Ometrics & Ochatbot
Founder & CEO
WEBINAR – GROWL Connex: How Chatbots Can Save Your Business in 2020
Libby Olson (00:02):
Good morning, everybody! Welcome to GROWL CONNEX. I’m Libby Olson, Principal and Founder of GROWL Agency. If you have any questions today during our conversation, please feel free to put those in the chat window if you’re watching us on Zoom, or if you’re watching us via Facebook live, put those in the comments. My guest today is one of my favorite marketers, and I’ve had the pleasure to know him for several years: Greg Ahern. Greg, why don’t you take a moment and introduce yourself and then we’ll dig in.
Greg Ahern (00:31):
Welcome everybody. My name is Greg Ahern. I’m Founder of Ometrics & Ochatbot . I’ve been in internet business since 1994. We specialize in conversion rate optimizations and AI chatbots. We have a suite of tools as well as doing consulting.
Libby Olson (00:47):
Awesome. So Greg, tell us a little bit more about chatbots. What is a chatbot?
Greg Ahern (00:54):
So a chatbot is a program where people can converse with the computer in a conversational way. It’s not live chat, which people get confused with. The live chat is a real person. It’s kind of like texting a real person. Although you can have a chat bot and you can switch to a live person if the user wants to talk to a human. A chatbot can review data really quickly so they can pull up the information that the person’s looking for and present it with a video or an image or text or sending a link, or a support ticket and chatbots can be accessed on websites and social sites like Facebook, Slack, you can even work on SMS for like texting on your phone, you can talk to a chat bot the whole time. They can be engaged by texting or pressing buttons or voice, you know, like Siri, that kind of thing.
Libby Olson (01:49):
So chatbots have grown a lot in popularity over the last couple of years. It’s rare that you go to a website and they don’t have some type of chat functionality. What do you think is driving this trend?
Greg Ahern (02:00):
You know, it’s really two parts of that. There’s the customer’s expectations and then the company’s bottom line. So from a customer side, you know, the society is becoming more and more used to using computers. I mean, there’s a whole generation now that grew up with computers. So they have certain expectations that we don’t older people don’t have. And the other piece is that everyone is more used to engaging on social systems and phone, um, like “Alexa” and “Siri” and things like that. So they’re very comfortable, engaging with the computer and talking to a computer. Whereas, you know, even five years ago, they weren’t as much, although these systems have been around since then. The other piece is the way that technology is changing, more and more people are using their phone to access information. Websites changed since 1990s, as far as navigation and finding information. And that way of doing things is really archaic, especially when there’s so much more information you need or want to look for, like, you know, thousands of products on an eCommerce site, for example. So instead of having this huge tree to find a pink sock, which is like in a site that sells everything from hammers to clothes it’s a lot easier to, just for a chatbot to say, I’m looking for socks and it really pops it up. And so this shift is happening now, and I think that’s part of what’s going on. On the business side, you know, companies want to service their clients and make them happy. The customers make them happy, and they also want to improve their bottom line. And so chatbots improve the revenue and they also increase leads, business leads and things like that. And so with happier customers and selling more products, companies want to invest in something that’s going to improve their revenue. And so that’s why they’re investing in chatbots and AI systems.
Libby Olson (04:04):
So to expand on that a little bit, you know, as marketers and business owners, there are opportunities to spend money in marketing are numerous, and obviously we all have limited resources and time and to get in front of customers. What are the benefits for businesses to invest their time and resources in a chatbot?
Greg Ahern (04:22):
Well, if you think about the sales cycle of any business, there’s certain times when you ask someone a question and we ask the prospect a question, and so chatbots can engage the customer at the right time in that sales journey that they’re going through. So for example, if you’re on a page and you’re on the homepage, the bot may ask a general kind of question, but if you’re on a particular page about a particular topic, then it can ask them more specific questions about that. Or, you know, for an eCommerce site, you try and check out, it might ask different types of questions. And so it can engage the, at the right time of the process timing as well. Like not right when they’re on it, but maybe a few minutes afterwards or seconds or whatever it is, or if they’ve clicked these many member pages, they know, “Oh, this person’s interested in this. Let’s ask him this question”, that kind of stuff. So it can detect where they are in that whole piece. You can also remember the conversation from the last time they’re on the website. And so then they can pick up the conversation from where they left off. It’s also good for lead funnels. So that’s more of an AI kind of system, whereas a simpler lead funnel type piece. You can ask particular questions and get more qualified leads, and get more leads that the company wants. And so the whole lead funnel system is very simple to set up with the chatbot.
