Balancing support efficiency with a great customer experience

Main illustration: Susan Payne

As customers expect more and more out of support experiences, support leaders can risk burnout on their team to meet the escalating demand.

In a recent study with over 600 support leaders, we found that smaller support teams (fewer than 100 agents) are less likely than larger teams (more than 100 agents) to say they can continue meeting accelerating customer expectations. But even for larger teams, an influx of customer questions can overload agents, leaving them frustrated and overworked, and in turn, not able to provide great support.

This poses a real challenge – how can you increase the efficiency of your support team, while still maintaining the high quality experience your customers want and deserve? With some experimentation, automation, and strong relationships across teams, it’s possible to strike the right balance between providing a positive environment for your support team and keeping your customers happy.

The growing tension between team burnout and customer demand

Today, most support teams are hyper-focused on reducing their first response time (FRT). The emergence of real-time chat has made responding to customers’ conversations in real time easier and in turn, has changed customer expectations forever. Today, customers expect fast, efficient, and personal help.

Although businesses are seeing the benefits of real-time chat – improved customer satisfaction, retention, etc. – not all support teams, especially small ones, can support this type of human support at scale. Even teams that could offer real-time chat may not feel that it’s the right solution for them. Responding to all customers in “real time” adds operational complexity and places potentially unnecessary pressure on your support team. Maintaining a 100% real-time experience typically requires:

  • Increasing headcount to keep pace with a growing customer base and their conversations.
  • And if you don’t increase headcount, you risk burning your team out. That’s less than ideal, especially if you see spikes in customer demand at different points throughout the year.

Instead, with a conversational support approach, you can ensure most simple customer questions are automatically addressed (and often resolved) through self-serve and proactive support. This leaves your team with more time to tackle complex and VIP queries.

“73% of support leaders say customer expectations are increasing, but only 42% are sure they’re meeting those expectations”

Define what a great customer experience looks like

Our research shows that 73% of support leaders say customer expectations are increasing, but only 42% are sure they’re meeting those expectations. That means that a crucial first step in balancing customer experience and efficiency is to define what great customer support looks like for your business – and it will be unique to your business.

“A good customer experience is rooted in setting clear customer expectations”

Still, no matter what your business looks like, unmet and unclear expectations are the root of all heartache. A good customer experience is rooted in setting clear customer expectations. For example, when a customer writes in, you can set up an automated bot to let them know how long they’re going to wait, and if that wait is on the longer side, then make sure to meet that expectation.

At Intercom, after we automatically connect a customer with the relevant support rep on our team, the interaction moves along pretty quickly. It can be close to real time in that capacity. Ultimately, providing a resolution in under four days is our goal. Usually, that customer will provide positive feedback if we meet that expectation.

That’s our ideal customer experience, but it’s not right for everyone. I’ve seen many companies aim for a really fast first response time, triage that conversation and get it to the right team, and then from there offer one response a day, as opposed to that near real-time interaction.

Both can work! It really depends on two things:

  • Your customer base: What kind of conversations are they actually having, and how urgent are their questions?
  • What kind of experience you actually want to provide to your customers: At the end of the day, there’s no right answer here. Both of the workflows I outlined above have their upsides. They both have their downsides. It’s up to you as a support organization to specifically define this and let your customers know what they can expect.

Take a fresh approach to satisfy your customers – and your team

At Intercom, we take more of a conversational approach to support – we use automation to answer simple, repetitive questions, and leave an escape hatch open should the customer need to speak to a human. Then, the conversation can bounce back and forth between the customer and agent until they reach a resolution.

About a year ago, our Customer Support team had about 100 people. But with all the automation technologies that we’ve rolled out, along with other efficiency initiatives, we have decoupled the growth of our Customer Support team with the growth of our customer count, even putting ourselves in the position to not backfill support folks who move on to another company or another role within Intercom.

Below, I’ll walk you through how we got there, and provide some tips for how you can gain back efficiency on your support team without sacrificing the happiness of your customers.

Experiment with first response time to see where you can gain efficiencies

Your first response time (FRT) is one of the first places you should look to if you need to create a more streamlined workflow for your support team. Increasing your FRT can buy your team back significant amounts of time, allowing you to provide a consistently great customer experience without having to scale your actual team.

We recently ran a test on FRT to see if we could increase first response time without sacrificing customer satisfaction and retention.

The test compared a longer FRT of one business day with our normal FRT of several hours, splitting a pool of lower spend customers 50/50 into these two experiences. If the longer FRT didn’t make a significant difference to customer satisfaction or decrease retention in excess of the financial savings from decreasing our headcount growth, then the test would be a success.

“We also saw a 4% decrease in questions asked and a 20% gain in efficiency per conversation”

After running the test for five months, we found, at a 95% confidence level, that there was only a 1.25% dip in our customer satisfaction rating (CSAT), and most importantly, no impact to net revenue retention. We also saw a 4% decrease in questions asked and a 20% gain in efficiency per conversation.

