Everything you need to know about custom chatbots for your business
Every single business needs to facilitate good communication between different stakeholders. Salespeople need to be able to communicate well with their customers to close deals. Marketers need to be able to create well-informed content so they can communicate their message effectively. CEOs need to communicate vision and strategy to their team so that they can ensure their organization moves forward as one single, well-oiled machine. Good communication is extremely important to making sure everyone is well-aligned. And since writing is one of the most common ways to communicate online, it's imperative that companies implement tools that can facilitate good written communication.
Custom chatbots are a great way to facilitate communication
Businesses have increased their use of chatbots in the past few years, and as large language models (LLMs) like ChatGPT and Bard have become more prevalent, new, companies can start to implement some extremely capable and sophisticated custom chatbots to help them save time, improve ops, and grow sales. Chatbots are becoming so important that today, around 39% of all chats between consumers and businesses are done through chatbots (Comm100↗).
In this guide, we’ll cover everything you need to know about using chatbots tailored to your unique business needs. We'll go over the benefits that different types of chatbots provide, dive deep into how internal chatbots differ from external bots, and outline both off-the-shelf and build-your-own solutions. By the end, you'll have a good sense for where you can implement custom chatbots for your own business and how to best do so.
Chatbots can offer a huge number of benefits
Chatbots present a huge opportunity to transform key business functions by being always-available software agents that can help you 24/7. Here are some of the key ways that a custom chatbot can help you grow your business:
- Self-service customer support: Providing customers with the ability to solve their own problems using a knowledgeable, informed chatbot is one of the most common use cases for chatbots. By letting customers ask their own questions and get quick, automatic answers, businesses can reduce the amount of manual work that their employees need to support customer requests.
- Internal research: The right chatbots, connected to internal documents and databases, can allow your team to ask questions in natural language and get trusted answers back, along with links to additional resources that they might need to better complete their tasks. These types of bots can also serve to implement a single, centralized point of access for your data, particularly when they’re connected to all of your internal and external systems.
- Business intelligence: With the advent of better AI techniques for natural language, businesses are now starting to implement technologies that let their employees ask questions about their metrics and data and retrieve reports and visualizations that would normally require SQL or BI tools.
- More effective business development: The right chatbots can help you collect leads and sign demos more quickly by putting your prospective customers into a flow where they can easily go from answering their own questions about your business to booking time on your calendar when they’re ready to move forward.
In general, custom chatbots are good for improving productivity and sales, and can thus be separated into internal and external bots. In the rest of this article, we’ll differentiate between the two, telling you where each type is best suited and how to implement the right one for your business.
Internal and external bots
In general, you can think about bots in two different ways: those that talk to your external stakeholders (website visitors, customers, etc.) and those that talk to your internal team members. It’s possible that the same bot might be capable of doing both, so we’ll group bots like that into “external” bots, since it’s important to specifically consider the experience that your customers get when they use these bots.
External bots drive business development
You’ve probably used an “external” chatbot at some point yourself. You’ll usually see these types of bots in the “Help” section of a software service or tool. Most companies like to use these bots as a way to let customers solve their own problems so that they don’t have to call the support center, since that’s both a nuisance for the customer and also a cost center for the company.
Chances are, the bot didn’t do a great job in helping you, because up until recently, most bots were simply a fancy way of doing a keyword search against a company’s help documents. They tend to not be very smart or plugged into the data for your own account. Bots like these generally lead to a lot of frustration on the part of the customer, and will either lead to a call to the support center anyway, or will simply result in the silent loss of a customer or prospective customer.
External bots like these provide little value on top of existing help resources, and customers can be very sensitive to chatting with these tools, since they’re keenly aware of how much of a waste of time they can end up being. It’s likely that many of these bots are created in a “rules-based” format, where you as the customer support manager need to define the exact questions, answers, and flows that the bot will follow.
