Building and Scaling a Customer Support Function
What are the business advantages of a strong support function?
Continuous improvement of the support function reduces cost – high-quality customer experience teams continually iterate on the defects reported by customers. By proactively addressing issues and fixing bugs, fewer problems occur in the future and companies can spend less on customer service.
Retention improves when customer issues are effectively repaired – customers generally contact customer support when they are having a bad experience, and it’s up to the support team to get those customers back on the “happy” customer journey. Strong support teams repair negative experiences so effectively that retention for complaining customers becomes comparable to baseline retention.
Excellent customer service is an invaluable branding tool – customer service shows how much a brand cares about its customers. Above-and-beyond customer service drives loyalty and brand love by forging an emotional connection and demonstrating that the company values the relationship.
What are the common growth/scale inflection points at which a company needs to evaluate its customer support capability?
Customer support strategy must be reevaluated consistently – while certain business changes necessitate more significant adjustments to customer support (CS) strategy, consistently prioritizing and investing in CS reduces the risk of negatively impacting the customer experience. The gradual increase of CS issues can be compared to “boiling a frog”; changes happen slowly enough that many companies don’t realize they’ve reached a critical point until they’re already experiencing significant challenges.
In a growing business, support becomes a significant cost center if left unoptimized – as your topline revenue grows, CS needs grow linearly unless improvements are made to your processes or products. For example, if your business grows by 10X, your support needs will also increase by 10X unless you reduce product defects. Eventually, the cost of CS becomes significant enough that companies see the value in taking a more strategic approach to optimizing the function.
CS improvements become necessary as business complexity increases – defect rates tend to increase exponentially as you add product lines, service more geographies, support multiple languages and develop adjacencies. If the rate of customer service contacts rises faster than the business is growing, you need to invest in a more efficient support team and prioritize eliminating the root causes of these defects.
The Customer Support Team
What are the common roles or sub-functions within customer support?
Team structure becomes more developed as companies grow – smaller businesses have CS functions driven by employees who are able to wear multiple hats. As your business gets bigger, roles become more defined and responsibilities become more focused through specialized sub-teams and individual issue specialists.
| Sub-function | What it is | |
| Workforce Management | This team is responsible for determining people and staffing needs. The Workforce team uses real-time analytics to optimize staffing and ticket handling. | |
| Learning & Development (L&D) | This team onboards and trains new representatives. They also create content for training, workflows, and upskilling. | |
| Quality Assurance (QA) | This team regularly audits individual agent performance, providing scorecards and feedback that help agents improve on an ongoing basis. | |
| Vendor Management | If outsourcing is important to your business, this team is responsible for selecting, negotiating with, and managing your providers. | |
| Customer Experience (CX) Generalists | CX generalists focus on understanding the customer journey and using insights from support interactions to drive improvements across the business. | |
What customer support KPIs should companies monitor? What does good look like for each?
Balance CS scorecards by prioritizing both efficiency and quality metrics – well-run Support teams resolve issues efficiently while providing a high-quality experience for their customers. Although quality metrics don’t translate directly to immediate revenue impact, they are an important indicator of brand health and can affect revenue by improving longer-term customer satisfaction.
| Efficiency KPIs Efficiency KPIs measure how well CS performs from a cost and employee utilization perspective. | ||
| Metric | Definition | Benchmarks |
| Defect Rate | Percentage of transactions resulting in customer service contact | Good: 1-5% Problematic: 10% or higher |
| Average Handle Time | Average time it takes to resolve a ticket | Varies by industry and complexity (must balance quality + expediency) |
| Cost per Minute | Average cost per minute of agent time | Varies |
| Overall Productivity | Combination of below metrics | 60-70% |
| Shrinkage | Percentage of shift time agents show up to work | 90% or higher |
| Utilization | Percentage of shift time agents are available to take tickets (vs doing something else) | 85% or higher |
| Occupancy | Percentage of available time agents are working on tickets | 85% or higher |
| Quality KPIs Quality metrics determine how well CS provides helpful solutions that meet the customers’ needs and improve the overall customer experience. | ||
| Metric | Definition | Benchmarks |
| Customer Satisfaction (CSAT) | How happy/ satisfied customers are with support interactions | 65% or higher |
| Net Promoter Score (NPS) | How loyal and likely to recommend your customers are | 50 or above |
| Customer Retention | Rate at which customers stay with a business in a given period of time | Varies by industry |
How does the customer support org structure change based on your business model or customer base?
