DSAT Study For A Leading Asian Travel Booking App company



With over 40+ million customer base across Asia, the travel and accommodation aggregator had a huge challenge of DSAT (customer dissatisfaction), their internal reports showed that it has increased from 12% to 18% in the last quarter and they wanted to identify the core reasons asap to improve the CSAT and retain their customers.

  • Measure CSAT with speech analytics
  • Understand drivers of negative customer experience, escalations, and churn
  • Sentiment analysis (positive/ neutral/ negative)
  • Repeat call rate
  • Reasons for non-talk (if over 1 min)

Customer Profile:

An Indian online travel booking company providing online travel and accommodation-related services including flight tickets, domestic and international holiday packages, hotel reservations, rail and bus tickets, etc.

  • No. of Agents: 1200+
  • 5 sites (A, B, C, D, E)
  • 400k calls/ month

Our Approach:

The project was guided by the following set of broadly stated objectives:

  • Ingested a sample of 50k calls, across 5 sites in focus (A, B, C, D, E), call type was “Customer Service” only
  • Agents in focus: 150, Language in scope: English
  • Our team of speech experts developed a DSAT query based on the feedback and suggestions received from the SMEs
  • SmartSpeech POD listened to a sample of 500 random DSAT calls (100 each site) from their banking customer service domain to determine its core call drivers and FCR related opportunity areas
  • SmartSpeech POD prepared the required number of DSAT driver queries and created related reports on the SmartSpeech Analytics platform
  • SmartSpeech team tracked further interesting elements on these call drivers through an exercise of hypothesis validation and shared the insights providing additional value to the client
  • Listed the actionable recommendations based on the opportunities identified to improve the current situation and processes

Our Solution:

SmartSpeech team identified that Site A was receiving the highest number of repeat calls interactions between 15-30 mins of bucket followed by Site C. 53% of repeat callers showed DSAT and 28% demanded an escalation asking for a supervisor.

Top Repeat Call reasons identified:

  • Callback not received (37%)
  • Update not received (21%)
  • Refund status not received (19%)
  • Ticket confirmation missing (8%)

Top DSAT reasons identified:

  • Long refund timelines (42%)
  • Refund not received (29%)
  • Hotel check-in issues (21%)
  • Taxi booking not confirmed (18%)

SmartSpeech team not only identified the top core DSAT and Escalation reasons for the customer service LOB but also prepared relative queries in the system, which will help the client to utilize the speech analytics platform to the fullest as these queries not only just showcase the latest results for a certain topic but also let the user/ analysts play with the platform’s reporting capabilities effectively, the users can also choose to export the data and apply various permutation and combinations over it to derive more complex and useful results. SmartSpeech team has identified the Organizational AHT for DSAT calls to be at 1099 seconds and average Non Talk Time at 24%, and site A was discovered to be handling long calls the most for call drivers like “Refund Not Received”, “Hotel Check-In Issues”, “Room Not as Expected” and “Room Cleanliness” etc.


In just 3 weeks, the list of all identified call drivers for DSAT was shared successfully with the client, required reports were prepared as well for future reference. The consolidated list of other related challenges and recommendations was also shared as a value add. The client said to believe that the study could help them curb the current challenges by the next 2 quarters through a strategic solution implementation plan and can help improve the existing CSAT by 20%, which figuratively results in profits by $2.5m. MattsenKumar proved to be a strategic partner to the client in helping deploy the solutions amongst their quality evaluations team for an overall improvement.