Although chatbots improve customer experience, provide 24/7 support, sell products, increase website conversion rates and automate many other things, small businesses still need to be convinced to test the technology.
In fact, 80% of internet users interacted with a chatbot at some point, but it’s still crazy to think that only 23% of customer service companies use chatbots. There’s more, only 0.5% of B2B companies integrated chatbots into their strategy.
This article briefly explains the basics of chatbots and draws on compelling use cases to show how conversational AI chatbots can be integrated into your marketing strategy.
🤖 What is conversational AI?
In essence, a conversational AI is a service for communicating with customers based on predetermined rules or artificial intelligence (AI). Using large volumes of data, machine learning, and natural language processing (NLP), conversational AI imitates human conversations, translating complex speech and text inputs into actionable outputs.
Chatbots usually reside in a website’s contact section, voice assistants like Amazon Alexa or Apple Siri, or messenger apps like Facebook or Telegram.
The applications for this technology are vast; that’s why we will focus on conversational AI chatbots in this article.
👩💻 How are chatbots built?
A chatbot’s system requirements and architecture should reflect the customer’s intent. Platforms such as website, Facebook Messenger, WhatsApp, Telegram, or Amazon Alexa are used by a user for inputting information.
The customer intent can be anything from a simple question about the services or something more complex like a product recommendation or account changes.
Thence the chatbot module analyzes a message and generates a response using several functions of the back-end module. A dialogue module aims to grasp what’s being said, process the user’s intent and fire back a reply imitating human speech.
I would also like to point out that dialogue modules use many processes, such as text normalization, morphological analysis, semantic affinity patterns, and language processing through machine learning (ML), natural language understanding (NLU), natural language processing (NLP), natural language generation (NLG).
Conversational chatbots can also be integrated with many web applications through API, extending their capabilities.
So if a user wants to see specific products, the chatbot can extract data from your product database and show products that match user’s query in an interactive interface, including carousel with photos, videos and product-specific textual information.
If the user decides to buy the product, the chatbot can initiate a checkout process inside the same dialogue window, enabling a truly omnichannel shopping experience.
🤼♂️ Traditional Chatbot vs conversational AI chatbot
There are primarily two types of chatbots: ‘traditional’ and ‘intelligent or AI’. Their fundamental differences can be summarized as follows:
- Traditional chatbots are rule-based and serve a very specific purpose – to automate recurring tasks, like answering frequently asked questions or scheduling appointments. They cost less to implement, and usually, their architecture is simple and easy to build. Although they have some AI capabilities, a human agent still must write scripts and provide a list of keywords that would trigger different workflows.
- Conversational AI chatbots teach themselves progressively through reinforcement learning. So the customer can use their own language, and the system would easily understand them and deliver an appropriate response. One study predicts that the global conversational AI market will reach $32.62 billion by 2030, a whopping 464% increase over 2020.
We’ve made an infographic to help you see the main differences between traditional and conversational AI chatbots.
|Conversational AI Chatbot||Traditional Chatbot|
|Omnichannel presence||Individual channel|
|Neural network enables deep learning of the customer||Rules-based, solves customer issues following predefined workflows|
|Handles a broad range of tasks||Algorithm handles a narrow range of tasks|
|Self-learning makes it possible to use for complex cases||Best for simple, rules-based use cases|
🦾 Four types of conversational AI chatbots
We classify conversational AI into four main types:
- Rules-based chatbots
- Chatbots have a less flexible conversational flow and can be predictable
- Conversational AI
- Algorithms enable these chatbots to improve their understanding of natural language over time without much input from an operator
- Virtual assistants
- Virtual assistants such as Siri, Amazon Alexa or Cortana use voice input to execute different commands
- Voice and mobile assistants
- Technologies that convert voice commands into machine-readable text
👌 What problem does conversational AI solve?
In the case of the service sector, chatbots aim to automate routine tasks, increase staff productivity and convert customers.
About 40% of staff productivity suffers from multitasking, as many believe they can save time by doing routine work while concentrating on large projects.
Scheduling meetings, answering frequent questions, providing customers with account information, and other routine tasks – all eat up time that could be spent on more pressing matters. So chatbots automate these routine tasks, allowing you to free up your time.
Regarding client engagement, chatbots simplify the communication between the customer and the company.
