Understanding Keyword Detection in SupportLogic Trend Analytics

Overview

SupportLogic’s Trend Analytics and Keyword feature allows users to identify and track specific keywords within case data. This functionality helps surface patterns and topics of interest across inbound/outbound messages and case notes.

How Keyword Detection Works

SupportLogic detects keywords by scanning:

• Inbound (IB) Messages: Customer messages within the case.

• Outbound (OB) Messages: Agent responses in the case.

• Case Notes: Internal notes added to the case.

If a keyword appears in any of these areas, the case will be flagged for that keyword in Trend Analytics.

Example Scenario – Keyword “Report”

A recent investigation revealed that cases were being flagged under the keyword “report” even though the word did not appear in the customer’s conversation.

• The root cause was an automated bot message containing the phrase:

  “BOT: If you’ve already reported an outage…”
  

• Although the keyword was not present in the customer or agent messages, the bot’s automated response included the word “reported,” which triggered the keyword match.

Recommendations for More Accurate Results

To achieve more precise and relevant search results:

1. Use Multiple Keywords: Combine related keywords (e.g., “report + outage”) to filter out cases where the keyword only appears in automated responses.

2. Review Case Context: Check the Bot responses, internal notes, and automated templates to identify if keywords are being triggered by system-generated messages.

3. Apply Additional Filters: Utilize filters such as date range, product category, or priority to narrow results further.

Additional Notes

• SupportLogic’s keyword detection is case-sensitive only to word presence and does not distinguish between customer-entered text and automated messages.

• Engineering is exploring potential enhancements to differentiate between bot-generated content and actual customer/agent messages.