Well, whaddya know, It’s 2021, and in our technology-fueled lives, artificial intelligence or AI is poised to play a bigger and better role.
Just take this statistic from Zoominfo, which found that by the time 2025 comes around (so a little less than five years away), almost all interactions with customers will be driven by AI, 95 percent of these.
Even skipping ahead a few months to 2021, experts believe that customer management will involve AI about 85 percent of the time.
In those interactions, there’s never a need for another human to be present; it’s just customer and chatbot.
One major form of AI that is already implemented by companies across the world is chatbots.
That’s not such a bad idea considering data from Juniper Research found that, by 2022, chatbots will save companies a yearly total of eight billion dollars.
Your company may be interested in AI chatbot marketing or sales, but admittedly, you’re not quite sure where to begin.
At first, you might have thought chatbots as a trend that would die off, but now you realize you may need them to augment your customer service.
If you want to learn more about chatbots and AI, this is the article for you.
In it, we’ll explain more about the SMB marketing chatbot, the different types of bots, and how they can assist in achieving your sales and marketing goals.
By the time you’re done reading, you’ll be ready to shape your chatbot marketing campaign.
Before we can begin talking about implementing chatbots into your sales and marketing campaigns, we must begin with a chatbot definition.
A chatbot is a robotic chat function that is used for entertainment and assistance, with the latter more common in businesses.
When a customer has a query or problem, they’d visit your website and connect with the chatbot.
There is no other person on the other line, just the bot, meaning all communication is automated.
The customer types in their question or concern, and the chatbot processes this information.
How processing occurs depends on the chatbot in question.
Some look for keywords in the customer’s text, searching through their databases for information that includes that keyword.
They then answer the customer, providing useful information.
Other bots have a natural language processing system, something that we talked about on this blog before.
These language processing systems can derive intention from a customer’s words through a mix of AI, information engineering, computer science, and linguistics.
Some companies will make it look like you’re talking to a real person with their website chatbots, including a picture (perhaps a stock photo) of a customer service rep.
Customers can often see through this though when the replies don’t quite read as human.
In fact, for all their quick advancements over the years, chatbots are far from perfect.
They’re not up to par on the Turing test, which measures the level of intelligence of a machine or AI.
This doesn’t mean chatbot systems are useless to sales and marketing teams; far from it..
To know more about what is a chatbot, check out this video:
A chatbot is much like a normal application with a slight difference that it also needs to understand the intent of the user’s request.
It can’t comprehend the plan of the customer on its own.
A chatbot is trained to analyze and identify the user’s intent of their request and once the analysis is done, the right response is delivered to the user.
A chatbot is trained at a significantly faster and larger scale than a human.
A chatbot is nourished with a large number of conversation logs, and from there, the chatbot starts understanding what kind of questions need what type of answers.
There are three classification methods:
In this case, bots utilize pattern match to group the text and then deliver an appropriate response.
A bot is able to get the right answer in the related pattern.
However, it can’t go beyond the related patterns.
AIML (Artificial Intelligence Markup Language) is a standard structured model of these patterns.
NLU is nothing but the ability of the chatbot to understand a human.
It is the process of converting the text into structured data so that a machine can process and understand it.
It has three specific concepts.
NLP bots help convert the text or speech of the user into structured data for a machine to process it.
This data is then used to choose an appropriate answer or response.
NLP follows the following steps:
The NLP filters the set of words into pieces that are linguistically representative.
Here, the bot tries to interpret the user response that aligns with their emotions.
It checks and processes the text to find out any typo errors and common spelling mistakes that can alter the intended meaning of the query.
This program model of bot searches for different categories of words, similar to the user’s name, addresses, name of the product, whatever is necessary.
The bot looks for common phrases, objects, and nouns in the user’s text to discover related phrases that users want to convey.
EngageBay’s live chat software enables you to deliver personalized one-on-one support to every customer.
Now that you understand more about chatbots and how they work, let’s take a deeper look into the types of chatbots for SMB marketing chat and more.
These include standalone, messenger, voice-enabled, context-enabled, service/action, and quick reply/scripted chatbots.
Your first bot option is the standalone chatbot, which exists in its own separate app.
If your company develops a standalone chatbot, it’s often solely for the purpose of answering customer questions and addressing concerns.