| Business Model | Support Approach | Org Structure Differentiators | |
| B2B | Higher touch, more personalized approach for enterprise clients – since churn is expensive and customer issues are more complex, B2B support functions often assign representatives to a specific account. Support acts as a complimentary team with the Product and Sales functions to help resolve high-stakes problems and create a positive customer experience that acts as the foundation of a long-term client-provider relationship. | • Account-based support • Tiered support levels • Consultative, cross-functional approach • Customized service levels based on contract value or product | |
| B2C | Efficient and scalable solutions can address higher volumes of simpler customer problems – customer issues are often simpler to resolve and have a smaller impact on future revenue. Workforce management and streamlining the customer experience are critical for B2C businesses that plan to scale. | • Efficiency team focusing on day-to-day operations and cost management • Experience team focusing on retention and product quality | |
| Marketplace | Components of both B2B and B2C Support organizations are necessary – because marketplaces have two sets of clients, they often have two entirely separate Support organizations that need to be managed according to different principles. | • Separate team for each side of the marketplace | |
| SMB | SMBs have significant variations in CS org structure – needs vary greatly depending on the nature of the specific business, the nature of its clients, and how sophisticated the SMB itself can be with regards to building out its Support function. | • Highly variable • Teams face the greatest difficulty as they try to figure out how to best serve their customers | |
How should headcount change as your customer base grows?
Headcount depends on your industry and the nature of the support your team provides – you always want to have as few Support representatives as possible, but there is no golden ratio for this function. For example, if you offer products that frequently have urgent issues (e.g. home security), it might be critical to your business to have a large team of highly paid representatives who take their time answering customer calls. On the other hand, some businesses have large customer bases that rarely encounter urgent issues, which allows those businesses to field smaller, less expensive teams.
| General Headcount Guidelines | |
| Role | Ratio |
| Workforce | 1 workforce manager: 200 representatives |
| RTA | 1 RTA: 60 representatives |
| QA | 1 QA: 30 representatives |
| Team leads/ managing agents | 1 lead: 10 representatives |
| L&D | 1 head of L&D: 1,000 representatives 1 L&D trainer: 60 representatives |
Tooling Customer Support
What are the different categories of tech tools for a customer support function?
| Tool Category | Key Functionality | ||
| CRM (e.g. Salesforce, HubSpot) | Foundation for customer data management–every company needs a CRM at some point. | ||
| Contact Center Software (e.g. Gladly, Zendesk, Intercom) | Manages multichannel communications and handles ticket creation, routing, and resolution. Might include telephony and knowledge-base features | ||
| Workforce Management Software (e.g. BambooHR, Workday) | Manages schedules, forecasts staffing needs, and tracks productivity. Helps maintain service levels, optimize resource allocation, and ensure that the team meets all legal requirements | ||
| L&D Software | Creates training materials and tests to educate and upskill representatives. Manage knowledge base, educational content, and employee training. | ||
| Telephony Systems (e.g. Five9, Dialpad, RingCentral) | Manages phone communications for your support reps. | ||
| AI Chatbots | Automates simple inquiries and assists agents. These platforms can be standalone or integrated into contact center software | ||
What AI use cases should you evaluate for your customer support organization?
Start with the most basic AI use cases and attempt to mature from there:
- Basic use case: Automating discovery – AI reviews a chat or email and summarizes it for an agent to expedite their reply time.
- Moderate use case: Live agent assist – AI identifies and surfaces relevant information (articles, manuals, etc.) to help an agent provide better live support.
- Complex use case: Support resolution automation – AI solves the entire customer issue without involving a human support agent. This solution can seem appealing but is often not as strategic as people think due to its reactive nature and current limits of the technology.
Customer Support Processes
How should you decide which mediums and channels to offer customer support through?
Understand what your customers want and try to meet them there – customers usually have very clear preferences around communication. For example, a B2B enterprise client might need to have a long phone conversation with someone about a complex issue, while a B2C consumer might prefer to issue a complaint about a delivery service through a text message or button in an app. The entire CS process feels more seamless to the customer if you’ve factored in their habits and preferences instead of forcing them to use media that doesn’t feel like a natural solution.
Urgent issues require real-time channels like phone calls or live chats – that will give them access to real people—and from a Trust and Safety standpoint, it is never in your best interest to force a potentially critical issue through a logic-based chat or auto-generated email flow. Less urgent issues can flow through slower channels like email.
Optimize the cost and quality trade-off – balance cost and quality for each type of customer support request in order to find a solution that works for your business.
| Support Channels | Cost | Speed | Quality |
| Phone | High | Fast | Highest |
| Chat | Low | Fast | Low |
| Low (easiest to manage from a workforce perspective) | Slow | Medium |
Start with fewer channels (it’s easier to add a new channel than to remove one) – removing options inevitably creates confusion and frustration for customers acclimated to your old support options. Be thoughtful about which channels you start with, and only expand if you identify a new customer need and have the capabilities to meet it effectively through that channel.
When do you need a call center?
Use a call center when your internal costs become a problem – don’t try to “proactively” save money by outsourcing your Support function to a call center. Instead, grow your internal capabilities until it’s no longer in the company’s best interest to do everything in-house. By the time you outsource, you should have optimized key Support processes, codified common workflows, and identified specific types of tickets to outsource.
In-house call centers are sources of innovation and optimization – the highest-performing Support functions create new ways of solving customer issues and transform problems and costly defects into opportunities for improvement that move the entire business forward. Once you begin outsourcing CS to a call center, you risk losing access to these insights.