80% of businesses saw an increase in the number of leads after implementing marketing automation
World Economic Forum has made a pretty cool table illustrating the primary purposes and intent of conversational AI. We’ve modified examples to help you see them in the context of your business.
Purpose of AI chatbots
|Transactional||To sell||Online sales|
FAQ & problem resultion
|Communicative||To inform||Maps & transport|
Food & restaurant
Cultural & social activities
|Get personalized content/event/task recommendations|
Access content swiftly and seamlessly
Book tickets or register for an event
|Educative||To learn||Teaching assistants |
|“How to” quick tips|
Special needs with voice commands
Speak to learn
Bounce off ideas, the bot will come up with something cool
|Social||To connect||Social networking |
|Verify users before they can proceed with checkout|
Get the latest news stories on topics you care about
The bot talks to a person who can’t fall asleep
|Play||To entertain||Gaming |
Play word-based games
|Quickly calculate mortgage or interest rate, etc |
Help users manage transactions
Develop a savings plan
|Diagnostic||To identify||Medication use|
|Help users create better meditation practices |
Examine user’s health by receiving their symptoms
Qualify leads eligibility for certain procedures
|Behavioural||To change||Healthy eating |
|Develop a nutritional plan |
A user can talk to a bot to cope with anxiety & other mental issues
Create a personalized training plan
🖐 5 conversational AI use cases
⬇ In healthcare
AI chatbots can be used to find a new family doctor and register a patient. AI will generate a list of available family doctors based on patient’s input such as name, postal code, insurance type, and healthcare number.
The user can also request additional information about each doctor, and when the user wants to register with the family doctor, the bot can help make an appointment.
Chatbots are secure as they utilize end-to-end encryption and often require user authentication to process personal data.
Schedule an Appointment
Appointment scheduling is seen as one of the chatbot’s most common features, as it’s able to integrate with a clinic’s booking platform, offering patients a simple appointment booking process.
Other functionalities may include looking up existing appointments, cancelling or making scheduling changes.
This is especially useful for patients looking for after-hours appointment information.
Healthcare professionals spend around 35% of their time documenting patient data.
Intelligent chatbots can help capture required patient data for a physician and facilitate embedding the data into the documentation.
This helps save significant time and helps staff focus more on critical tasks.
A study on customer experience found that 92% of customers prefer using a knowledge base for self-evaluation instead of going to or calling a medical clinic.
AI can lead a patient through a series of questions in a logical order, like a real doctor, to understand their condition.
By analyzing symptoms and even recognizing injury patterns from images, AI can provide the answer and connect a patient to a real doctor in case of an emergency.
⬇ In ecommerce
Answer frequently asked questions
“What’s your return policy?” or “what is the difference between X and Z products?” Many similar questions are being asked by customers every day.
Chatbots can quickly address common questions, and help customers resolve any issues like changing their order status, shipping address and more.
By 2023, chatbots are expected to generate over $100 billion in eCommerce transactions.
Chatbots can recommend products based on defined conditions:
black shirt, new collection, budget <$90, 4 stars and above
Their ability to recognize voice messages allow customers to browse personalized products faster.
After processing the request, the bot generates products in a carousel view.
The user proceeds with payment inside a dialogue box.
Given that your store already has an automated order processing system, all you need to do is to ensure the payment has gone through, and the customer received their order.
Personalized product suggestions based on purchase history
A rather interesting case is the integration between CRM and chatbot, offering personalized product recommendations based on the customer ID and budget.
The system runs a query in the CRM to find the customer’s ID and past purchases.
Then, using an ecommerce module, it searches for products that match the customer’s input.
The customer is presented with a carousel with personalized products on the front end.
⬇ In tourism & hospitality
According to one study, 87% of travellers would use a travel chatbot if it could save them time and money.
Below are some of the use cases that create value for travellers:
Act as an itinerary planner
Skyscanner’s Facebook chatbot does a great job of suggesting the cheapest destinations from the nearest airport and categorizing flights by price, location and flight time.
Another feature worth mentioning is called anywhere, when a user types this word, the chatbot returns a random destination anywhere in the world.
Using Skyscanner as an example, it makes sense for hotels to include chatbots in their strategy.