Since this bot isn’t somewhere prominent, such as on a social media platform or perhaps even your website, it’s easy for leads and customers to not see the bot.
If your customers need information, they might call your customer service representatives since they didn’t even know your company had a chatbot feature.
Another deterrent to using a standalone bot is that since it’s an app, it’s going to take time to plan, test, and deploy, not to mention it will be costly to do so.
All that cash and effort might not be worth it if people don’t use the chatbot.
That’s not to say a standalone bot can’t ever have a role in your company, but it’s better to use a messenger app first and then switch over.
The reason for that is because messenger chatbots have become incredibly common in recent years.
These are the chatbots that exist in messaging programs and apps like Kik, Discord, Telegram, Slack, and the like. When your customers use these services or ones like them, they see the chatbot right away.
Should they need to reach out and ask a question, it now becomes incredibly easy and convenient for them to do so.
Messenger chatbots work well, too, so proves this 2018 chart from Statista.com.
In it, you can see that 15,000 people used WhatsApp’s chatbots in a month and 13,000 favored Facebook Messenger’s bot in that same timeframe.
That’s followed by WeChat with 1,000 users in a month, QQ Mobile with 763, Skype with 300, Viber with 260, Snapchat with 255, LINE with 203, and Telegram with 200 users.
Image courtesy of Statista
Next, we’ve got voice-enabled chatbots, with two of the most well-known and beloved examples Siri and Alexa.
Each time you ask either of these voice assistants a question or request that they schedule an appointment, play a song, or search something for you, you’re using a voice-enabled chatbot.
These bots even know your name, creating a sense of personalization that almost makes it feel like you’re talking to a real person, not a bot.
How voice-enabled chatbots work are as follows: you express your request verbally, Siri or Alexa processes it, and then they give you instant information or fulfill your request.
By relying on voice recognition APIs and text-to-speech services, the answers are mostly accurate.
Combining elements of artificial intelligence and machine learning, context-enabled chatbots may be some of the smartest around.
Sometimes referred to as contextual chatbots, context-enabled bots rely on context, as you probably guessed.
Not only do these bots have recollections of past conversations then, but they can take this information and use it to inform the data they provide you now and in the future.
In a way, it’s like these bots develop a memory, much as we people do.
As the context-enabled bot continues building these memories, it learns how to serve you better.
Google Assistant is a context-enabled bot, as are Alexa and Siri.
With a service/action chatbot, the bot needs contextual information (the service) to do something for you (the action).
These bots tend to have limited functionality, meaning they’re designed for a specific purpose.
For example, maybe you use an airline chatbot that tells you when flights are booked or canceled, how much a light costs, and if your flight is leaving on time.
You couldn’t ask a bot like this to schedule your haircut appointment or write down your grocery list, that’s true.
However, for what they do, service/action bots are very useful.
Finally, there are quick reply/scripted chatbots, which are like service/action bots in that they have limited interactivity.
Again, these bots are very good at what they do, but they’re not as all-encompassing as some of the other types of chatbots we’ve discussed thus far.
The bot might be the one asking you questions, sometimes in a multiple-choice format and other times open-ended.
You then fill in the responses and continue until the bot has gathered enough information.
Image courtesy of Daffodil
The above overview of bot types may have fueled your imagination, giving you several ideas of how to use a marketing automation chatbot system for your own company.
Once you integrate these bots into your company protocol, you’ll see that your business changes in a myriad of ways.
Here are all the uses of chatbots, with many of them beneficial to small businesses, mid-sized companies, and even bigger brands.
The most basic function of a chatbot is also among its most important.
That is, it’s supposed to provide information to customers and would-be customers that have curiosities or questions.
In today’s phone-averse society, chatting with a bot online is much more convenient than picking up the phone.
It’s also faster than email, in which you could wait hours, sometimes even days for a response.
Not every chatbot is all about work, work, work, as others are designed for fun.
These are likely not the types of bots you’d use for your company, but that would depend on your industry.
Rather than answering questions about a product’s pricing or where to find a company’s return policy, these bots are here to entertain.
They may ask the questions or the user can, but either way, the conversations between (wo)man and bot certainly make us people smile (or at least chuckle).
Whether you’re curious about how these bots work and want to learn through communicating or your goal is to trip the bot up by asking weird questions, you’re certainly entertained.