Libby Olson (05:53):
So are there different types of chat functionality more than live or automated?
Greg Ahern (06:01):
Yeah, so even in the chatbot world is different types of bots. I mean, as I was just saying earlier that there’s a le bot, which is a very simple lead funnel type. I’m going to ask some questions and begin to answer those questions and that’s it. Then there’s “e-commerce bots”, which are going to pull data in and support bots do this too. They pull this data in and then you can ask particular questions and it will pull up that information quickly. And so, you know, for an e-commerce system, the effect is that you have so many products that can find exactly what you’re looking for quickly for same with support type things. And then you have bots that learn. So there machine learning is part of artificial intelligence or “AI”, and there’s different ways of doing that. So there’s a number of different types, basically, there’s your AI chatbot, you’re Levi e-commerce type bot your Support type chatbot. And so the back analog systems are similar, how you approach and set them up are different.
Libby Olson (07:06):
So how do I know where to start? If I feel like I need to connect with my customers better, where would I kind of start this process in the discovery?
Greg Ahern (07:19):
I think first you have to step back and figure out what the company goals are. You know, where’s the issue the company’s having? Is it people you have a very complex product and there’s lots of questions before they can decide to make a decision on which one to get? Is it selling the products is easy, but there’s often a couple of questions in the set up of the product or service that they have. And so that’s bogging down your whole onboarding of your product or service. So things like that can speed that up and then they can start there. Maybe they’re turning revenue for the monthly subscription, which he can’t kick off unless they set up the thing that you just sent them, you know, things like that. Or yeah, I mean, you have to look at that piece and then once you figure that out, then you can step back and say, “okay, this is what our customer is looking for, how are we going to get through that quickly?” We want them to know there’s a coupon. We want them to know or some deal, right? And then we want them to know that we offer these three types of things for the three types of use cases that we serve. And then we have to find out which use case they want right away, and then start providing the information they need. So that’s kind of one example. It really depends on when you get to backup and look at that piece and then building it out is kind of straight forward after that. Especially if you already have data, like we have all this call center data from a support tickets, or we have a huge back in system like that. Well, we have the database on Shopify and we just pull that right in, right away. We’re done in two minutes, you know, that kind of stuff. So then it’s just building it out and then once that’s done, then we have to train the bot to learn. And so that’s kind of the next step after that.
Libby Olson (09:02):
So how long does it take to create a chatbot and what’s required from the business? If ,you know, someone like us worked with Ometrics to create a bot on our site, what’s required for me as a business owner?
Greg Ahern (09:16):
You know, if it’s a simple bot like a le bot, then we just need to know what are those questions you want to qualify the lead and to draw them into, “Oh, that’s a good question. That’s exactly what I need here. I’ll give you my email address now, because I know you’re meeting my needs”. So that funnel building out that, that can take, you know, two minutes that that’s pretty straightforward. The, e-commerce and support type bots. Those can take, you know, a week to a couple of weeks, depending on how complex they are. The real key with building a more complex bot is to build the first part and then launch it. And then we watch it every day. And when there’s an issue with a fallback, we learn, “oh, wait, they’re asking for this thing. Well, they’re asking it this way”. Okay. We need to change the intent. So we understand that intent and answer that question question quickly. And that’s how it becomes smarter and smarter over time. There’s machine learning, and then there’s a human in the loop. So machine learning is when the computer does all by itself, and that’s quite dangerous. As far as society’s standpoint, there’s a bunch of failures in that realm. Human loop is where most systems work and the bigger systems, they have a combination of machine learning and human loop to make sure that everything is correct. You know?
Libby Olson (10:40):
And what do you mean by human in the loop? What does that look like?
