Keeping the Customer/ Reducing Churn

The reasons that customers leave are many, and typically addressed retrospectively, or at the time that the customer has already decided to leave at which point any incentive to try and retain the customer may be futile, costly and often worsen the relationship even further.

The reasons a customer has left may appear obvious in hindsight.  Declining spend, contract expiry, negative interactions with support, frequent credits.


What if we know the customer is unhappy

The best time to intervene and retain the customer is before they have made the decision to transition to a competitor.  Within modern CRM systems, almost every interaction and activity of the customer is documented, and within these millions of records lies a similar number of clues as to how the customer truly feels about your services.

For a service agent to go through every account and read every case note, every transaction, every call for every customer in order to determine if a customer is ‘at risk’ of leaving is probable but in reality not possible due to cost, and the approach likely flawed due to bias, fatigue, or the sheer repetitiveness of the task.


Artificial intelligence and machine learning

What we can do, in replacement of service staff monitoring customer records, or workflows that trigger on system events such as contract expiry, is use artificial intelligence techniques to look at data in near live-time, making decisions on the fly about your customer, using insight and trends/patterns that you wouldn't have even guessed existed in order to predict if/when your customer is thinking about leaving.

For example, our AI systems may discover a correlation between customers who have a contract term that expires 2 weeks after christmas, of whom have recieved more than 2 negative sentiment calls in the past 6 weeks, and a heightened chance of leaving.  The models then update your CRM system, alerting retention teams to jump into action and try to incentivise the customer to stay.



hoW this fits in with you

Our Keeping the Customer API works directly with the world's largest CRM platform: Salesforce.  

Customer privacy is of utmost concern and we will work with you to see what data we can and cannot have access to.  We then perform data science on the set, inline with your customer definitions, and let you know which of this data is useful to us and which isn't.

We then take this data and process it on some very high performance computers.  We extract sentiment from text and find patterns in spend, contract lengths, regions and anything else that may be unique to your business.  The result is a trained AI model of which has accuracy in predicting customer churn of 90% and up. 

Then, Daily/Weekly/Monthly (depending on your needs) fresh data from your Salesforce instance is put through the model, and the model predicts the likelyhood of your customers leaving.  Customer status is updated via the Salesforce APIs for actions within CRM.

The model undergoes continuous improvement, learning as it goes.  You don't have to engage change management teams to discover new business flows, the AI does it for you, discovering new trends as the year progresses!