Predict churn of a subscriber using a multi-input model to predict churn. Read the full code story here: (https://microsoft.com/developerblog)
For a service business, there are two ways to drive growth: grow the number of new customers, or increase the lifetime value from the customers that you already have by retaining more of them. Improving customer retention requires the ability to predict which subscribers are likely to cancel, and to intervene with the right retention offers at the right time. Recently, the use of deep learning algorithms that learn sequential product usage customer behavior to make predictions have begun to offer businesses a more powerful method to pinpoint accounts at risk. This understanding of an account’s churn likelihood allows a company to act to save the most valuable customers before they churn.
Managing churn is fundamental to service businesses. The lifetime value of the customer (LTV) is the central measure of business value for a subscription business, with churn as the central input. It’s often calculated as Lifetime Value = margin * (1/monthly churn). Reducing monthly churn in the denominator increases the LTV of the customer base. With increased LTV for the customer base comes increased profitability, and with that increased profit comes the economic support to increase marketing activity investment in growing customer acquisition, completing a virtuous cycle for the business.
Predicting that a customer is likely to churn requires understanding the patterns in user state and the sequence of user behavior for churners compared to non-churners. Modeling of these user states and behavioral sequences requires using tabular source data, coming from transaction systems, incident management systems, customer and product records, and then turning these data into a model-friendly set of numerical matrices. This pre-processing of the source tabular data into model-friendly sequences is in and of itself a significant piece of technical work.
For the prediction task we had to choose whether to predict against the attrition event itself or the inactivity that might presage a later attrition. In our modeling approach we predict churn itself. Despite the fact that the actual cancellation decision is a lagging indicator, our model delivered sufficient precision. In future modeling efforts, we will label long-standing inactivity as effective churn.
We defined a performance requirement in terms of accuracy, precision and recall. Once a model performed at the required levels, we could experiment with retention offers to those predicted to cancel, and to do so we needed to deploy this predictive model in a secure and operationally reliable environment where we could retrieve batch predictions daily and drive retention offers and the associated workflow.
We applied a multi-input model construction. We used the Keras functional API and combined the textual and categorical data coming from incident history with the time series transaction and state sequence for the account. The multi-input approach allows us to essentially concatenate these and fit the hybrid model.
Our sequential non-text information is best harnessed in a Bidirectional LSTM – a type of sequential model described in more detail here and here – that allows the model to learn end-of-sequence and beginning-of-sequence behavior. This maps to domain experts' knowledge that distinctive behavior at the end of the subscription period presages churn. And the progress of events over time holds patterns that can be used to predict eventual churn.
On the other hand our textual and categorical data need a separate model to learn from this differently structured data. We have several options here. The simplest option that we developed was to convert the sequence of textual and categorical data, coming from our incident data, into a 1D CNN. We created row-wise sequences of textual incident information for each account, tokenized these words, and applied Glove word embeddings for each. These were right-aligned and pre-padded to better learn patterns from most recent date in order to discern sequences and events that presage churn.
In future iterations, we could add additional input models to learn from non-volatile state and profile information to improve the signal, and categorical embedding modeling for the categorical incident information and other categorical data. We developed and fit each model input independently to understand its performance and adjust the hyperparameters to a good starting point to fit the hybrid model. Then we combine them into a hybrid construction and fit against the two inputs. Find the notebook in the notebooks folder of this repo.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent