1.Analyze your customer service queries. Identify simple, repetitive questions that can be handled by a customer service bot.

2.Calculate the percentage of chats that are made up of these simple queries.

For example, your support gets 10,000 queries monthly, and you identified that 80% or 8,000 questions are eligible for a chatbot to take over.

3.Calculate the agent time spent on those basic, repetitive requests.

After you identified that you have 8,000 tedious, repetitive queries, you can calculate how much agent time goes to these questions.

For example, on average, your agent answers the question in 4 minutes. So, 8,000 queries x 4mins = 32,000 minutes or 533 hours of your teams’ time goes to FAQ’s.

If you say that customer service bot answers the questions twice as fast as agents, then you’ll have: 8,000 queries x 2mins = 16,000 minutes or 267 hours of bot’s time.

4. Estimate the annual costsof handling these requests by multiplying the hourly salary of the agent by hours spent monthly on these requests. Then, multiply the number by 12, and you’ll get the annual costs of answering FAQs by your agents.

Let’s say your agent earns 15$/hour. So, 15$ x 533 hours = 7 995$. And that’s how much costs to answer FAQs monthly. Now, let’s calculate the annual spendings: 7 995$ x 12= 95 940$ per year!

5. Now, compare the annual costs with quoted customer service bot cost. Don’t forget to count set up, integration, and any additional chatbot fees.