Greg Ahern (10:43):
Human in the loop is we monitor the questions and fallbacks that are coming in every day. And then we correct a new way of saying, thank you, maybe “TY” now, instead of “thanks”. And so this new acronyms in our languages are always changing and those have to be updated, or it could be a new competitor came out with a new product and they’re saying, “Hey, what do you do with my, are you like Blue Jay?” We are like blue Jay? What does that to do with anything? Oh, it’s competitive product. We didn’t know about that. You know, we let the client know, and then we have an answer for Blue Jay. Yeah. But it doesn’t work like RSS because of it’s whatever. So there’s always stuff like that coming in. There’s also other things we see, like there is on the site, coupon codes, not working, stuff like that. But as far as the learning goes, it’s better to build the base part. And then in the next, that first month, the number of fallbacks drops significantly. And then it takes another two or three months to really get the nuances. And then after that you’re always monitoring this, but you pick up these slight insights, marketing insights, how the consumer is really thinking. I mean, one way to think about a chatbot in general is imagine if someone walked into the store, the physical store and they asked the question and the sales person came up and they introduced themselves, they maybe has a small chatter. And then they ask these questions to find out exactly what that person wants. That’s what the chatbot is doing. There’s lots of nuances in that kind of conversation, which we can think about and plan all day long. But the reality is you gotta send it live to see what’s really going on.
Libby Olson (12:23):
Yeah. Do you think consumers ask more questions via chatbot than if they were either with a sales representative or in a, you know, a form message or do you think you can gather more information from them or they’re more willing to give data, give information?
Greg Ahern (12:44):
Yeah, it’s ironic because I mean the latest tests, so like, you know, 63%, prefer to talk to a chatbot than a live chat. And you think about it, it’s like, well, they think it’s, they don’t think anything’s being copied or, and something personal is being shared, which of course and true, like collect all the data. Not that we would read all of it it’s too much, but it’s all there. And then there is no feeling of like, this guy’s going, gonna sell me. I’m not going to get pressured into a sale. I don’t want to talk to a live person right now. I just want to know, does it come in pink or whatever? And so that’s why people converse more with the bots and they do live chat. Now, in some cases like I can’t load your software. What’s, what’s the driver I need for it. I’m on this particular system. And, Oh, by the way, there’s this weird thing I’m on satellite. And so I have a lag of whatever and, you know, so I can get some stuff like that. Then it’s better to talk to a human and they’ll be fine. I mean, there’ll be more willing to talk to a human then the chatbot, but I’m most upper level funnel stuff, you know, learning about getting over the obstacles of a product or service, finding out what kind of product they want, they talk with a chatbot more.
Libby Olson (13:59):
Interesting. So how do we know if it worked? How, what are the typical ways that you measure success and how does a business measure success with a chatbot?
Greg Ahern (14:10):
So, most systems in our system included have dashboards that show your data that says, you know, how many people interacted, how many of those purchased, you know, how many fallbacks you have …that kind of thing. So we see a 15 to 35% lift in revenue for eCommerce sites when they engage with the chatbot, and they buy more, they spend more money when they engage with the chatbot. Other companies I’ve seen reports from other companies saying around 17%. So it’s a good number to take 15 to 30 that I view. On the business lead side, we see it five to 15% lift in leads, but, you know, these are tricky and it’s really the quality of leap. So measuring just, I got more leads, it could be junk. But, in general, they engage more and provide stuff, you know, compared to like a pop up, for example, there’s all these other nuances. Like, I had one client, they’re a bed and breakfast and so service type industry, and I was concerned that they weren’t getting increased in bookings and working less, but they weren’t getting increased. And so I talked to the client, and they were like, “Oh no, this is great. You saved me 110 phone calls last month.” Like, Oh, so she’s not measuring bookings. She’s meant by, is, am I getting interrupted with these silly little questions for about my BNB? You know, they have like 20 different BNB’s. So I mean, it all depends what your pain point is. B2B customers, one of our B2B customer that sells products, and, they get a lot of questions about zippers, like, Oh, something’s going on with our zippers. So they, you know, now talk about their zippers and how, what the warranty is, blah, blah, blah. That same one had a huge increase in people asking for a mail catalog. So, I mean, that’s kind of interesting that mail is so important for that particular business and client. And so that told them that, you know, direct mail can really help us here because they want to, they want something physical before they make the decision and we’re passing around something. So we want these insights you can learn. The CBD clients that the customers know nothing about CBD. So there’s a lot of education to learn. Like, no, it’s not going to fix this. It may fix that. We don’t know if it fixes this. This is the strengths you want, or you want to, what’s the difference between taking it as an oil or a gummy or putting it on your skin. All those things that most public doesn’t know about, this is a great way for them to educate them. There’s also a reduction of support tickets for some of our clients, what I mean by that is that it’s answering the simple questions so that the support team can spend time on the more complex ones and help out the clients. And that goes both ways that the employees are more happy to answer complex ones and say, “no, did you turn your monitor on” or things like that compared to, you know, more complexity than just about whatever’s being done. So that reduces the, monotony of their job and the consumer likes it, cause they’re immediately getting the answer they need. They can go off it instead of waiting for someone in live chat and everything else. So, you know, the insights we get that we repourt can vary from technical issues, like, you know, coupon’s not working, to more, you know, things you wouldn’t really notice about a customer looking for something that, Oh, we sell that, but gee, we don’t have it on the homepage. And then who knew that, you know, toilet paper and masks would be the big thing to sell right now, but everyone’s asking for it. So let’s get it on the homepage instead of being a very somewhere else in the site, so things like that.
