Chatbots look good on paper. Still, you shouldn’t invest in any technology – no matter how appealing – without first estimating what the ROI of that product will be. Here’s how to calculate the ROI of Chatbots, so that you will have an idea of how much money Chatbots can save your business:
1. IDENTIFY ELIGIBLE QUERIES
First, examine what queries your company currently receives over live chat. Then, identify the top simple queries that could be resolved by a Chatbot. These queries could include questions about return policy, shipping times and rates, store location and other noncomplicated issues. These questions may already be answered in your present knowledge base, or consist of easily available and communicable information.
Depending on your bot’s offerings, queries may also be expanded to include customer questions about personalized account information. Bots with these capabilities can give more specific responses to queries by consulting the customer’s available details. Applicable questions may include the following:
As bots grow more advanced, some may even be able to complete simple tasks for customers, such as changing passwords or billing addresses upon request.
By familiarizing yourself with a Chatbot’s present capabilities, you will be able to better categorize what sort of questions it will be able to answer. When identifying these questions, it helps to ask live chat agents for their input, since they are experts on common queries. Another way to do this is by asking agents to identify or tag eligible queries post-chat in wrap-up mode.
2. CALCULATE THE PERCENTAGE OF CHATS DEALING WITH ELIGIBLE QUERIES
Once you know what questions to look out for, it’s time to calculate how many of your chats are made up of simple (what a bot could answer) vs complex queries (what an agent should answer). What percentage of your total chat volume do the Chatbot-friendly queries make up? Are they 10%? Twenty percent? Fifty percent?
The higher the number, the more your company stands to gain in operational efficiencies for the dollar value of the Chatbot.
For the sake of this example, let’s assume that your company receives 6000 chats per month. The eligible questions that you identified make up 3000 out of 6000 chats in one month, or 50%.
3. CALCULATE AGENT TIME SPENT ON ELIGIBLE CHATS
Once you have this number, it’s time to focus on the people who are currently handling these chats: the agents. How much time do agents spend handling eligible chats daily, monthly, or yearly?
Calculate the estimated handle time for these types of simple inquiries. if your data suite is advanced enough, you might be able to tie in the simple query types taken from wrap-up data and find out the handle time totals for each of these categories.
Continuing the example above, let’s say that your agents spent 10 minutes on average on each of these 3000 chats (the average chat time is 12 minutes and 20 seconds according to our latest benchmark), or 30,000 minutes total (that’s 500 hours).
4. ESTIMATE THE ANNUAL COST OF HANDLING ELIGIBLE CHATS
To see how much a Chatbot will save you, you have to look at what you’re currently spending. That starts with measuring the average hourly cost of a live chat agent.
Multiply the average hourly pay of your live chat agent by the number of hours spent on these kinds of interactions. Once you have done that, figure out your annual cost by multiplying your monthly cost by 12.
For example, imagine that the average hourly cost of a live chat agent at your company is $15 per hour. Multiply that by 500 hours (your time spent per month on Chatbot-friendly interactions), and you get $7,500. To get the approximate annual cost of those interactions, multiply that by 12 (months) and you get $90,000.
In this example, $90,000 is what it costs per year for your company to handle questions over live chat that a Chatbot is capable of handling on its own.
5. TIME TO COMPARE: HOW DOES THE CHATBOT STACK UP?
Once you know how much your company is spending handling simple queries, it’s time to compare that annual rate to what you might spend on a Chatbot. For this part, consult a Chatbot provider on the annual fee of their Chatbot, and place the two values side-by-side.
For example, let’s say your prospective Chatbot costs $50,000 per year. In this scenario, it has clear advantages over continuing to pay $90,000. The $50,000 that you would spend on a Chatbot is the equivalent of nearly two full-time human agents (and this doesn’t account for employee turnover and training costs).
Don’t forget that Chatbots also have their own deployment costs, which you may want to factor in as an aside when calculating your first year’s ROI.
Incoming chat counts are rising every year – from 2016 to 2017, businesses saw a 180% increase in average chats per month. As these numbers go up, so will the amount of easily answered chats that your Chatbot is capable of handling. And, as Chatbot providers further push the boundaries of what customer-facing AI systems are capable of, Chatbots will be able to handle increasingly complex queries. This means that even including any Chatbot development and maintenance costs, your bot will grow more profitable as the years go by.
Remember that this ROI calculation model only works if your Chatbot can solve the incoming queries that it is designed to, and not make your queue worse. To ensure proper use of your Chatbot, implementation and ongoing learning are key. Be sure to map out your Chatbot’s journey, and make use of customer feedback. By simply asking your customers to rate whether an answer was helpful or not, you and your Chatbot can make tremendous strides in learning what is working and what isn’t.