How a hypergrowth Fortune 1000 company provides customers support at scale

May 4, 2020 | Automation

One of the leading providers of the Application Infrastructure, Data Center automation, and cloud orchestration solutions created significant value for their customers by accelerating their application performance, simplifying the application management, and delivering applications in a high availability mode. 

With hypergrowth, comes complexity in customer support

The right market fit, and the great value prop propelled the company into a hyper growth mode. The rapid growth of the customer base introduced new complexities associated with the customers application environments.

  • Customer support cases and situations involved highly complex application and infrastructure-related issues. 
  • The hyper-converged customer environments required a full understanding of their infrastructure, application architecture, and in several cases, the engineers needed to search, research and deduce the issues across multiple different providers or components of the architecture or infrastructure of the application to present the solutions and resolve the cases.

While the complexity associated with on-boarding and supporting the customers was increasing, the company did not want to slow down the rate of growth.

Scaling customer support

This posed a significant challenge –

How to scale customer support teams smartly without slowing down the growth and without adding more support engineers, linear to the growth curve? was tasked with solving the situation and was presented with these high- level challenges –

  • Help manage 10+% growth rate in Support Case(per month or year?) volume without having to increase the cost of support in a linear manner.
  • Improve on-boarding time and improve customer’s time to value.
  • Improve customer experience by improving First Call Resolution, Total time to resolve and help deflect customer cases by creating a better self-help infrastructure. engaged with this company in early 2019 and performed an analysis of the Customer Support issues that this company was facing and found –

  • Support Engineers when presented with complex customer issues, used an enterprise search tool to find the information needed to resolve the issue. While the enterprise search tools promised a ‘Cognitive Search’ ability, it failed to capture the full situation and the holistic aspects of the customer situation or the complexity of the issues. 
  • This lack of ability to understand the full context of the issue, leading to a poor set of search results required the Support Engineers to perform multiple searches, browse through a long list of results, read technical documents and consume information from other sources.
  • While this ‘searching’ wasn’t the most optimal way to find the resolutions quickly, it also created a huge amount of loss of productivity –
    • found that the Support engineers were treating every problem as a new problem and were not able to identify applicability of previously resolved issues that were same or similar in nature.  The inability to identify duplicate cases was resulting in huge inefficiencies, given that about over 40% of the issues are duplicates.
    • When the Tier 1 engineers were not able to resolve these issues and the cases were escalated to the next level of escalation, that next level also performed similar type of search – duplicating the efforts of the previous tier and causing further delays. 
    • Resolving customer issues using the ‘search for information’ approach heavily depended on the Support Engineers ability to deduce the resolution from the several options presented.  This was an iterative process requiring customers to try the solutions first before the best resolution could be deployed. exceeds expectations for customer support automation

In the first six months of deployment, was successful in resolving several of these challenges with its Autonomous Support platform.  Quark was able to interpret the customer issues, including their subject, description, and comments, using Deep NLP to automatically recommend resolutions from the reference documents with high accuracy, thus significantly reducing the need for support engineers to determine what to look for, perform several searches accordingly and read different documents.    This helped reduce the time needed to resolve a customer issue rapidly.

  • In benchmark tests done by the customer, they saw that the resolutions presented by were 50% more accurate than the enterprise search.
    • Learning from the previously resolved cases, the platform prevented duplication of efforts and eliminated the need to research every problem as a new problem.
    • Several aspects of BI, analytics-based reports and a full view of the efforts performed at each level of troubleshooting helped reduce duplicative efforts across the support tiers and helped improve the time to resolve the customer issues.
    • By reducing the time spent of the customer cases the company was effectively able to train its engineers for new technologies and new product launches.

Delivering a superior customer experience with

Today, this customer of can onboard its customers faster, delivering support in a manner that helps drive adoption and has significantly improved their time to value. 

The NPS from the existing customers has improved and the company has achieved its objectives and converted its customers into promoters.