From your company’s perspective, though, that may not be such a bad thing. If you’re implementing a bot for a very limited set of reasons, you might only need a rule-following bot that responds in a well-known and predictable way every time.
It’s important to think through the reasons you’re implementing an external bot in the first place. By determining the objectives you’re trying to achieve with such a bot, you’ll be able to make a better decision about the best bot to implement for your company. It’s important to consider the following items when deciding what you want your custom bot to do:
- What’s the primary objective?
Are you trying to capture leads? Let customers answer their own questions? Reduce headcount at your support center? By defining your main metric for success, you’ll be able to more effectively choose the best custom bot for your needs. - Who will be talking to the bot?
Will it be logged in customers only? Will it be for visitors who come to your website? Depending on the audience, you might need to support different levels of integration. For example, logged in users might expect the bot to know details about their account, but a generic, website-based bot may only need to be able to answer more generic questions about your business. - Do you need to collect leads?
If you need to collect contact info for visitors who chat with your bot, you’ll need a lot of additional capabilities than you would if you just needed a simple bot to answer questions. - Do you need the ability for a human to take over control of the chat?
If you need the ability for someone from your team to be able to jump into a chat, you’re going to need a much more sophisticated bot, which will be more expensive than if you had a simple bot that only supported automatic conversations. - Do you want to support the ability to follow up? To see past conversations?
A full-featured support bot needs the ability to provide contextual information to your team so that they can follow up, or at least see the historical context of the bot’s conversations. - Do you need to integrate with your internal chat tools?
Some bots have the ability to integrate their conversations with your internal tools, like Slack, so that you can not only jump into a conversation but do it from the tools you’re already using. - How is the bot supposed to get its information?
Does it plug into your existing documentation? Can it even access that information? Is it behind a firewall or stored in a system that requires authentication, like Google Drive? Depending on how you want your custom bot to answer questions, you’re going to need to provide it with the right context and data, and if your bot doesn’t have access to that data to begin with, it’s a clunker right out of the box. - Do you need to configure it?
In an ideal world, you could simply connect your data, embed your bot into your website or web app, and let it do its thing. But historically, most chatbots haven’t been able to handle the wide array of customer requests that they need to, which means it has almost always been the responsibility of the business setting up the bot to set the rules and flows that it will follow. While this can ensure the bot provides well-known and trusted answers, it also leads to very inflexible and brittle bots that customers don’t like.
As you can tell, there are a lot of things to think about when implementing a customer-facing custom bot, and it’s rarely as easy as just setting it up and forgetting it exists. But with the right bot, you can dramatically improve your customer experience and increase your pipeline.
How can you find the best external bots?
There are several custom chatbot providers that can help you implement an external facing bot to improve your customer experience.
Intercom is the leader in this space and provides a comprehensive solution for custom chatbots in addition to customer support backend tools, help center articles, leadgen, and more. They power a large number of chatbots across the internet, and allow you to specify a rules-based approach to how your bot should respond to customers, and also recently rolled out Fin, an AI-powered chatbot that doesn’t require the same level of configuration as their primary offering.
Unfortunately, they don’t have a free tier offering, so you’ll need to pay to use their product.
Crisp is a great alternative to Intercom, as they provide a lot of the same features, with a great chat interface, ticketing system, customer support features, and they offer a free tier that you can use for a long time before you need to upgrade.
Internal chatbots are great for boosting productivity
Just like external chatbots are great for helping customers help themselves, internal chatbots can help your team improve their own productivity and efficiency by making it easier for them to find what they need, when they need it.
Internal chatbots haven’t been widely used in the past, but with the advent of LLMs, we’re beginning to see a lot of new, productivity-focused chat tools crop up and make their way into companies’ day-to-day operations.
When you’re implementing an internal custom chatbot for your team, chances are you’re hoping to make it easier for your team members to find what they need or collaborate with each other with the help of an AI assistant that has context around your company’s data and/or procedures.
