4-Step Formula for Calculating Your Chatbot ROI

A meeting that could’ve been an email has the same energy as chats that could’ve been handled by a chatbot, full stop.

Valuable time and resources are sometimes spent on answering the same questions over and over when those simple inquiries could have been handled by automation.

cartoon of R2D2

Chatbot adoption is predicted to lead to annual business cost savings of $11 billion by 2023, up from an estimated $6 billion in 2018. Which leads to the bigger question — “how much can my business save and earn with a chatbot?”

We’ll break it down equation-wise so you can get a ballpark idea of what you can save and earn. If you’re ready to learn how to calculate the return on investment from implementing a chatbot, then let’s get our math on.

How to calculate chatbot ROI

We highly recommend having several months experience with live chat for an accurate calculation. Phone and email inquiries will also give you a sense, but chats are easier to determine if they’re eligible for chatbot takeover. Live chat experience will also make the introduction of chatbots a lot easier.

1
Calculate number of inquiries that can be answered by a chatbot

Chatbots are ideal for answering simple questions to free up your agents’ chat slots and capture more leads. To get an idea of how many chat inquiries a chatbot can help with, we recommend setting up a system that marks these eligible questions.

In our solution Userlike , for example, agents can tag “FAQ” or “canned” on conversations in which they used a canned message. Then you can check how often that tag is used in your Dashboard analytics. That result will give you an idea of your automation potential.

Once you’ve identified chatbot-worthy questions, you can perform the calculation in the next step. We recommend using at least 100 conversations for a quick estimate.

You’ll use the following formula: number of messages that can be answered by the FAQ / total inquiries

Let’s say out of those 100 conversations 30 used canned responses. Those 30 responses / 100 conversations = 0.3 or 30%. We will call this the FAQ ratio.

2
Determine current costs spent on answering simple inquiries

Those FAQs your agents are taking over can be costly, both in money and time. This next step in determining your chatbot ROI will help you calculate how much you’re spending on answering simple questions per month and year. Then we can multiply that result by your FAQ ratio.

First, you’ll need to determine the average monthly or yearly wage of your support agents. We’ll use a monthly wage of $2000 as an example.

three cartoon chatbots

Next, estimate the percentage of their work time spent on chat support. If an average employee spends 100% of their day on live chat, then you would just multiple their salary by one (which is just still 2000 in our case) and move on to the next step. But if an agent spends, let’s say, 80% on support, then you would multiply their wage by 0.8%.

Lastly, multiply the wage from the percentage of chats that could be taken over by chatbots from step one.

This is what the equation looks like:

Average monthly wage: $2000
80% spent on chat support
0.8*2000 = $1600
1600*0.3 = $480 a month spent on chatbot-eligible conversations, $5,760 a year on average

3
Estimate the extra lead value you can gain from chatbots

Chatbots often answer simple inquiries but they can be the first contact for complex questions too. This increases the amount of hot leads your agents are taking over instead of wasting time on collecting standard information.

For example, our live chat solution Userlike lets you set your chatbot parameters, like forwarding specific questions to the appropriate agent. It can also collect pertinent information from the contact before forwarding, like their name and email address.

image of a chatbot forwarding the chat to an available agent

Chatbots are also available 24/7 and can help capture leads outside of your business hours. Late night shoppers can be guided through a curated sales funnel and chatbots can then forward serious buyers to your agents.

Here’s how to evaluate the extra lead value you can earn from implementing a chatbot:

Determine your lead to chat ratio. You can do this by having your agents mark relevant conversations with “lead” to see in your analytics what percentage of conversations are leads.

Then estimate how many leads that may have been missed while your agents were offline. For example, in Userlike we have an analytics section called "missed opportunities" to make tracking these leads easier. These are the potential customers a chatbot will capture.

screenshot of missed connections section in userlike

Multiply the number of missed opportunities by your leads/conversations ratio. Then multiply that result by the average value per lead.

This is what the equation looks like:

Let’s say that your lead/chat ratio is 20%, and on average you have 500 missed chat opportunities per month. Then we can assume that 20% (your lead/chat ratio) of those missed chats would have been leads, so 500*0.2 = 100 leads.

Then let’s say the average value per lead is $350. You would multiply 100*350 for a result of $35,000 extra generated lead value.

4
Choose your chatbot provider

Now that you have an idea of what agents answering canned messages is costing you and the extra lead value you could generate with a bot, consider the type of chatbot provider you need. This will determine implementation and maintenance costs.

There are different starting cost differences between decision tree and contextual bots, so in the context of calculating ROI, let’s consider the potential monthly costs.

Here’s a simple breakdown of what monthly costs could entail:

  • Monthly fee: Often ranging from Free to $500+ per month
  • Extra chats costs: e.g. $10 per 1,000 chats per month
  • Labor: Consider the hours spent on development times employee salaries (unless you hire a development team for a set amount)

Your costs will also vary if you plan to hire a team or the chatbot provider to build and integrate the chatbot for you.

Another chatbot option is a bot that is already integrated into your live chat. Userlike has one as well as Freshworks , ChatFlow and iAdvize . If you have a Business plan with Userlike, you can use our Logic Chatbot for a simple, personalized decision tree chatbot. You can then offer automated service across your customers’ favorite channels like WhatsApp, Facebook Messenger and more.

infographic of messaging support in the adoption curve, a major customer service trend
BMW uses Userlike’s Logic Bot to navigate site visitors

We highly recommend connecting your chatbot to your live chat solution. In our research , we learned that customers are okay with talking to chatbots first as long as there’s an easy way to escalate the conversation to an agent. A chatbot tied to your live chat/messaging solution is going to have more value than just a standalone chatbot.

Calculate your lead value and the time spent on canned messages above first to help you decide what type of chatbot is most suitable for your business.

Calculate your chatbot ROI and get in touch with us

If you need help driving organic traffic to your website and/or find that a healthy selection of your inquiries are FAQs, then a chatbot may be a good fit for your business. Chatbots are there for your customers 24/7 and can help answer simple inquiries, freeing up chat slots for your agents. By guiding visitors through your website, your chatbot can help turn curiosity into legitimate interest and generate hot leads for your agents to focus on.

As chatbot enthusiasts, we’ve read many chatbot success stories in our years of research and have since created our own solution. Userlike can help get you started with a chatbot based on YAML so you can write a customized script in no time.

If you already have another chatbot service in mind, we can tell you more about how you can use our chatbot API to connect it to our solution. Sign up with Userlike and contact our customer service team if you’re ready to talk bots 🤖