Outsourced call centers follow directions and achieve consistent performance – external call centers are reliable because they depend on compliance. If you provide an external call center with a comprehensive playbook on how to handle certain types of customer tickets, they will do as they are told at any hour of the day and for a fraction of the cost.
How do you implement an effective escalation process?
Effective escalation processes require excellent filters – in order to escalate issues in a time- and cost-effective manner, you must establish a system or logic that can identify all issues that are urgent, sensitive, complex, and high-value. False negatives are potentially damaging to your business and false positives create unnecessary costs, so filters must be well-designed and well-implemented.
Create a culture that encourages appropriate escalation – agents should never be punished for accidentally escalating an issue. Instead, focus on training agents how to determine when to escalate, and demonstrate how escalation can be good for the business.
| Escalation Type | Process | |
| Trust and Safety | 1. Implement a clear “red line” criteria that triggers immediate escalation. 2. If possible, build an AI solution where Trust and Safety issues get escalated before they ever reach a regular representative. 3. Train agents to recognize and escalate sensitive issues without hesitation. 4. Create a dedicated queue for Trust and Safety concerns. | |
| Customer Satisfaction | 1. Establish clear guidelines for when to escalate dissatisfied customers. 2. Create an escalation queue managed by team leads or senior agents. | |
| High-Value or Complex Issue | 1. Use dynamic routing to direct complex issues to specialized teams from the start. 2. Implement a tiered support structure that matches the type of agent to the type of customer. 3. Create escalation paths for VIP customers or high-stakes situations. | |
How do you implement self-service options effectively?
Focus on self-service on high-volume issues – identify the 20% of issues that drive 80% of support volume and create clear, easy-to-use self-service flows for these common problems. You should regularly examine trends in support issues to ensure that your self-service options are prepared to assist with any new high-volume issues that might emerge.
Keep self-service as simple as possible – where you can implement button-based flows instead of chatbots for basic issues. When customers aren’t able to access a live agent, they’re less likely to escalate if the flow is straightforward.
Provide an “escape hatch” for users to discuss with a person – always offer an option to connect with a live agent (via phone or chat) if self-service doesn’t resolve the issue. Failing to do so will create an influx of tickets that don’t solve your customer’s problem.
How do you effectively train new customer support representatives?
Representatives require extensive support throughout the full span of their employment – gradually ease new representatives into their roles by providing significant support during the 4-6 week onboarding process. However, even agents who’ve reached full productivity benefit from upskilling, emotional support, and an “always improving” team culture.
Prioritize training quality over expediency – it is tempting to rush the onboarding process, especially since Support can have a high turnover rate. Rushed onboarding negatively impacts retention, and well-trained agents are happier and perform better than those who lack sufficient support.
| Customer Support Training Process | |
| Weeks 1-2: Classroom-style training | Bring agents up to speed on your products, systems, processes, and customers. |
| Weeks 3-4: “Nesting”-style training | Gradually introduce hands-on experience by starting new representatives with a small volume of tickets (5% of typical volume). Agents should side-chair with one another and debrief with their managers after each ticket. |
| Weeks 5-6: Ease into higher call volume | Side-chairing might continue into this stage as you slowly increase ticket volume from 50% at the beginning of Week 5 to 100% by the end of Week 6. |
| Weeks 7+: Ongoing support | Assign mentors or buddies for new representatives and create a culture that acknowledges the stressful nature of CS work. |
Overall
What are important things to get right?
Customer support should initiate constant feedback loops across the organization to cut down on tickets – the information CS has about what is and isn’t working is invaluable to Product, Operations, Sales, and leadership. Support must be in consistent, meaningful communication with each of these teams.
Customer support costs should not scale linearly with your customer base – as your organization grows, ensure that your Support function is capable of handling increased volume with high levels of efficiency.
What are common pitfalls?
Replacing one tool with another usually creates as many issues as it solves – Support tooling options are famously imperfect, and there will always be problems with whichever system you use unless you are in the 10% of companies that are able to custom-build a high-functioning solution. When new Support leaders come in, they often put the department through the disruption of changing contact center solutions even though the new solution isn’t objectively better than the old one. Be thoughtful about your tooling choices and focus on solving the most important business problems instead of implementing the perfect tool that does one thing well at the cost of creating new problems for other parts of the business.
Overestimating the value of external call centers prevents CS from living up to its full potential – at the highest level, customer support perceives opportunities based on consumer insights and presents those opportunities to the rest of the business to help it become better. External call centers can follow directions and provide reliable support when internal capabilities are overwhelmed, but they can’t replace the real value of the Support function.
The appeal of pure automation AI solutions is a distraction from solving real problems – developing “pure deflection” solutions is trendy, but its main benefit is saving (expensive) agent time. A CS function that is focused on solving the underlying problems that create the issues that “pure deflection” would be handling will have a significantly larger impact on the company as a whole.
Responses