Sources of data for multi-channel analytics
The sources of customer interaction data are vast. Interactions can come from internal (i.e., company-solicited) and/or external (i.e., consumer-generated) feedback sources. Much of this data—perhaps 80 to 85 percent—is unstructured, meaning that it lacks a defined, standard structure that would enable simple quantification.
Each of the two dimensions shown above—structured/unstructured and internal/external—provides its own set of challenges and benefits
- Voice Calls
- Focus Groups
- Social Media
- Word of Mouth
- Customer Surveys
- Agent Disposition
- Quality Scores
- Online Ratings
- 3rdParty Market Research
Internal – Unstructured Interactions
- Provide immediate, unbiased feedback from customers about your company.
- Offer great depth of insight, due to free-form nature of conversations.
- Traditionally more difficult to analyse due to unstructured nature of the data.
- Can provide early warning signs of significant service issues before they “go viral.”
External – Unstructured Interactions
- Provide unbiased dialog between customers about your company.
- May have impact on other customers and even your competitors.
- May be difficult to distinguish real customer insight, from non-relevant “noise.”
- Traditionally more difficult to understand full context of the interaction.
External – Structured Interactions
- Provide quantifiable measures of overall customer ratings.
- Offer wide scope, but can be less actionable because it is harder to identify root cause without the full context.
- Supported by mature analysis methods. Most useful for benchmarking and future planning.
Internal – Structured Interactions
- Typically provide quantitative analysis on key internal metrics.
- Can help answer specific questions, but can miss important new emerging trends.
- Drive operational decisions.
The balance between structured and unstructured sources is rapidly shifting to a significant increase in unstructured data sources. These have traditionally been harder to mine and quantify, but are very rich in information and customer insight. This leaves the analyst who wishes to create a complete picture of customer interactions the daunting task of making sense of disparate data sources and formats.
- Debbie Hage of Verint