Bots for Business
Listening to Your Bot: Chatbot Metrics and Analytics
Chatbots are powerful tools for helping your businesses engage with customers. You can use chatbot analytics to measure your bot's success and to improve your customers’ experience. So which are the best metrics and KPIs for measuring their success, and what can you learn from your chatbot?
November 6, 2018
Chatbots have become incredibly powerful tools for helping businesses engage with their customers. This 24/7 virtual staff allows companies to communicate with their clients, answer questions, and drive sales—all of which may help your company grow. But the real benefit of chatbots comes from their ability to collect data.
Information gleaned from your customers’ chatbot interactions can be useful. The data often shows how well your bot is performing and gives you a deeper understanding of the overall success of your marketing strategy. Chatbots can also solicit information from your consumers in passive ways to help you better understand your clients.
4 Chatbot metrics worth tracking
When a customer begins interacting with your chatbot, the data from the exchange is captured. This information is then sent to the backend of your bot’s host for you to analyze and to help you better understand how well your bot is operating. Many platforms offer chatbot analytics tools, which allow you to see key data points, such as:
1. Response rate
Chatbots may initiate conversations with visitors wherever they are on your site. The first step to having a successful bot/consumer interaction is to get the attention of your potential customer. A simple “Hello”—or a more direct “How may I help you today?”—might work.
The platform hosting your bot reports the number of times that visitors respond to your bot’s initial attempt at contact. This information may then be utilized to adjust your chatbot’s opening statement to maximize its response percentages. A bot can’t help you if users won’t interact with it.
2. Session duration
As soon as your customer begins interacting with it, your chatbot can start a clock to track how long the engagement lasts. If you’re trying to drive traffic to a site or product, keep your bot’s script short to reduce the interaction time with the user.
If your bot is programmed to answer questions or to collect information, your aim may be a longer engagement time. There isn’t a golden rule for how long your customers should be interacting with your bots, but the ability to see engagement durations allows you to notice any obvious problems and make adjustments to increase your bot’s effectiveness.
3. Goal completion
Conversion rates are the true measure of your bot’s effectiveness. After your bot has captured the attention of your customer, the bot script you wrote should guide the user toward a set goal. You may want your bot to lead your customer to a certain product or to convert traffic from one page to another.
You can gauge the effectiveness of your chatbot by tracking how many of your customers interact with it but do not pursue these goals. For instance, if your customer stops their interaction with the bot at the point when the conversion point should occur (when a goal is triggered), you need to make adjustments to increase your conversion percentage. Bots must be fined-tuned to maximize the amount of work they can do for you. Conversion rates can help you to better understand when to update or rewrite your script so your bot can do the job it was created for.
4. Customer retention
Watching to see how many people delete your chatbot conversations is a great way to gauge whether the interactions were useful. A deleted chatbot interaction can show that your bot is not doing its job and may even be annoying potential customers. A worthwhile exchange between a user and chatbot might be saved by a consumer, especially if the user viewed the interaction as beneficial.
It’s also important to limit your deletion numbers so your chatbot can follow up later with potential customers. If a consumer does not delete the conversation, your chatbot can be programmed to reach out to the client and try to reconnect. Since scripting is time-consuming, it is common to have a bot send a new greeting message designed to pull the costumer back into the script. More elaborate scripts can be written to include specific follow-up questions that address the reasons why a user ended the interaction.
3 ways to leverage customer insights through chatbots
Running a successful business hinges on your ability to know what your consumers want and to provide it. After you understand this, you can shape your products and services to meet these expectations.
1. Build out your knowledge base
It’s hard to predict every question a potential customer might have. When a person is interacting with a chatbot, they may ask it something outside of the provided script. But a bot can be programmed to respond with a default message to use when it doesn’t understand how to answer a user’s question. Usually, the default message includes information on how to contact a human customer service representative.
After this default message is sent, you can look at the interaction to determine what the question was. If it’s significant, you can rewrite your bot’s script to include the answer. Now your bot can answer even more of your potential customers’ pressing questions 24/7.
Many businesses have relied on FAQ pages to help alleviate their clients’ concerns. But FAQ pages are static and often can’t collect information about questions that a consumer might have, other than the answers already on the website. This can cause frustration for consumers and might result in lost sales for the business. The worst part is that, unless the potential customer sends you an email or calls your customer service department, you won’t be able to answer their questions. Chatbots allow you to identify and address the questions your potential customers have, without requiring an additional email or phone call.
2. Generate new leads
Sales representatives spend a lot of time finding new customers. Rather than spending staff time on cold calls, you can have a chatbot collect information from your potential customers. It can interact with them, learn what they are looking for, and determine whether they are ready to make a purchase.
When a potential customer reaches the end of their questions or meets the limitations of the script, your bot can prompt them to enter their contact information, which can then be accessed by your sales staff. Bots can even categorize the lead information, based on which product or service your potential customer is interested in. The ability to passively and remotely generate leads saves your staff time and your business money.
3. Improve customer satisfaction
Not many people are fond of taking surveys, which might explain why some companies, like Taco Bell, offer chances to win prizes when a customer completes one. Rather than providing cash rewards for surveys, you can have a chatbot conduct surveys for you.
Your bot can be scripted to offer surveys at important points in an interaction with a user. You can have it ask the customer questions about the quality of your services or products and even have the user rate the helpfulness of their interaction with the chatbot. Working a few survey questions into the conversation between a bot and a user is a less intrusive way to ask for opinions without having to offer a reward.
Listen and learn from your bot
Bots are always collecting information about your customers and about their own performance. Data gathered by chatbots can help you increase your sales, website traffic, and customer satisfaction—without the need for additional staff. Listening to your bot can even help save you money and increase your brand presence.