5 Keys to Implementing an Omni-channel CEM platform

By: Clarabridge Team

June 1, 2015

by Shorit Ghosh, Director of Enterprise Solutions Services

Today’s customers provide feedback in a myriad of ways. They can use surveys and call center channels to talk to you, review sites to talk about you and social media channels to rant and rave about your products and services.  The enterprise today must be equipped to not only collect and listen to all this feedback but also intelligently analyze (and operationalize) the findings.

There are a few keys that enable this intelligent multi-channel listening for today’s enterprise:

1. Ingesting data from multiple listening posts: If you cannot collect all your data, you cannot act on it. Select a technology partner that has the ability to source data from all of the enterprise’s listening posts. These sources may be as varied as voice transcriptions and flat text files from a call center, direct API connections to rating and review sites, connectors to social media sites like Facebook and Twitter as well as social media aggregators and web scrape mechanisms to scour through obscure and not so obscure websites for content related to your industry, competitors and products.

2. Listening to both structured and unstructured feedback: Today’s listening posts are increasingly text heavy (social media, call center, web forums) and less focused on traditional sources like surveys that provide you structured scores. It is important that your technology partner should be able to process both text and structured feedback. In fact you should be able to overlay structured scores on text feedback to truly understand the drivers of customer satisfaction. Consider the scenario where you are mining your call center transcriptions to understand top volumes of conversation and then overlaying Net Promoter Score/Overall Satisfaction Score (NPS/OSAT) numbers on top of these same conversations to understand if they are impacting your satisfaction scores.

3. Tuning sentiment by source: Your customers will talk to you very differently depending on the source you are analyzing. In a truly omni-channel environment, you should have the ability to modify sentiment analysis to account for highly sentiment bearing social media, not-so-sentiment-bearing call center notes, and the biased-sentiment survey sources (think survey questions that bias responses like “What did you like about your experience?”).

4. Analyzing common data points from varied sources: Many times you will want to analyze data holistically across all sources. For example, you may want to analyze all feedback from your high net-worth customers or from a particular region regardless of the source. Many times these “high net worth” or “region” data points are called very different things in your various data sources. It is important for you to be able to funnel these data points into a common cross-source reportable attribute to be able to report holistically and get a true omni-channel view of your customers.

5. Analyzing feedback in multiple languages linguistically: Today’s customers speak to you in multiple languages. Demographic trends within the US as well as rapid international growth have made it imperative for enterprises to listen to customers in the language of their choice. In fact, your CEM technology should not only be able to listen in multiple languages but be able to understand linguistic connections, sentiment, taxonomies, and concept rules in these languages. Additionally– you should have the ability to surface trends holistically regardless of language to understand a complete picture of your customers’ feedback.

These five keys provide the basis for a robust, intelligent omni-channel platform that can turn raw data into actionable insights.