Alexa and Siri are often referred to as virtual or digital assistants because they can do a lot of the same things people in a receptionist role can.
For instance, you can use either of these context-enabled, voice-enabled bots to schedule something on your calendar, be that a meeting or an event.
You can even make appointments using bots.
Image courtesy of Daffodil
Some more sophisticated bots may even be able to tell you if you’re already busy for a calendar day so you don’t accidentally overbook yourself.
If your company could create a bot that implements elements of scheduling and appointment-making, that could majorly help organize your business.
Your sales representatives are very busy people, and even that feels like an understatement.
They may contend with hundreds of calls a day and thousands a week, perhaps even more than that! Anything that can lighten the load even a bit is helpful, and chatbots can do just that.
The efficiency and speed with which the bot answers questions means they can take care of more customer queries so the customer service reps don’t have to overload themselves.
This frees up the customer service rep’s time somewhat, allowing them to spend it on other tasks.
For example, if a customer wants advanced help and requires the rep to spend a lot of time on the phone (like 30 or 40 minutes), they can get the kind of thorough attention they deserve to resolve the problem.
The combined efforts of your sales reps were effective enough, but now you’re thinking that overall efficacy can increase even more with the addition of a chatbot.
As we just mentioned, these bots are fast, pulling information from a database in seconds or minutes when it would take us, people, far longer to do the same.
Even if you only used the bot for website queries and your customer service reps for phone calls, altogether, more customers are getting their questions answered.
If you’ve ever called a company, got put on hold for a half-hour, and then got hung up on, you probably felt dejected and quite angry, right?
That’s no way to increase customer satisfaction, and yet it happens (hopefully not on purpose).
Customers want a company to take their concerns seriously and provide instantaneous answers.
They don’t wish to wait days for an email response or hours for a social media messenger reply, as that’s too slow for their liking.
Chatbots provide the kind of instant satisfaction that has become a trademark of today’s society.
Customers get what they need right away and might tell their friends, family, and colleagues about their good experience, further increasing your clout.
Another way to improve customer satisfaction through bots is in the availability of chatbots. The average person doesn’t work all day and night, as that’s not smart nor healthy.
Instead, they’re in the office a set number of hours and then they come back the next day unless it’s a weekend.
If one of your customers has a question late at night or on a weekend, they used to have to wait because customer service reps only worked for so long each day.
Chatbots have no such restrictions, so customers can get their queries answered even if it’s at a weird time of day or night.
When we talked about how bots can schedule appointments and plan things on your calendar, those are two forms of automation.
As a part of AI, chatbots are almost entirely automated, as that’s one of their perks.
As we said, someone doesn’t necessarily have to oversee everything the bot says, but rather, they can trust it to answer questions accurately and adequately because it was programmed that way.
Automation is something we talk about a lot on this blog, and that’s because when it’s done right, it can be a major time-saver for businesses.
Instead of employing people to do smaller roles, technology like bots can do it for us.
This also reduces the rate of human error when automating workflows and other tasks, ensuring accuracy is always at the forefront.
As handy and helpful as chatbots are, the pendulum can swing the other way and they can hinder more than anything else.
If someone were to try to use a bot for malicious reasons, it wouldn’t be hard to do so.
One such example of that is an AI bot known as Tay, which was created for Twitter by Microsoft Corporation in 2016.
You may remember the Tay saga when it unfolded, or perhaps you missed it, but here’s what went down.
Tay was a context-enabled bot that got into the wrong hands on Twitter so it would say rude and vulgar things to other Twitter users.
The plug had to be pulled, so to speak, a mere 16 hours after Tay was introduced to the world.
The bot didn’t even last a full day, but it was soon followed by a similar bot named Zo.
Microsoft Corporation had an egg on its face after the failure of Tay, as the whole incident was both embarrassing and offensive.
Not only were people appalled by Tay’s behavior, but they wondered what programming oversights existed that could have made the bot so easily exploitable.
While Microsoft has since recovered and learned from its mistakes, we’re also talking about one of the biggest brands on the planet.
If your company had a similar incident befall you, it could severely undercut all your progress.
The main purpose of using customer service chatbots is to eliminate 90% workload of the customer service team.
The chatbot does this by solving customer queries through two actions — automation and deflection.
Automation and deflection apply to most chatbot use cases.
But the method of calculating its benefits is quite similar across all cases.
Here’s how it is calculated.
Deflection is when a customer service agent and their chatbot works during the same hours.
The chatbot initiates conversations with every customer and tries to solve their issues.
If it successfully solves their queries, the chatbot has deflected the customer away from human customer support reps.
But in case the chatbot wasn’t able to answer the customer’s query, it transfers the conversation over to the available customer support agent.
Here’s the formula:
Deflection Rate = (1 — No. of conversations transferred to agent/all chats within live chat opening hours) * 100
Automation is when the chatbot works without any human agents as a backup.
It is generally the case when the chatbot is providing service outside office working hours.
Here, the chatbots behave a little differently outside the office hours when there is no human agent back-up.
If the chatbot knows the answer, things are sorted.
But if it doesn’t, then it apologies to customers and offers them alternative options.
For instance, the chatbot will ask them to leave their details saying the agent will call them soon.
Or, it will open a ticket in the service ticketing system with the details of the issue.
This way, customers will be able to check back their status at a stated time.
There are two ways to calculate the automation rate.
The first is general automation rate where:
Automation rate = (1 — No. of unresolved conversations/Total no. of chat conversations) * 100
Then there is ‘Outside of Service Hours Automation Rate’
Automation rate = (1- No. of unresolved chats outside of office hours/all conversations outside of office hours) * 100
In the list of the above perks, nowhere did we mention a chatbot on a website to increase sales.
Is that something you can do, use an AI chatbot for sales?
Like a bot can aid your customer service representatives in answering customer questions, they can also work with your sales team to make the entire sales process more efficient.
Chatbot sales can also occur in several ways, such as being the first point of contact between your company and the lead.
By answering questions and providing the information the lead seeks, the lead may move through the sales funnel faster.
Your bots, as they communicate with more and more leads, can also generate data that could be of major use to the sales team.
A lead’s interests, needs, pain points, and challenges can all be gleaned from these chats, giving the salesperson a one-up when communicating with the lead by phone or through email.
You’ve read on this blog that the average sales funnel has at least four stages: awareness, interest, decision, and action.
Sales chatbots can play a role in each of these stages, from beginning to end, increasing the productivity of your sales team and even boosting your revenue.
We’ll now discuss how chatbot for sales can be applied for each of these sales funnel or sales cycle stages.
The awareness stage is first and entails the very beginning of the customer journey.
As we’ve written about on this blog before, during this first phase, your would-be customer knows next to nothing about your company.
They also have yet to make a purchasing decision, which may come later if it happens at all.
Your marketing and advertising methods are typically what attract these leads to your company.
The chatbots on a company’s social media profiles (such as Facebook Messenger) or on their website are something the lead is likely to come across as they do their initial research.
The bot can write a general greeting, such as asking the lead if they need help with anything.
It’s also possible to customize the bot’s dialogue so it mentions the way the lead found the company, such as through an ad, social media, or email marketing.
The lead may decide to take advantage of the chatbot, peppering it with questions to gain information.
After all, as we said, your lead knows pretty much nothing about your company, and they may be curious to learn more.
When the lead has a good experience with the bot, they could feel inclined to move onto the second stage of the sales funnel, or interest.
This is also the first form of engagement your company may have with a would-be customer, chatting via bot.
The bot can address each lead that may arrive at your social media or website, helping them learn what they want to know in the early days.
It does so tirelessly, perhaps going through dozens of leads, and maybe even hundreds a day or week.
If you as a human tried to do the same, it would take you days or longer.
By the time you finished with one group of leads, the next would have trickled in, creating a never-ending job.
The lead, if they engaged with your chatbot during the awareness stage, likely only asked the most cursory of questions.
Perhaps they wondered when your company was founded, who runs it today, or what you sell.
The positive experience of having interacted with your chatbot may bring the lead back for stage two, or the interest stage.
Now the questions the lead would ask would be more specific, such as the pricing of a certain product or how much shipping costs.
This would also be the time that the lead may decide to opt in to an email newsletter or reach out for a consultation with a salesperson by phone.
Like your bot provided the most basic information before, it can also dig through its database and answer more complex questions now.
Should the lead ask a question that even exceeds the chatbot’s knowledge, then the bot can provide a phone number or email address so the lead can get in touch.
Now we’re getting into the advanced part of the sales funnel or the decision stage.
This is stage three of four, so the lead is pretty serious about your company and its products or services by this point.
They still have yet to buy anything, but they’re gearing up to make a purchasing decision sooner than later.
This is typically when the marketing team does most of its work, enticing the lead with information and promotional materials that will hopefully convince them to buy.
You may wonder if chatbots can play a role this late into the sales funnel, and they absolutely can.
First, let’s talk about the chatbot’s duties from the marketing side, then the sales side.
By now, you have to anticipate that your chatbot has interacted with hundreds or thousands of leads.
In logging all that data, common threads can emerge that your marketing team can use.
For instance, your marketers can identify common problems and challenges a lead may have and then use these to create pain points for various customer avatars.
Now the marketers can segment the leads and target the most qualified ones with personalized, tailored materials that hit home.
While chatbots don’t work directly in this regard, the past duties they’ve done and are continuing to do for new leads proves invaluable.
As for the sales team, bots can work in conjunction with your sales team, almost as if you hired on several more sales reps (but without the extra expense of doing so).
Your chatbot can pass along relevant content to leads, continue to address queries, and fill in the gaps salespeople can’t because they’re well, human.
Like with your marketing team, you can rely on the log of conversations the bot has had with other leads to target and personalize the sales approach from there.
Now we enter the fourth and last sales funnel stage, which is all about taking action.
The lead will finally buy something from you and convert from a lead to a customer.
Admittedly, chatbots don’t always do much for this last stage, as people are needed to draw up contracts, have phone calls, or sell a product in-person.
Bots can provide information about other relevant products, ask questions to the customer about what the buying experience was like, and even reach out to the customer sometime later about a reevaluation.
If you want to design your own chatbot for sales, it’s important to keep these points in mind:
We’ve gone quite in-depth regarding bots for sales, but what about chatbots and marketing?
Whether you want a chatbot for automated marketing or a B2B marketing chatbot, it’s possible to use AI to supercharge your marketing efforts in 2021.
The following methods for marketing automation chatbot use are sure to get you on the right track.
This is something we just touched on, how in the initial stages of a lead entering your sales funnel, your chatbot could be the first form of communication they have with your company.
It’s much the same in chatbot marketing, in that this conversation could pave the way for further engagement.
It’s a lot less daunting speaking to a bot than it is to a person.
With sales or marketing teams, a lead may assume they’re going to be sold to, while with a bot, they don’t necessarily have the same preconceived notions.
It’s even better if your bot initiates the first conversation, so this way your lead doesn’t have to take the plunge.
After all, this can be scary for some leads too, and so they may never get around to doing it.
If you use popular messaging platforms like Facebook Messenger, Slack, and others, you don’t have to start from the ground up if you want to add a chatbot.
Instead, it’s often as easy as clicking or selecting a setting that activates the bot going forward.
Some services, such as Slack, have more than one type of chatbot, including the GrowthBot (for CRM, sales, and marketing), the BirthdayBot, and the busybot that keeps all team members occupied.
Image courtesy of Neil Patel
It’s no secret that personalization is a key component of hooking in customers and keeping them engaged.
Chatbots can be personalized as well, addressing leads and customers by name or mentioning how the lead found the company.
While some people can ignore bots with a generalized, canned greeting, it’s a little harder to pass off a chatbot that’s talking to you by name.
If your company has a diverse team of marketers, they may not always be in the same place at the same time.
Some may be out-of-office, attending events or networking while others are hard at work crafting new marketing materials.
You don’t want to have overlaps where two people are doing the same thing, as this wastes everyone’s time, but how can chatbots help with that?
We’re glad you asked, as chatbots can absolutely lead to a more organized, productive marketing team.
You can try a bot that tracks productivity and activities so even if marketers who are working remotely never feel out of the loop.
While AI like chatbots can be intimidating to some, they’re going to be the future of business. It’s already started, so it’s a good idea to hop on this trend bandwagon now.
After all, chatbots can greatly assist your company in both sales and chatbot marketing automation.
With the information we provided in this article, you can now map out how chatbots can work for you and make 2021 your most productive, successful year yet!