Prediction Studio now offers multistage predictions that you can use to calculate business outcomes that occur one after another, over a longer period of time. For example, when deciding which web banner to display to your customer, you can take into account not only how likely the customer is to click the banner, but also how likely they are to accept the corresponding offer and convert. This way, you can address the customer with an offer that they are most likely to accept in the long run, and in consequence achieve your long-term business goals.

Multistage predictions calculate the probability of your customer performing two subsequent actions by multiplying the customer's propensity to perform one action by their propensity to perform the other action.

Single-stage vs. multistage predictions

Conversions are long-term outcomes that you can predict in many ways, depending on your business needs and the data that you have. To decide whether a single-stage or a multistage prediction is the better choice for you, your team should answer the following questions regarding your customers, data, and storage:

Data scientist

• Are the customers who click and the customers who convert similar or different? 
  Conduct a data science analysis to compare the profiles of both groups:
  • If the profiles are similar (except for volumes), you can predict conversions with a single-stage prediction. Clicks are often a good proxy of conversions, so you can use a click model. Alternatively, if you have access to conversion data, you can use a conversion model.
  • If the profiles are significantly different, consider using a two-stage prediction to assess which customers out of those who are likely to click an ad are also likely to convert.

System architect

• Do you have access to conversion data?
  You might have to retrieve conversion data from another system, for example, a fulfillment system. This might pose technical challenges, such as a longer waiting period.
• Can you retain impression data until customers convert? 
  In systems that process large volumes of data, it might not be feasible to keep the impression data long enough for customers to convert.
   

For more information, see Multi-stage predictions and Creating predictions.