Try : Insurtech, Application Development









Dev Ops(2)

Enterprise Solution(12)




Augmented Reality(7)

Customer Journey(7)


User Experience(19)

AI in Insurance(19)


Product Innovation(27)


Artificial Intelligence(79)



Cognitive Computing(6)

Computer Vision(5)

Data Science(11)


Intelligent Automation(19)

Machine Learning(41)

Natural Language Processing(5)

Here’s How You Measure the ROI from Chatbots

Nidhi Agrawal
5 minutes, 6 seconds read

IBM reports that globally businesses spend over $1.3 trillion/year to handle roughly 265 billion customer calls. Chatbots spring up to minimize the expenditure on handling customer queries, especially the most redundant ones.

It’s quite common for businesses to assess the return on investment before adopting new technology.

However, ROI from chatbots may vary according to the purpose it serves. For example, an insurance chatbot ROI differs from that of an HR chatbot. Here are certain parameters to consider for calculating the return on investment from chatbots.

#1 Average Human Live-chat Cost

The total number of tickets raised per month and the number of agents involved gives an idea of the average price per contact.

According to Help Desk Institute, the average cost/minute for a live chat is $1.05, while the average cost per chat session is $16.80. Assuming an organization handles 10,000 chats in a month, the cost incurred sums up to $168,000/month.

Depending on the number of people involved and their compensation, you can calculate the amount you’re spending on your organization’s customer support. Here’s a salary reference, which can be used in further calculations.

sample customer support operational cost

The salaries mentioned are referred from Job Futuromat 2019 wrt 12 months, 18 working days, 8 hours.

The actual operational cost also depends on material resources invested like office space, conveyance, communications, gadgets, etc. You can consider these aspects on your chatbot ROI calculator.

#2 Bot Installation Cost

The phases of bot installation cost involves brainstorming sessions, integration, and training both bots and agents.

During kick-off sessions, stakeholders discuss the scope of the bot, define goals and responsibilities, and make a project plan. After this, programmers and managers integrate the bot on the organization’s website and other platforms. Customizing the bot according to the client’s support cases covers the bot training phase. Testing the bot and training agents to use it are also factored into the ‘bot’ installation costs.

According to Ometrics, the average development charge for a chatbot may range from $1,000 to $5,000. But, this is a one-time charge, and after that the bot-developer may bill for maintenance charges.

chatbot roi calculator: installation cost

If the chatbot requires a higher level of customization, then the bot-developer may also claim additional charges. Also, the number of days spent for bot installation varies according to industries and organizations.

#3 Gains through Bots

Here we’re assuming all the customer queries are routed through the bot and it is accurate 50% of the time. Out of the 50% queries handled by a bot, if half of them are self-served and the remaining required human intervention, then monthly gains from the bot can be-

chatbot roi calculator: gains from chatbot

You can find the exact cases and accuracy from your bot’s analytics dashboard.

#4 Monthly Maintenance Cost

Like humans, bots also require human assistance for its successful operation. Its monthly maintenance cost is a summation of the organization’s human resources it needs and developer’s charges. Here, let’s assume a chatbot maintenance fee, which ranges from $100 to $1,000 a month. Similar to the bot development charges, maintenance fees vary according to bot capabilities.

chatbot roi calculator: montly maintenance cost

#5 Chatbots Return on Investment Calculation

The return on investment is a ratio of benefit from the investment to the cost of investment. It evaluates the efficiency of an investment. Mathematically, ROI = (Current Value of Investment – Cost of Investment) / Cost of Investment.

Since chatbots incur a one-time development cost and recurring monthly maintenance cost, here’s the chatbot ROI calculation from both perspectives.

Chatbot ROI during the first month: This includes the bot installation charges. 

For the above case,

ROI = (Gains through bot – Installation charge – maintenance charge)/(installation charge + maintenance charge)

ROI = ($63,000 – $9,292 – $3,647)/($9,292 – $3,647)

ROI = 3.9 or 390%

Chatbot ROI after the first month: This excludes the bot installation charges. 

For the above case,

ROI = (Gains through bot – maintenance charge)/(maintenance charge)

ROI = ($63,000 – $3,647)/($3,647)

ROI = 16.3 or 1630%

Using this method, you can build your own chatbot ROI calculator considering your own business parameters.

