DSPs Can Proactively Improve Broadband Performance

67% of customers who contact their Digital Service Provider’s (DSP’s) customer service report broadband issues. Around 26% of customers report inconsistent internet speed causing work from home issues during the pandemic. A recent E&Y report on ‘exceeding customer expectations’, clearly shows that good customer service is a critical element for value creation and customer loyalty. Almost 32% of customers are willing to pay more for stable internet speed in return for good customer service.

Building a 360 degree zero-touch service assurance framework is key to value creation and customer loyalty. It enables continuous remote monitoring to proactively detect connectivity issues and provide automated resolution resulting in a higher Net Promoter Score.

Fig: Zero-touch service assurance framework

This service assurance framework helps Digital Service Providers (DSPs) to proactively solve broadband performance issues and provide improved customer experience. The three-step process that enables DSPs in achieving improved serviceability is detailed below:

1) Transformation of customer data into actionable insights

In the process of data transformation, it is suggested to monitor the customers’ current speed to analyze the event patterns in real-time and compare those to expected behavior. Build a correlation engine that decides target customers for performing the diagnostic procedure, thereby generating a genuine and actionable targeted customer list.

Following are the detailed steps to convert customer data into actionable insights:

Smart monitoring of speed

• Monitor customer’s current speed data on an hourly basis
• Use data science software platforms like Rapidminer, Alteryx, KNIME, which analyzes event patterns in real-time and compares those to expected behavior
• Set up thresholds, based on business needs. Below are the examples for setting threshold –
  • Customers with less than 80% of their current vs. billed speed
  • Decline in rate of speed in the last 24 hours and 1 hour
    • 20% for SMB
    • 40% for residential

Smart correlation 

• Correlate the speed trend with historic customer data
• Check if there are past tickets raised, customer complaints, or pending orders
• Use Python-based correlation engine to create a pipeline for these data and decide target customers for performing the diagnostic procedure

Filtered output

• Generate genuine, actionable targeted customer list which goes as a feed into the diagnostic engine

Recommendations for adding intelligence 

• Use data science software platform for acceleration and implementation of smart monitoring
• Smart correlation avoids acting more than once on the same problematic customer

2) Classify the customer issues identified for the right resolution

In the process of data issues classification, it is suggested to use python scripts to perform loopback tests and classify resolution steps such as port bounce and modem reboot. The test also identifies issues that cannot be auto-resolved remotely. Following is the major resolution classification:

• Category 1: Port bounce
• Category 2: Modem reboot
• Category 3: Issue cannot be resolved through automation. E.g., hardware failure, parent device performance issues, connectivity issues, over-provisioning etc.

Fig: Issue identification using Loopback Testing

Recommendations for adding intelligence

Identify linear usage time using tools like M-curve report

Proactively check the service operation for high profile customers

3) Fix identified issues by executing autonomous actions

For the identified issues, it is suggested to use auto-resolution to fix them even before the customers are aware of it. This can resolve 45-50% of connectivity performance issues, when implemented correctly.

Fig: Scenario of a day in a leading DSP in North America

• Based on the issue categories identified, sets of remediation actions are created like modem reboot and port bounce to improve speed issues
  • For category 1: Modem reboot
  • For category 2: Port bounce
  • For category 3: Issue cannot be resolved through automation. Service request/auto ticket is created
• Actions are triggered when an alert meets specific criteria
• Automatic trigger tasks such as incidents creation, change requests, security incidents, field service work orders, or customer service cases can also be created for issues under category 3

Recommendations

• Do the modem reboot only during the maintenance window so that customers are not interrupted
• Perform resolution status verification to check the performance of closed-loop post-remediation

4) Leverage advanced analytics techniques on data to tap into rich insights through dashboards

Analytics and interactive visualization platform can help DSPs to query, visualize, alert, and understand key metrics to proactively resolve connectivity issues. Use analytics and interactive visualization applications like Grafana, Dynatrace, AppDynamics, etc., which help to visualize data, seamlessly define alerts and thresholds, and generate meaningful insights. The custom dashboards can provide operational and business insights that matter most.

In Conclusion

Digital service providers (DSPs) need to move from reactive to proactive service assurance methodologies to drive significant improvements in customer experience. Below are the benefits that we have seen DSPs achieve, by implementing the three-step process explained in this article:

• 18-20% improvement in NPS score
• Resolved 25% of speed issues through auto-resolution
• Speed improved for 45% of customers
• Issue resolution time reduced from 72 hours to 1 hour

I thank Murugan Chidhambaram – Senior Technical Architect, Archana SR – Lead Engineer, Lalithkiran K – Lead Engineer, and Neha Sehgal – Manager, Strategic Insights for their valuable inputs for this article.

By Muthukumaravel Sakthivel