Chapter 3:
Putting theories to the test

What went wrong?

Most customers believe in second chances and don't jump ship right away—unless something severe occurs that affects them negatively. In this segment, we'll review a few real-world examples of poor customer support. We'll analyze the series of events that snowballed into an overall drop in quality of service and resulted in a loss of revenue for these businesses.

1. Customer support for dummies idiots

After a dispute with their wireless network operator, one customer found their account nickname changed to "Idiot idiot". According to the customer, they had been trying to contact support every five days to get information on credits that weren't added to their account. Presumably, after one such interaction, an irate customer rep changed the username, unaware that it would be displayed for the customer as well.

2. Press one for...

An all too familiar situation, almost a rite of passage, is the permanently inactive or busy customer support line. This isn't an isolated incident. Social media is filled with complaints from frustrated customers trying to contact businesses through the numbers on their websites, only to find the numbers either disabled or ignored. After a few failed attempts, some customers opt for competitor products or services just for their prompt customer support.

3. Infinite transfer loop

A customer was facing technical issues with their online shopping application. They ordered a product through the app but were facing unusual delays in delivery. The app offered an option for a refund. However, the customer was unable to claim the refund, and the support staff only offered template responses and no real resolution. When asked for further explanation, the customer was transferred to another agent. This agent transfer action occurred multiple times. Finally, the customer expressed their concerns on social media. After a couple of months and multiple back and forth emails, the issue was resolved and the customer received their refund.

A former customer service representative later admitted online that employees don't often stay with the company for long. With limited work experience, agents are expected to handle scenarios, most of which vary from the theoretical scenarios taught in the training sessions. They also shared that some product features were not addressed during the training period, making it tougher for agents to identify a resolution.

What successful customer support looks like

Regardless of whether an organization sells products or offers services, nothing is foolproof. ManageEngine's road to becoming a sustainable enterprise is paved with some unsuccessful experiments and minor missteps. However, what sets an organization apart is its ability to take these errors in stride and turn them into learning opportunities. Our customer support is no different. We know our products are a constant work in progress, as we are continually trying to solve new business problems for our customers. This also means we hear feedback—some that came from a place of frustration. Let's take a look at a real example of how ManageEngine's support team was able to navigate through a challenging situation.

In this scenario, we'll review the steps taken by the ITSM (Cloud) product support team. First, the roles. At the time of inception, the team had six members. Today, they stand tall with over 80 members managing ITSM support worldwide. The team is comprised of:

• Customer support manager
• Product specialists
• Technical support engineers
• Support coordinators
• Community specialists
• Customer success representatives
• Technical trainees
• Help desk coordinators

New recruits engage in a two-month immersive training session—a concept backed by the global head of ITSM support at ManageEngine, who firmly believes practical sessions provide more value than theory lessons alone. This training period is split into:

• A three-week conceptual training session where trainees study the framework that bolsters the product and understand why each feature is built.
• A two-week technology-related training session to understand the operations.
• A nesting period where they are assigned minimal support cases (L1) under the supervision of a lead support engineer.

Following this, trainees are encouraged to go into full production and take on support requests—emails first, followed by chats and calls when their supervisor is assured of their product and domain understanding.

In the early days (over a decade ago) of our flagship product, ServiceDesk Plus, there were times when the product was not stable, as with any growing tool. Occasionally, the application would slow down and users experienced performance issues. This was a problem, especially when it affected customers from major industries like airlines and banking, where time is of the essence. Unfortunately, we were unable to review the logs because we didn't have the mechanism to collect relevant logs at the exact moment of impact. By the time the customer reached out for a remote session, the application's performance had often returned to normal.

There were instances where the support team was on call with the customer's IT team for hours just to be able to collect the logs at the precise time the performance issue was happening. This level of engagement often exceeds the time allotted for a support request, so it can disrupt other support operations.

When this issue occurred again, the support team realized it was time for a focused internal collaboration. The technical support engineers collected customer feedback and worked with product specialists and the customer support lead to brainstorm solutions. The product engineering team was then notified and provided its input on how the tool could be adapted to collect logs related to the performance issue more efficiently. How does it work?

However, when an application is experiencing performance issues, one log file isn't enough for analysis. The team requires five to six types of log files to compare, analyze, and identify the root cause of the issue. The engineering team came up with a feature called performance debug URL. When a customer is facing performance issues, there is a URL the customer can utilize to efficiently collect the log files.

Once this new tool was implemented, the support team began monitoring incoming tickets to verify that it was accomplishing its job. As anticipated, many customers found this URL to be useful, and it was described as a resounding success. The support team shared their feedback with the product team, and decided to incorporate this URL as a feature in the product. For example, if the application is found to be consuming more CPU or memory than expected, now the URL will automatically activate and capture logs to determine the root cause of the performance issue. Even if customers can't or don't reach out in time and advise about a performance issue, the support team can still obtain the applicable logs as they've been recorded by the application. Utilizing this helpful feature, the support team can now quickly and efficiently identify the root cause and fix it promptly for customers.

Being the voice of the customer

(i) Customer success management (CSM)

ManageEngine's customer success program strives to foster long-term relationships with customers. The CSM team begins interaction with customers a few months after purchase of solution. Here, the point of contact isn't senior leadership but the IT manager who knows the product best and oversees day-to-day operations and implementation of the tool.

With a continued focus on retention and expansion, the CSM team conducts health checks on customers. They collect information on the customer's environment, training, consultation, and implementation processes. This data is crucial to identify gaps in product usage, customer education, user engagement, and product adoption. After every consultation, the customer success team shares a report with the customer to study their current practices and recommend changes in line with their business requirements. They also conduct specialized training sessions and workshops to encourage optimized product usage. Additionally, they interact with departing customers to identify potential problem areas.

(ii) Customer Advisory Board (CAB)

ManageEngine's CAB is a platform that facilitates open dialogs between key customers and senior leaders in the organization. The company invites select customers of different sizes from various regions and industries to connect with their peers. These customer-driven conversations focus on industry trends, common IT challenges, and what they expect from ManageEngine which help us design the roadmap for our solutions.

AI in customer support

Predictive and generative AI tools hold the potential to elevate an organization's customer support game. We've talked about 24/7 chatbots, but here are a few more ways to implement AI and provide customers with an all-in-one personal butler:

Wait time monitoring

Customers aren't patient. We know this because we aren't patient either when we're the target of the seemingly endless elevator music. Giving customers a definite response is a great way to retain their attention and resolve their issue. AI systems can monitor queues and give customers an estimated wait time or their position in queue. This is calculated using a combination of real-time data analysis, historical call data, and predictive algorithms.

Intelligent routing

AI can read incoming tickets and queries and identify the most suitable support agent. Let's say we have an online store that sells tech gadgets. A customer is facing issues with the printer they purchased and contacts support. Intelligent routing systems, rather than relying on round-robin scheduling, can direct the ticket based on the technician's success rate related to printer issues and their availability.

Call transcriptions

AI-powered transcription tools can document customer calls in real time or after the call is recorded. This usually serves two purposes: first, it facilitates faster response and follow-up actions from the support agents. Second, having a written record of the conversation often helps with dispute resolution. Transcripts can be used to train new hires and speed up the onboarding process for the support team.

Ticket responses

Support agents deal with a flurry of tickets on a daily basis, which can be hard to track without the right tools. Incorporating AI, specifically natural language processing (NLP) tools, into the help desk system can assist agents by providing them with a summary of the ticket, an analysis of the sentiment of the customer, and generate a suitable human-like responses.

Multilingual support

We've established the importance of having multilingual support for a global user base. If it isn't feasible to set up exclusive customer support for smaller regions, organizations can benefit by using AI-enabled translators to respond to customer queries. Of course, AI cannot replace humans. But NLP can translate the content so the AI-enabled translator can detect cultural subtleties or offer appropriate responses the way humans do. AI cannot replace the real deal.

Setbacks and strategies

a. Identify emerging digital hubs

The "Ok" sign made with the thumb and index finger signifies approval, at least in Western countries. In some non-English speaking European countries, the same symbol is considered rude and disrespectful. A support technician could make an innocent mistake via chat if they aren't aware of this. However, this is not common knowledge, which is exactly why it's crucial to invest in a local support team.

Understanding cultural differences and social subtleties can be a challenge with customer support, especially if you're operating from one location. However, it can be expensive to set up teams in popular cities to cater to your target market. Setting up smaller teams and shifting focus to localized recruiting at emerging digital hubs is a great way to respect these nuances and address customers' needs simultaneously. Zoho Corp. has hub and spoke offices worldwide, most of which are situated in rural or suburban regions, with staff native to the region and fluent in the local language.

b. Automate wherever possible

Let's not forget, support agents are humans too. Even the most diligent agents can make errors, especially when handling monotonous tasks or when operating under pressure. Moreover, the sheer volume of incoming inquiries can be overwhelming for support teams in growing organizations. In order to scale operations without staggering costs, automation should be an integral part of the customer support domain.

Workflow automation can free up agents from redundant tasks like sending welcome mails for new tickets, responding to FAQs, and generating reports. It can also be combined with AI for automated ticket routing based on a predefined criteria. ManageEngine's help desk allows workflow automation to auto-create/auto-close tickets, assign tags, initiate approvals for escalations, and generate responses.

c. Support your support team

Working in customer support can be demanding, frustrating, and sometimes a little monotonous. One negative customer interaction can affect a technician's productivity for the rest of the day. In the long run, low morale can take a toll on the team and, in turn, on customer satisfaction. It's important to take employee experience into account and make sure support agents feel heard and valued.

Support managers need to be more hands-on and try to understand the operational challenges faced by the team. Remember that support goes beyond numbers; team members are humans too. It's important for an organization's management to take the time to hear support employees and encourage open discussions. Implementing a reward system for providing exceptional customer service is another great way to incentivize their work and show the support team that their contribution makes a difference to the company.

d. Under-promise and over-deliver

Getting an Amazon package a day before the estimated delivery date is probably one of the top ten joys of life. Amazon announces to customers conservative delivery dates that take into account multiple variables like supply chain issues, logistic challenges, or unexpected increases in demand. Customers might not recall when a business meets expectations, but they will remember when a business fails to do so. Providing conservative customer service estimations gives the organization the buffer time it might require to handle unexpected hiccups without disappointing customers. A similar ideology is important in customer support.

If a business promises response to tickets within three hours, customers expect response within three hours. However, the support team might actually target a response time of within an hour. Additionally, throwing in complimentary goods or services is a great way to keep customers happy. For instance, tickets to an upcoming tech conference in their city could be a pleasant surprise for the customer.

Wrapping up

Customer support is crucial to succeed and stay in business. Being the bridge between customers and the product teams is no small feat and shouldn't be dismissed as "just responding to queries". Conversations with senior leadership at Zoho and ManageEngine highlight a common theme in support—our actions must mirror our words; and our goal is to provide a world-class support experience. The framework we've provided today helped us achieve exactly that, and we hope it works for other organizations as well.