Because of this, it’s important to ensure that your internal chatbot has the following features:
- Search across your data
An internal custom chatbot needs to be able to find answers about key company products, offerings, and operations wherever they’re stored. As such, it needs to be able to access important company information wherever it’s stored and provide a comprehensive and consolidated view of the information available to your team members. - Easy to collaborate
Internal bots need to be able to provide information to the right people at the right time. They should be able to respond to multiple team members with the right context, and team members should be able to easily share information that they receive from a query to the chatbot with other team members. - Integration with your existing chat ecosystem
Good internal chatbots should be able to plug into your existing chat tools, like Slack or Google Chats. - Clear permission management
Internal bots should be able to implement good permissions management and it should be easy to know what data your internal chatbot can access in order to answer your team members’ questions.
Good internal chatbots will help your team move faster and spend less time searching for details. They should act as internal assistants that are well-versed in the company’s policies and operations and can lend a helping hand whenever they’re called upon.
Where do you find a good internal custom chatbot?
Locusive’s (hi, that’s us 👋) chatbot for Slack is a good choice for teams that want to be able to consolidate access to all of their data to a single, well-known chat interface. Our chatbot lives in your existing Slack workspace and you can provide it access to data sources that you’re already using, like Google Sheets, Google Drive, or Salesforce.
These examples show how you can start to think about implementing a custom chatbot for your business. But what if none of these options quite fit what you need for your company? In that case, you might want to build your own bot, and if you do, there’s a lot you’ll need to consider before you jump in.
Rolling your own custom chatbot
It’s possible that you’ll want to create your own custom chatbot that handles your company-specific use cases in its own way. While it is possible to build everything from scratch, you might be able to benefit from existing API providers that make it easier for you to focus on the basic infrastructure, while you focus on your custom functionality.
Components of a custom chatbot
When you decide to build your own custom chatbot, you’ll need to be prepared to create a lot of different components and pieces of infrastructure to make it all work. You’ll need to create a frontend to let users sign up, implement your own chatbot design and functionality (or plug into an existing chat tool that allows for third party integrations, like Slack), create data management and refresh capabilities for internal tools, CRM integrations for external tools… the list goes on.
Chances are, you’re looking to create your own custom bot so that you can implement features that other apps don’t have and that are specific to your business, so as much as possible you should focus on creation of those features and use other providers for everything else.
If you’re looking to leverage data retrieval functionality without having to build your own vector database or search indexes, or if you want to be able to invoke a chat tool within the context of the data sources you’ve already added, there are APIs available to you that you can use to power the backend of your custom chatbot while you focus on the frontend.
Chatbase is a new company, powered by LLMs, that provides you with the ability to plug in your external data sources and use their API to create, manage, and message your own custom chatbots. By using their backend API, you can send and receive messages programmatically and incorporate those messages into your own custom UX.
We also provide a developer-friendly API that’s designed to handle the hard work of storing data, finding relevant context, and sending contextual chat messages to an LLM. Our API makes it easy for developers to send messages without having to handle all of the infrastructure maintenance and management that comes with plugging in to business data, including authentication, authorization, and data refreshing.
Custom chatbots are a great way to automate tedious parts of your business, and as Generative AI gets smarter in the future, these chatbots will help you run things in a leaner and more efficient manner.
Chatbots of the future
As you implement new custom chatbots for different areas of your business, you’ll likely want to stick to chatbots that are powered by Gen AI, because they will provide added flexibility and functionality with minimal configuration.
An added bonus of using AI-powered chatbots is that these bots will begin to get very sophisticated, and will start to support additional capabilities, like the ability to execute commands on demand.
Imagine the power and efficiency of a bot that can look into your CRM, get the list of customers that need to be contacted today, send a personalized email to those customers on your behalf, and then automatically process responses from those customers and categorize them into low/medium/high urgency.
The world is changing quickly, and custom chatbots are going to be one of the main ways to interact with this new world going forward.
.png)
