What exactly are chatbots?

A chatbot is programmed for people to conversationally engage with the computer. People mistook it as live chat, wherein you talk with a real person. Users who have a chatbot but want to talk to a human can still switch to livechat. Chatbots can quickly review data and gather information about what a person is looking up in a website, they present outputs with an image, video, message or even by sending a link. These chatbots can be found on web and social sites such as Slack and Facebook.  It also works on SMS, you can talk to a chatbot by just texting from your phone. Chatbots “Siri” and “Alexa” are examples that they can also be accessed through typing, clicking or via voice-recognition.

Factors that drive chatbots into popularity

There are two main factors, first the customer’s expectations and then there’s the company’s bottom line. Firstly for customers, people are getting more used to being computer-aided through the years. We are approaching a technologically advanced generation, so there are people who grew up with computers and they have certain expectations which the older generation do not have. Compared to five years ago, almost everyone nowadays already had phones and comfortably use them to socially engage with the internet. Another thing is the way technology is changing, websites have changed since the 1990s, from navigation to finding information. Things were done archaically back then, especially when you want to search for one thing but there are thousands of products to look through on an eCommerce site. Now, instead of having to find a pink sock out of a huge pile of laundry, like in a site that sells almost everything it’s a lot easier to tell a chatbot what you are looking for, it will then filter every product in its site and give you results of the product you’re looking for. This shift from archaic to advanced is currently happening. It’s one the factors why chatbots continue in the trend. In using them for business, companies want to give the service that their customers want to receive, it is what improves their respective bottom lines. The chatbots in their website help with the company’s income and business leads. Companies invest in AI systems and chatbots which lead them good feedback and improved revenue.

What benefits businesses that invest in a chatbot?

In any business there’s a sales cycle. This cycle happens when you ask a prospect a question, but chatbots can do this job for you wherein they engage with the customer at the right time in their sales journey. For instance, when you visit the homepage of a website, the chatbot may ask a general question. Whereas a chatbot may ask specific questions if you’re on a particular page with a topic, the same thing goes for eCommerce sites which might ask different questions. There’s also a situation where a person has clicked on the same page several times, the chatbot may think that he’s interested in things related to it and so they’ll ask him questions about the page. With this, chatbots process their timing and can pick up from where they left off in conversations on the website as well. It’s more of an AI kind of system which is good to set up for lead funnels. Asking particular questions leads to more qualified leads which a company wants.

What are the different chat functions other than live or automated ones?

There’s also different types of bots in the chatbot world. As mentioned above, there’s the lead bot which is a simple lead funnel type, mainly just asking questions and the person just answers it. Another type are the e-commerce bots that pull data in. This is how support bots work. They will pull up information by asking particular questions and will quickly find what you’re looking for among many products in an e-commerce system. Then there are also bots that learn. Their machine learning is a part of artificial intelligence or AI, hence the name AI chatbots. Basically, there are different types, the AI chatbot, the leadbot, the e-commerce type bot and also the support type bot.

How will I know where to start, to connect with my customers better?

First figure what your company’s goals are, what kind of issue the company is having. For example, you have a complex product which results in difficulties in deciding which one to get. Another one is that selling the products is easy but the product or service that the company has is questionable. So this affects and bogs down the launching of the product and its service. Once you figure those things out, step back, think about what the customers want and get some answers. We want them to know that there’s some deal, that we can offer the service that they want. It depends on the back up you get and how you build it up, which is kind of straight forward. Especially if you already have data. In Ometrics, we get all call center data from support tickets. We also have the database on Shopify that’s quickly pulled in which can be done in two minutes. Then after building it out, the bot has to be trained to learn.

What requires a business to create a chatbot and how long does it take?

For a simple bot such as the leadbot, it only needs to know the kind of questions you want to qualify leads and draw them in. This funnel building can only take you two minutes, while the support type and ecommerce bots, depending on their complexity, will take you a week to a couple of weeks or so. More complex bots are built with its first part developed and launched. We look after it every day, and when an issue arises, we learn that we have to change the intent, understand it and quickly answer the problem. Over time, it becomes smarter. Another one is machine learning, where a computer is independent and does everything by itself, which is somewhat risky and has failures in that area. Now, the human in the loop is where most systems work. Bigger systems combine machine learning and human in the loop to ensure that everything is correct.

Human in the loop? What is it and what does it look like?