Think about it, whenever a guest has a question about their booking status, room condition, restaurant availability, or nearby attractions, the chatbot can become their best travel buddy.
According to a new study, 59% of unemployed hospitality workers are looking for jobs in other industries unrelated to tourism and hospitality.
This figure presents a number of challenges, including reduced staffing, increased load on frontline hotel employees, longer wait times, and unsatisfied guests at hotel doors.
While chatbots aren’t a panacea for all the problems hotel management teams face today, they can alleviate the manual work associated with reservations and help your staff focus on providing guests with a memorable experience.
Some other use cases in hospitality industry include:
- Guide customers through the booking process
- Handle refunds and cancellations
- Display available hotels and flights
- Answer frequently asked questions
- Upsell products and services
- Handoff customers to a live agent
- Update customers on the status of their trip
⬇ In real estate
Chatbots can be used to qualify leads with greater accuracy and speed.
For your potential leads, the experience is new and thus exciting. While for your team, the experience offers more time to nurture quality leads.
Another great benefit of using chatbots in real estate is the ability to collect data on those who didn’t qualify as your sales lead.
By analyzing their responses, you can quickly detect the behavioural pattern that brings poor-quality leads into your funnel.
There’s a 451% increase in qualified leads if marketing automation is paired with chatbots. – Hubspot
Automated appointment setting:
Appointment scheduling is a great feature that chatbots offer to real estate agents.
Get more calls and schedule more viewings by having the clients book directly inside a dialogue window.
Integration with a calendar or a special booking software is easy, thanks to webhooks and special web-based automation task software like Zippier and IFTTT.
Once the client has booked a viewing, the chatbot automatically adds an event, sends out invitations to both parties and populates the calendar with relevant information required for the viewing.
A chatbot platform Chatfuel saw an increase of 567% in appointments by adding a bot to one of their client’s websites. They went from around 30 to 200 appointments a month, all through the power of bots.
By integrating chatbots with your listing directory, you can show properties to your clients based on multiple parameters:
- 🏚 Square footage of a home
- 🏠 Number of rooms
- 🏡 Information on neighbourhoods like schools, shops, crime rate, etc.
- ⛪ Price, or potential future value of a house, or calculated mortgage and insurance rates
- 🕌 On average, how much is a utility bill for similar houses in the area?
Some future homeowners consider over 80 factors when buying a house. Provided you have data for the bot, it will become a powerful lead nurturing machine for you.
⬇ In construction
One peculiar use case of a chatbot in the construction industry is daily reporting and information management.
This research paper describes that updating, analyzing and managing construction-related information is one of the critical success factors in project management.
Some construction contractors started using instant messaging (IM) applications like Slack, WhatsApp, and others to share daily construction information. However, one significant drawback of IM apps is that subcontractors enter data in an unstructured form.
IM applications are not designed to organize data in a structured form. That’s why these channels are not suited as data management systems.
The authors propose to collect and process the required information through conversational AI chatbots, which would help subcontractors effortlessly submit reports, and project managers automatically generate and share daily reports with the key stakeholders.
I strongly recommend reading their paper to see the implications of chatbots in the construction industry.
Some benefits of chatbots in the construction site
- Real-time info about machinery and team performance
- Construction site activity-related photo and progress sharing
- Notification for urgent needs and requirements
- Easy sharing of contractor, warehouse and material info
- Publishing daily progress reports efficiently and accurately
- Tracking real-time activity progress
- Access to construction documents and drawings
👍 Advantages of conversational AI chatbots
Now let’s quickly run through some of the top advantages of conversational AI chatbots:
- ✔ Help to sell more
- ✔ Increase sales by 67%, rendering your lead gen more effective
- ✔ As submission forms increase engagement up to 90%
- ✔ Can reduce customer service costs up to 30%
- ✔ Save time, the bot at JPMorgan saved over 360,000 hours a year doing tasks in just a matter of seconds
👎 Disadvantages of intelligent chatbots
Not a single system in this world comes with only positive qualities, below are some disadvantages of chatbots:
- ✖ Takes time to implement, be patient
- ✖ Not humans, so remember that they’re not able to solve all problems
- ✖ Require consistent maintenance, so if you run a small business, you might think about outsourcing some help
- ✖ Can be costly depending on your project