NLP and AI-powered chatbots can yield a better return on investment. For instance, Religare has incorporated a service chatbot on its Web portal and WhatsApp integration to handle customer queries. It has resulted in 10 times more customer interaction and 5 times more sales conversion.


For the above case, where bots are able to handle 50% of customer queries, there’s a direct 50% capital gain to the organization. The human-time saved can be utilized for more productive tasks, which can eventually accelerate the organization’s productivity. 

Powerful bots result in better success rates for customer facing operations. For example, Diageo’s iDia chatbot has led to a 55% drop in help desk tickets. 

Here are more enterprise chatbot use cases.

Researchers predict that by 2025, chatbots will accomplish more than 90% of the B2C interactions. Also, chatbots can cut operational costs by more than $8 billion per year in the next three years.

We specialize in developing industry-specific AI-powered chatbots. Drop us a word at to learn more.


Knowledge thats worth delivered in your inbox

Across the Insurance ecosystem, a special fraction within the industry is noteworthy for its adoption of new technologies ahead of others. However slow but sure, uberization of insurance has conventionally demonstrated a greater inclination towards digitization. Insurers now more than ever, need big data-driven insights to assess risk, reduce claims, and create value for their customers. 

92% of the C-Level Executives are increasing their pace of investment in big data and AI.

NewVantage Partners Executive Survey 2019 

Artificial Intelligence has brought about revolutionary benefits in the Insurance industry.

AI enriched solutions can remove the ceiling caps on collaboration, removes manual dependencies and report errors.

However, organizations today are facing a lot of challenges in reaping the actual benefits of AI.

5 Challenges for AI implementation for Insurers

5 AI Implementation Challenges in Insurance

Lack of Quality training data

AI can improve productivity and help in decision making through training datasets. According to the survey of the Dataconomy, nearly 81% of 225 data scientists found the process of AI training more difficult than expected even with the data they had. Around 76% were struggling to label and interpret the training data.

Clean vision, Process, and Support from Executive Leadership

AI is not a one time process. Maximum benefits can be reaped out of AI through clear vision, dedicated time, patience and guided leadership from industry experts and AI thought leaders.

Data in-silos

Organizational silos are ill-advised and are proven constrictive barriers to operational productivity & efficiency. Most businesses that have data kept in silos face challenges in collaboration, execution, and measurement of their bigger picture goals. 

Technology & Vendor selection

AI has grown sharp enough to penetrate through the organizations. As AI success stories are becoming numerous investment in AI is also getting higher. However big the hype is, does AI implementation suits your business process or not – is the biggest question. The insurtech industries have continued its growth trajectory in 2019; reaching a funding of $6B. With the help of these insurtech service firms, Insurance organizations have made progress, tackling the age-old insurance ills with AI-powered innovations.

People, Expertise and Technical competency

‘Skills and talent’ in the field of AI is the main barrier for AI transformation in their business.

Still playing catch-up to the US, China, and Japan — India has doubled its AI  workforce over the past few years to nearly 72,000 skilled professionals in 2019. 

Are you facing challenges with your Insurance process but have no idea where the disconnect is? Is your Insurance business process ripe for AI in the year 2020?

What is the right approach?

Join our Webinar — AI for Data-driven Insurers: Challenges, Opportunities & the Way Forward hosted by our CEO, Parag Sharma as he addresses Insurance business leaders on the 13th of February, 2020.

Register for the live webinar by Parag Sharma (AI Thought Leader & CEO Mantra Labs). 


Knowledge thats worth delivered in your inbox

Ratemaking, or insurance pricing, is the process of fixing the rates or premiums that insurers charge for their policies. In insurance parlance, a unit of insurance represents a certain monetary value of coverage. Insurance companies usually base these on risk factors such as gender, age, etc. The Rate is simply the price per ‘unit of insurance’ for each unit exposed to liability. 