Libby Olson (18:03):
Yeah, exactly. Well, I’m thinking back to that CBD example that you were sharing, you know, a lot of times, I think patrons don’t know what to ask in a product like that, where there’s so much education that needs to happen. So the prompting, I think can help educate in little bits without requiring the user to know what to ask or know what to look for on their website and can kind of learn in that and those little tidbits. That’s interesting!
Greg Ahern (18:29):
Yeah. And you would see, you can imagine, like someone walked into the store and the salesperson would immediately say, “Oh, you have this particular illness or you have this type of pain or you can’t sleep”. Whatever the issue is. And then narrow that down to the right product and educated to get over it. No, you’re not going to get high when you take this back, that stuff. And so now the bot can handle that.
Libby Olson (18:52):
Exactly. So with chatbots, growing in popularity and people digging into sites less and less and can kind of relying on chatbots to have that conversation, does this change an SEO strategy at all?
Greg Ahern (19:08):
It really, so it doesn’t change the SEO or could change. Let me put it this way: It doesn’t affect your SEO. If you have a chatbot on your site, you know, I can get ranked better, but, you’re going to gather information that you can use in your SEO strategy. For example, you’re going to see new keywords that are coming up. This could help, not SEO, but also SCM, you know, for paid advertising. Cause you’re going to start seeing longterm keywords coming in or phrases that you can use in advertising and in your organic positioning and you can track all those keywords and see what kind of questions are being asked. And that comes to the next piece where if they’re always asking a question about X, Y, Z, maybe you should write a blog on that and it’ll give you like content ideas to do certain things. And that’s what, that’s the other piece where it can help your strategies. You can get content ideas, you can understand the pain points of your clients and what they’re really looking for, which is one of the things Google ranks, because those questions are coming in and then you get all the data on keywords and key phrases.
Libby Olson (20:14):
Yeah, that’s awesome. As we’re closing in kind of at the end of our half hour, what are those key takeaways people need to think about if they’re thinking about chatbot technology.
Greg Ahern (20:28):
So, like I said earlier, you need to think about where the pain point is for your company as well as for the consumer. And then that’s the base of what you’re going jump. It’s kind of dysfunctional life is going to be. When you’re building the bot, think, start slowly, get something out there so you can see how it’s being used and then adjust, constantly adjusting as you build the bot and what it’s going to learn. You know, the key thing to do is don’t build a Siri or an Alexa only focus, very narrowly on what you’re offering. This will allow the bot to be very smart in “pens” and know nothing else about anything else. But when a person asks about a particular product, it is very deep and that will make it more efficient for both the consumer as well as is building it out. Because if you’re, if you want to make it so that it talks about the weather, the politics, you know, what kind of shoe was popular in 1970? You know, it’s, it’s endless game, you know, every time it’s finished.
Libby Olson (21:33):
Yeah, absolutely. And I know you have a really great offer for people who are trying, you know, want to learn more and discover more about chatbots. You want to talk about that a little bit?
Greg Ahern (21:43):
Sure! So we have a course that we’re launching soon, a chat bot course it’s free. So you can sign up for that at ometrics.com/chatbotcourse. And then when we launched that I will be here every other day. We’ll send out an email that will go over a different segment, everything from, you know, what kind of chat should I do to how to build it, what it needs me to think about and all that when you start getting too intense and replies and all the other stuff, and then we offer them free to le bot and free chatbots. So, you know, you’re welcome to come and play with them on our chatbot.com. And of course we are always here to help you.
Libby Olson (22:25):
Awesome. Well, thank you so much for your time today, Greg. Really appreciate it. We will put that chatbot course on our website, growlagency.com/connex/. So you can find that, we recording of this video or webinar on demand and thank you guys for joining us today!
Greg Ahern (22:44):
Thank you!