Text mining your customer feedback for patterns

Structured or rated/scaled feedback is easier to manage than unstructured comments and customer feedback. While many companies get a scaled and organized report of how they doing with their customers, reading through the text comments from customers is cumbersome and most organizations don’t have the time and resources to allocate to this effort even though it is a valuable source for product/service improvement and customer retention. I provide a few simple steps that can help you automate this task and also ensure customer comments are routed to the right team for action.

Setup an Index/Collection of your customer comments

If your comments are are logged anywhere then they can be indexed and searched. Setting up a collection to index your customer feedback along with customer or product attributes can help you not only search them but also drill down on the specific comments to see if there are patterns with certain products or issues that have cropped up during a timeframe. Whether your customer comments are in Salesforce, databases, flat files or on your website, indexing them for search is the first step towards enabling any type of action. The comments once indexed can be setup as a tag cloud so analysts can watch the tags grow or shrink based on the terms used in the comments.

Setup alerts on customer comments

Identify keyword triggers that you want to be notified on when a customer comment is placed. Trigger words could be “dissatisfied”, “bad”, “awful” or “service” or anything that could be unique to your business. You can setup keywords for specific departments if they want to be notified when certain products are mentioned or a certain product ID is used. These triggers can automatically alert an email address so they can go through the comment in entirety rather than reading through all comments and figuring out if the comment applies to them. Alerts work similar to news alerts where you get notified when a topic or term of interest is encountered by the automated system.

Get notified when key phrases or trigger language is used on the comments

Once an alert is triggered then an email message is sent to the right email address to notify that the new comment is awaiting a review due to the term used. Multiple email addresses can also be setup within SearchBlox with a combination of different types of alerts so the organization can effectively deal with the feedback. The notification process can enable deeper analysis and reporting as required.

Just using the above steps can automate your customer feedback analysis process as well as provide insights into your customers.

Learn more about how you can gain insight into your customer comments by contacting us. SearchBlox can help your text mining effort and help you identify patterns in customer feedback.