Typically, a unit of insurance (both in life and non-life) is equal to $1,000 worth of liability coverage. By that token, for 200 units of insurance purchased the liability coverage is $200,000. This value is the insurance ‘premium’. (This example is only to demonstrate the logic behind units of exposure, and is not an exact method for calculating premium value)

The cost of providing insurance coverage is actually unknown, which is why insurance rates are based on the predictions of future risk.  

Actuaries work wherever risk is present

Actuarial skills help measure the probability and risk of future events by understanding the past. They accomplish this by using probability theory, statistical analysis, and financial mathematics to predict future financial scenarios. 

Insurers rely on them, among other reasons, to determine the ‘gross premium’ value to collect from the customer that includes the premium amount (described earlier), a charge for covering losses and expenses (a fixture of any business) and a small margin of profit (to stay competitive). But insurers are also subject to regulations that limit how much they can actually charge customers. Being highly skilled in maths and statistics the actuary’s role is to determine the lowest possible premium that satisfies both the business and regulatory objectives.

Risk-Uncertainty Continuum

Source: Sam Gutterman, IAA Risk Book

Actuaries are essentially experts at managing risk, and owing to the fact that there are fewer actuaries in the World than most other professions — they are highly in demand. They lend their expertise to insurance, reinsurance, actuarial consultancies, investment, banking, regulatory bodies, rating agencies and government agencies. They are often attributed to the middle office, although it is not uncommon to find active roles in both the ‘front and middle’ office. 

Recently, they have also found greater roles in fast growing Internet startups and Big-Tech companies that are entering the insurance space. Take Gus Fuldner for instance, head of insurance at Uber and a highly sought after risk expert, who has a four-member actuarial team that is helping the company address new risks that are shaping their digital agenda. In fact, Uber believes in using actuaries with data science and predictive modelling skills to identify solutions for location tracking, driver monitoring, safety features, price determination, selfie-test for drivers to discourage account sharing, etc., among others.

Also read – Are Predictive Journeys moving beyond the hype?

Within the General Actuarial practice of Insurance there are 3 main disciplines — Pricing, Reserving and Capital. Pricing is prospective in nature, and it requires using statistical modelling to predict certain outcomes such as how much claims the insurer will have to pay. Reserving is perhaps more retrospective in nature, and involves applying statistical techniques for identifying how much money should be set aside for certain liabilities like claims. Capital actuaries, on the other hand, assess the valuation, solvency and future capital requirements of the insurance business.

New Product Development in Insurance

Insurance companies often respond to a growing market need or a potential technological disruptor when deciding new products/ tweaking old ones. They may be trying to address a certain business problem or planning new revenue streams for the organization. Typically, new products are built with the customer in mind. The more ‘benefit-rich’ it is, the easier it is to push on to the customer.

Normally, a group of business owners will first identify a broader business objective, let’s say — providing fire insurance protection for sub-urban, residential homeowners in North California. This may be a class of products that the insurer wants to open. In order to create this new product, they may want to study the market more carefully to understand what the risks involved are; if the product is beneficial to the target demographic, is profitable to the insurer, what is the expected value of claims, what insurance premium to collect, etc.

There are many forces external to the insurance company — economic trends, the agendas of independent agents, the activities of competitors, and the expectations and price sensitivity of the insurance market — which directly affect the premium volume and profitability of the product.

Dynamic Factors Influencing New Product Development in Insurance

Source: Deloitte Insights

To determine insurance rate levels and equitable rating plans, ratemaking becomes essential. Statistical & forecasting models are created to analyze historical premiums, claims, demographic changes, property valuations, zonal structuring, and regulatory forces. Generalized linear models, clustering, classification, and regression trees are some examples of modeling techniques used to study high volumes of past data. 

Based on these models, an actuary can predict loss ratios on a sample population that represents the insurer’s target audience. With this information, cash flows can be projected on the product. The insurance rate can also be calculated that will cover all future loss costs, contingency loads, and profits required to sustain an insurance product. Ultimately, the actuary will try to build a high level of confidence in the likelihood of a loss occurring. 

This blog is a two-part series on new product development in insurance. In the next part, we will take a more focused view of the product development actuary’s role in creating new insurance products.


Knowledge thats worth delivered in your inbox

Loading More Posts ...
Go Top

May i help you?

Our Website is
Best Experienced on
Chrome & Safari

safari icon