Helpshift enables brands to deliver superior digital customer service by combining asynchronous messaging, intelligent automation, and data-rich conversations.
Digital customer support is new and exciting – bringing a world of possibilities for boosting customer loyalty and lifetime value (LTV). But how do you measure results?
Key performance indicators (KPIs) can help prove your success in digital customer support and can also help you continuously gauge customer service performance and gain executive buy-in for new and expanded initiatives.
Here are a few KPIs you should be using to measure the success of digital customer support.
For decades, customer satisfaction (CSAT), time to first response (TTFR) and time to resolution (TTR) have served as the top indicators of a customer service organization’s performance. That has not changed. Yet, how we consider those measures changes dramatically when supporting customers over digital channels because the technology and features embedded in digital support are so fundamentally different from traditional support infrastructure. Let me explain:
When customer service becomes digital, an organization can manage most of its customer traffic using message-based interactions. We’re not talking about live chat. Instead, the most effective messaging in digital customer service is asynchronous messaging — the way people message one another in apps like WeChat, iMessage, and WhatsApp. With asynchronous messaging, threads can be both real-time and time-lapsed. As a result, a customer can start a conversation, leave it for a while and return later to complete it. This gives customers greater flexibility and convenience.
Automation and AI can deliver additional customer convenience when embedded within messaging. An AI-driven answer bot that uses natural language processing (NLP) and machine learning (ML) can immediately surface useful knowledgebase articles based on a customer’s query, enabling them to easily self-serve. NLP bots can also detect the intent of each customer message to streamline the classification and routing of support issues. Finally, chatbots can also automate routine workflows, such as returns and exchanges or account updates, further allowing customers to gain assistance without needing help from a human.
Combining the convenience and speed of bots with the flexibility and convenience of asynchronous messaging drives up CSAT to new highs. CSAT under asynchronous messaging blows away CSAT in other digital channels. (Messaging outperforms other channels in terms of CSAT, according to Helpshift’s Performance Index Benchmark Report, 2019.)
One of the biggest reasons for this is because customers no longer have to restate their issue over and over again when they engage with support in messaging. Imagine the typical banking customer trying to apply for a loan. They might engage with a number of people across channels — email, phone, and perhaps live chat — and will likely have to repeat all of their details every time they begin the conversation again on a new channel. Each time a customer has to repeat themselves, think of it as another point deducted from their support rating. Asynchronous messaging does away with this problem by keeping the conversation in one dedicated channel, causing a substantial boost to CSAT.
In addition, when brands introduce smart automation and AI into messaging to further streamline and improve the customer experience, it correlates with a further boost to CSAT. Thus, comparing the CSAT results from digital customer support against your previous legacy platform is a simple way of proving its value.
Able to provide instantaneous responses, automated, intelligent bots drive down ‘time to first response’ dramatically – a median TTFR of just 30 seconds among the top-performing customer service organizations. That compares quite favorably to top-performers using email, with their median TTFR of 2.6 hours. Believe it or not, that’s a gain in response time of an astounding 31,200 percent! Think of how important that improvement in TTFR is when you consider that more than half of all consumers now expect a customer service response within one hour.
Beyond responding to a query nearly instantly, automation and AI can also enable large scale ticket deflection — where a support issue is never created because the customer has their question answered via a chatbot response. This creates a new KPI to measure alongside TTFR — deflection rate. This KPI is measured as the percentage of users who start a messaging conversation but don’t end up speaking to an agent. So if only 10% of messaging interactions result in a conversation with an agent then your deflection rate is 90%. That means agents receive far fewer tickets, and the ones they do receive are for unique or time-sensitive issues that require human skills to resolve.
Do you solve problems in an acceptable time frame? That’s the biggest question in customer support. The advances of asynchronous messaging and bot-driven customer self-service impact TTR in two ways. First, bots shrink TTR, because customers receive links to the content they need immediately.
Second, on the flip side, the time-lapsed convenience of asynchronous messaging can make your TTR appear artificially high. This is due to your customers enjoying the convenience of not having to stay in a live interaction. They can raise an issue, go about their daily routine and return later to complete the process. For example, rather than sitting through an hour-long live session, a customer uses asynchronous messaging and spends just a few minutes of their day to resolve their issue. Yet, the time-lapse involved in this messaging discussion makes TTR look artificially high. The elapsed time between start and finish could look like eight hours when in reality the customer only dedicated a few minutes. That is why we suggest using an additional KPI in conjunction with TTR — engaged Time. This is the amount of time the user spent engaging with the messaging thread.
TTR also improves when customer interactions become semi-automated. This means that agents and bots work together to resolve issues faster. For example, in retail, when a customer issues a request to return a product, a bot can collect information about the customer and their order upfront. The bot then passes that information along to a live agent to resolve the issue. Brands that follow a digital-first strategy, unifying digital with live agent support in semi-automated workflows, lower their TTR significantly. Response times average just 18.5 minutes for agent-assisted digital channels like messaging, compared to 70 minutes for social media and 411 minutes for email.
Today, customers are happy to let brands direct them to channels that provide the most efficient support. The combination of asynchronous messaging and AI-driven bots enables automated workflows that carry context from channel to channel. This means customers never waste time repeating themselves and obtain service in a way that works for their schedule.