Vtiger Prediction for Potential Customer

Using Machine Learning to predict Potential Customer 

Early we discussed about Lead Scoring in Vtiger, today we discuss about Vtiger Predictions and how it would be used to predict Potential Customers.

What we mention here is called as Supervised Learning which could be used for Classifications and Predictions.

In simple term we Train system with historical data and then create a Data Model which is used to Predict an outcome for a given set of attributes.

We use this concept to predict if a Contact is a potential customer. We have a historical data of our customers (around 3000 records) with details like age, education, income etc., with identification if they bought a product from us or not in last 2 years.

We use BigML to build our Data Model. Which would be then used to do Predictions.

Our historical data:


Similarly we have created custom fields in our Contacts Module.

Suppose we have imported a new list of Contacts captured from a Tradeshow and would like to predict if they are Potential Customer and what is the confidence in it.

Our Contacts in Vtiger CRM:


Check the first 2 columns “Prediction Confidence” and “Is Potential Customer”. Prediction confidence field indicates the confidence of the prediction made, which is the second field. We have implemented a job which takes this data and does the Prediction through API calls to BigML.

We can generate various Models in bigML and choose our algorithm for it, set the threshold etc.. A sample model:


Now we run our job which would perform the Prediction. Check the below screen shot on how the values for Confidence and Is Potential Customer have changed.


You can observe that the system has predicted that Jeff is a potential customer and the confidence for that is 59.64. Which is considered good.

Based on the scores your Sales person could focus more towards engaging that Contact and can save their valuable time.

Prediction and Machine Learning could be used for various application in CRM like:

  • Customer churn analysis
  • Lead Scoring
  • Sentiments Analysis
  • Social data Analysis
  • Duplicate record matching
  • Fraud detection
  • Sales Forecasting

In our next blog we would show you how we could use Vtiger with BigML for Sales Forecasting.

Let us know if you would be interested in improving your sales and customer retention using Predictions.


No Fields Found.
Please follow and like us:
Social media & sharing icons powered by UltimatelySocial