Basics of Uplift Predictive Analytics Models
So how do you know that the customer you targeted using predictive analytics wouldn’t have purchased anyway? To clarify this question, you can restate it in a couple different ways:
How do you know the customer wouldn’t have purchased even if she didn’t get the marketing contact from you?
How do you know that what you sent to the customer influenced her to make the purchase?
Some modelers claim that the problems with response modeling are as follows:
You’re taking a subset of your customers whom you’ve predicted will have some interest in the product or service already.
You’re wasting marketing dollars on customers who don’t need the extra influence to convert.
You may be decreasing your net margins because the discounts you’re using to entice the customer to buy may be unnecessary.
You may be reducing your customer satisfaction because some customers don’t want to be (constantly) contacted.
You’re incorrectly taking credit for the response in your evaluation of the model.
Uplift modeling, also called true lift modeling and net modeling among other terms, aims to answer those criticisms by predicting which customers will only convert if contacted.
Uplift modeling works by separating customers into four groups:
Persuadables: Customers who can be persuaded to purchase — but will only buy if contacted.
Sure Things: Customers who will buy, regardless of contact.
Lost Causes: Customers who will not buy, regardless of contact.
Do Not Disturbs: Customers whom you should not contact. Contacting them may cause a negative response like provoking them to cancel a subscription, return a product, or ask for a price adjustment.
Uplift modeling only targets the Persuadables. That sounds promising, but an uplift model has proven much more difficult to create than a response model. Here’s why:
It generally requires a larger sample size than for response modeling, since it has segmented the sample into four groups and only uses the group of Persuadables. It then has to be further split up for measuring the effectiveness of the model.
This group will potentially be much smaller than the target size for response modeling. With a smaller target size and complexity, however, the operating effort and cost may not justify the use over response modeling.
It’s difficult to segment the customers perfectly into those four distinct groups, just as it’s hard to measure the accuracy of the segmentation.
It’s difficult to measure the success of such a model because it’s attempting to measure change in a customer’s behavior, not the concrete action of whether the customer purchased after receiving contact.
To measure a single customer’s behavior accurately, you would (in effect) have to clone her and split the identical clones into groups. The first (treated group) would receive the advertisement; the second (control group) would not. Setting aside such sci-fi scenarios, you have to make some concessions to reality and employ some alternative (more difficult) methods to get a useful estimate of the model’s success.
Even with these difficulties, some modelers argue that uplift modeling provides true marketing impact. They consider it more efficient than response modeling because it doesn’t include the Sure Things in the targeting (which artificially inflates response rates). For that reason, they feel uplift modeling is the choice for target marketing using predictive analytics.
Uplift modeling is still a relatively new technique in target marketing. More companies are starting to use it and have found success using it in their customer retention, marketing campaigns, and even presidential campaigns.
Some pundits are crediting uplift modeling for President Obama’s 2012 presidential campaign win. The campaign’s data analyst used uplift modeling to heavily target voters who were most likely to be influenced by contact. They used personalized messages via several channels of contact: social media, television, direct mail, and telephone. They concentrated their efforts to persuade the group of Persuadables. They invested heavily in this strategy; apparently it paid off.