In today’s competitive world, every enterprise is under pressure to minimize costs. At the same time, they cannot afford to miss the new technologies bus and need to make massive investments therein to meet business demands.

More than 70% of all costs of a Service Desk (SD) are labor-related viz. salaries, benefits, incentive pay, and contractors. The high labor involvement can result in costly human errors and lead to longer response time - this affects the quality of service and leads to lower customer satisfaction.

*Gartner estimates that Ongoing reductions in outsourced head count due to intelligent automation will eventually force sourcing and vendor management leaders to redesign the workplace services for their organizations' users. This will result in a corresponding drop in the numbers of staff required on the service desk, so that when 70 percent of the workload is dealt with by Intelligent Automation Services (IAS), only 30 percent of the staff will remain. 

The WHY

The three drivers that make a case for steering this change to a fully automated service desk are the 3 Es – Economics, Efficiency and Experience.

Economics

Intelligent automation tools can be up to 65% less expensive than offshore-based staff. In the next three years, the costs of commodity services are expected to decline by up to 25% annually, as they move toward this price point.

Efficiency

The best way to reduce majority of the costs in a Service Desk is by optimum utilization of an agent as high agent utilization leads to lower average cost per call. Intelligent automation will play a significant role in complementing humans and enhancing productivity. Below is a graph of how agent utilization can bring in significant cost optimizations:

Experience

End user experience is critical to business success. Customer expectations are soaring and it is essential that these are sufficiently met as today, customers are impatient and exacting. They want quick issue resolution. With the business complexities increasing, service desks are unable to keep a check on incident inflow, which, if not addressed effectively and quickly, causes customer dissatisfaction.

Disruptive technologies like Artificial Intelligence (AI) and Machine Learning (ML) can help enterprises achieve the 3Es. The difficulty, however, lies in discerning the right technology mix. The same is captured in the image below with the “Unmanned Service Desk” journey for all enterprises.

The WHAT

 An Unmanned Service Desk – Completely Automated

When it comes to automation, “Eliminate first, Automate next and Shift-left finally” is the name of the game. The above illustration depicts the high-level ticket flow in a zero/unmanned service desk. By leveraging the right technology, everything but vendor management should be automated through technologies such as:

 Self-heal – agent-based solutions, which will monitor systems real-time and fix the issues immediately. Today, a majority of the automation is reactive with a focus on speed to resolution.

 Self-help – virtual agents (chat bots) assisting end users in resolving their issues and queries by giving contextual answers from the huge knowledge base. End users do not care where the support comes from, all they care for is if the support is there when needed!

 Self-service – creating user intuitive web portals, which are at the backend, connected to intelligent orchestration tools to fulfill end user requests with necessary approvals and integration, will greatly enhance the end-user experience.

The greatest bugbear of IT users, around 30% of all IT helpdesk calls are for password related issues. To address this, Wipro has- internally and across multiple customers, totally eliminated password resets (SAP and AD) using its HOLMES platform with a completely web-based self-service portal, which is accessible from every mobile device.

Auto-Classification and Assignment

Enterprises have suffered from plummeting CSATs and high MTTR (mean time to repair) due to one of the following reasons:

• Analysts (even L3) end up taking a huge amount of time to assign tasks to the right queues
• Wrong resolver group assignments lead to higher ticket hops

By leveraging Machine Learning models that can train on historical data, auto-classifications to the right resolver groups based on confidence scores can be done. Assignment accuracies can reach as high as 85% - 90% in just over 6 months.

Analytics

Service desks collate all issues across an enterprise and hence have tons of data pertaining to the health of the enterprise. Whether it is a CXO who wants to have an operational insight at the click of a button or a cross function manager looking at an efficient problem management, analytics will play a crucial role. Machine Learning models can now be trained on historical data which will enable enterprises to move to the 3Ps- “Predictive, Prescriptive, Proactive” model.

The HOW?

Customer Involvement and User Adoption

With CIOs across industries wanting to transform enterprises and embark on a digital journey, we have often seen that despite a successful implementation of a chat bot or self-service portal, the user adoption is abysmally low - in some extreme cases as low as ~1-5 % after 2 years of implementation! Users still prefer to log tickets through emails, call up the service desk to reset a password and the list goes on. This is because enterprises tend to commit the cardinal sin of not realizing that “automation is a journey”.

Customer involvement plays a very critical role in adoption of these technologies. Organizations will have to educate, engage and enforce users in a smooth and planned manner, which will help phase out older channels of communications and increase adoption through regular communication.

Automation is no more an option but is essential for every enterprise. Before embarking on its adoption, an enterprise needs to:

• Make a proper assessment of its current IT estate from the historical ticket data available to find the top call drivers and prioritize accordingly.
• Set a clear roadmap for automation and align the rest of the team.
• Choose the right technology mix.
• Have regular governance of automation projects to ensure proper maturity and adoption are achieved.
• Estimate the potential opex savings by moving to an AI-enabled service desk and use these estimates to build the business case for the capex allocation for the transformational SD.

In conclusion, it will be the early adopters of this transformation who will reap the maximum benefits. Service desks have evolved and the demands on them will keep getting bigger in the future - some will rise to the challenge, while others will fall behind. Which line will your organization toe?

References

[1] Gartner, Press Release, November 13, 2017 “Gartner Says Intelligent Automation Will Transform Workplace Outsourcing”  https://www.gartner.com/newsroom/id/3827166

[2]  https://www.thinkhdi.com/~/media/HDICorp/Files/Library-Archive/Rumburg_SevenKPIs.pdf

Raj Srirangam - Pre-Sales Solution Consultant

Raj Srirangam is a Pre-Sales Solution Consultant for Wipro’s Automation and AI ecosystem. He transforms businesses across varied industries through infrastructure automation solutions.