Indicator  Relative change
Questions asked -4%
CSAT -1.25%
Net revenue retention 0%
Efficiency gained per conversation +20%

As you start to test your own FRT, keep in mind the context of who a customer is, what they’re asking about, the importance of the question, and how big that customer is.

In addition to running a test, consider increasing the FRT for bigger customers who you often give a very fast FRT to. For example, if they’re writing into you about something that is not at all urgent, like a feature request or a suggested UI improvement, this won’t require the white glove experience you’d usually provide. Despite the fact that they are a high value customer, if there’s nothing urgent about the actual conversation, then it doesn’t need an immediate response.

The downside of that sort of action is that not all of your customers might agree with your determination. So tread carefully here, and weigh the value of the efficiency gain with the actual revenue value of the customer. And if you’re going to move those conversations to a longer FRT, make sure it’s a substantial difference so you gain a substantial amount of efficiency – real time to a couple of hours, for example.

Leverage automation to empower your team to take on more meaningful work

Automation is one of the most powerful ways to relieve your support team from repetitive questions that are often the culprits of team burnout. If implemented correctly, we’ve seen automation immediately resolve 33% of customer issues.

In a recent survey we ran with support leaders, we found that 73% of companies that automate support with chatbots have seen an increase in conversation/ticket resolution over the past 12 months.

“The key to mastering automation is to keep it personal”

The key to mastering automation is to keep it personal. And you can do that by using a sophisticated business messenger to provide support that’s dynamic, and that feels like you’re having a conversation with an agent – even if it’s not in real time.

If you’re concerned that introducing automation into your support strategy will feel impersonal to your customers, remember that the tools that you have available to you now aren’t at all like your old impersonal, legacy support tools. Here are a few ways you can put automation to work for you, while still keeping it personal:

1. Automate the transactional

Setting up automation to handle the transactional side of things really respects your end user’s time. For example, you can set up a bot, like Resolution Bot, to answer simple questions like, “how do I change my profile picture,” or “how do I update my billing details,” or even “how can I change my password” that really don’t need to involve a human.

But you can also set up a Custom Bot to set customer expectations for how long it’ll take to get a response, or to collect information that will speed things up, like the customer’s email address or order number. And then, at the end of the day, when the conversation is closed, you can use automation to send a survey to see how your team did.

2. Personalize the experience with context

You can also use bots to automatically capture custom attributes that will help it route the customer to the appropriate team for a response. Plus, those attributes can provide important context about the customer, like if they’re a high value customer – in which case, you’ll want to provide them with a great, VIP experience every time.

“Sending a customer through a logic tree to figure out what they’re asking about can save them time”

The other key here is to make sure you ask relevant questions. Sending a customer through a logic tree to figure out what they’re asking about can save them time. For example, say you have products A, B, and C, but the incoming customer only uses product C. You can use logic to figure out what product they use to make sure you aren’t asking them questions about a product they aren’t even using.

3. Integrate with helpful apps

Bake apps into your automation to streamline the experience. For example, use app integrations to schedule meetings, check order statuses, or send a typeform for a feature request.

4. Measure the impact

Finally, you’ll want to measure how your team is doing and impacting your business’ bottom line. By using automation, you’ll already have a head start – companies that use chatbot technology for support are 35% more likely to be able to track their support team ROI.

“Last year, our ROAR was around 4%, which translated into a savings of $400,000”

We recommend measuring:

  • Customer happiness through a combination of CSAT and an Internal Quality score.
  • Rate of automated resolution (ROAR), which is a metric that I’ve designed internally. ROAR is the main metric that we look at to see what the efficiency of our automation is in terms of answering questions without a human actually being involved. That efficiency translates into real revenue for us. Typically, our ROAR hovers between 4 and 5%. Every increase of one percentage point that we have in ROAR means that we don’t have to hire an additional head. Last year, our ROAR was around 4%, which translated into a savings of $400,000.

Create efficiency gains with a strong relationship with other teams

Though automation is extremely useful, it isn’t the only way to gain efficiencies. One of the biggest opportunities lies in fostering a tight relationship with your product team and sharing customer insights with them so they can proactively solve common problems in the product.

“By proactively addressing problems upfront, you can prevent these questions from arising in the first place”

For example, you can identify the areas that are confusing customers, the features they really need, and their pain points. Then, you can package that feedback up and bring it to the product team. This is one of the biggest efficiency wins – by proactively addressing these problems upfront, you can prevent these questions from arising in the first place. We prioritize and get a lot of value from this type of proactive support at Intercom.

Use the Conversational Support Funnel to create efficient, positive experiences at scale

As you start to think about balancing a great customer experience with maximum team efficiency, you’re really setting out to optimize the three fundamental components of the Conversational Support Funnel: proactive, self-serve, and human support.

By fostering an open feedback loop with your product team, you’ll activate the proactive leg of your support strategy. And mastering automation means bringing a really fantastic self-serve support experience to your customers, so that they can get their questions resolved almost instantly or get routed to the right human in record time.

Ready to make your support as efficient and personal as possible? Get the Conversational Support Funnel guide👇