This is when we monitor questions and fallbacks that come every day. New acronyms in our language are always changing and it needs to be updated, say for example “TY” as a new way of saying thank you. There are also times when coupon codes in a site are not working. It is better to build the base part because when fallbacks drop significantly, it takes two to three months before you can really get the subtle differences. Then you’ll always be monitoring this, and while you’re at it you encounter marketing insights, insights that are from consumers. One way chatbots are generally viewed is that they’re like a salesperson in a physical store who can pick up small chatter and eventually ask questions to find out what the customer wants. The only difference is that they’re programmed and all the chatter is done digitally, but reality is that you have to send the questions live to know what’s happening.

Are consumers more comfortable with giving their information to chatbots or a live person?

The latest tests show that 63% of customers prefer talking to a chatbot than a live chat. This is because they don’t feel that any of their personal information is shared if it’s a chatbot they’re talking to, which is true. They don’t get pressured into buying the product unlike when it’s a live person that’s talking them through sales, then there’s a feeling of guilt if you don’t buy their product, because sometimes, as a buyer, you just want to ask inquiries before making your final decision whether to buy it or not. In some cases which chatbots can’t fully handle such as conflicts with loading a software or a lagging problem, it is better to talk to a human. They’ll be able to respond to the problem more quickly than a chatbot would, especially when the chatbot doesn’t know anything about the problem. 

How does a business measure success with a chatbot?

Most systems include dashboards that show data of how many fallbacks a company has, the number of people who interacted and purchased. For eCommerse sites, the 15-35% lift in revenue that we see reflects that people spend more money when they engage with the chatbot. Other company’s reports have around 17%. 15% – 30% is a good number to take in view. Now for business leads, we’ll see 5% – 15% lift in leads, but these are quite tricky, because more leads could actually mean more junk. But generally, they engage and provide more stuff compared to pop ups. For example, CBD clients have customers that know nothing about CBD. There’s a lot of questions coming in. This could be put about product strength and also as an opportunity to educate about what the public doesn’t know about CBD. Answering simple questions will help the support team to focus on more complex ones and also help out clients at the same time. This is a win-win situation since employees are happier to take care of complex ones because it reduces the monotony of their job, and at the same time the consumer also likes it since they’re being immediately attended to instead of waiting for someone in the live chat. So the insights we get can vary from technical issues, for example coupons not working, to unattended things, like if the customer is looking for something that the company has but it’s not posted on their page. Another example: knowing what product is in demand, such as how toilet paper and masks are right now. With that information the company could put products related to it on their homepage so that more customers will see it and eventually buy it.

Does the reliance of people on chatbots affect and change an SEO strategy?

Having a chatbot on your site can get you ranked better, but you have to gather information that you can use for your SEO strategy. For instance, new keywords could help not just SEO, but also SCM for paid advertising because you’ll start seeing long keywords or phrases coming in which you can use in your organic positioning and advertising. You can also track those keywords to see what kind of questions people have. Now, if they’re always asking about “xyz”, it could give you blog or content ideas. So no, it doesn’t affect your SEO but can help your strategies instead.

What are the key takeaways people need if they’re taking chatbot technology into consideration?

You have to take your company’s pain point as well as the consumer’s into consideration. It’ll serve as your starting point. When building the bot, think, it is okay to start slowly to have room for adjustments. Go and find out how it’s being used and what it can learn. Don’t try to build a Siri or an Alexa type of bot, only focus on what you can offer, because eventually it’ll be very smart in nothing else but “pens”. So that when a person asks it about a specific product, it could go deep which is efficient for both the consumer and itself since it’s also building out more.

To know more about chatbots, sign up for our free chatbot course which will be launched soon. Once it’s launched I will be here every other day. We also offer free lead bot and chatbots, you’re welcome to play with them by signing up for our free plan.

Follow Me

Greg Ahern

Greg Ahern Founder and President of Ometrics® is a fanatic about conversion rate optimization, AI chatbots and lead generation. Greg has been a successful Internet entrepreneur since 1994. He speaks at conferences and webinars and has built a number of internet businesses, including web marketing, web development and internet lead generation, which have been successfully acquired. Greg is the Denver Chapter Leader for the Digital Analytics Association. You can follow Greg on Twitter @gregahern and join his CRO Hacks Groups on Facebook and Slack.
Greg Ahern
Follow Me

Latest posts by Greg Ahern (see all)