realm of making assumptions on what's important and letting data tell you what matters," he said.