Using Intent to Engage Your Visitors

Reading Time: 18 minutes

Guest: Eliav CohenSeasoned executive with experience running large organizations.

Overview: This week on digital conversations our guest is Eliav Cohen of Ultra Cool. He teaches how to get started using keywords for Drift and 3 key points to getting the most value from chat.

Reference:

Billy Bateman: Alright everyone, welcome to digital conversations. This week, I’ve got Eliav Cohen, the founder of ultra-cool joining me. Eliav, thanks for joining the show.

Eliav Cohen: Yeah thanks for having me.

Billy: So, we’re going to talk about a lot of cool things that you’re doing with drift but before we get into that just take a minute and tell us a little bit about yourself and ultra-cool.

Eliav: So, I got into drift for a very specific reason, after having 20 years in sales, running my own companies, helping billion-dollar startups with their sales teams, I decided I was going to start a hot air balloon company because I’d always been into hot air ballooning- that I got into when I was just out of college. As the hot air balloon company started moving up on Google, I was using Chatra back in the day, (that was like live chat which was big in 2015) and what happened was I started getting so many chats that I literally, couldn’t get through books with my kids.

I get too many phone calls and chats and most people would say well that’s a good thing, right. Google, like you’re getting lots of people to your site wanting flights, but I was finding it  8 o’clock at night I’d be sitting there reading a book with my three year old and I couldn’t get through a book. I felt like well I have to answer the chat or I have to answer the call because if I don’t, I don’t make anywhere from 400 to 1,200 dollars.

And so what the biggest challenge was is that there was I decide right, do I answer it or do I read a book and be a good dad. That’s why I looked up automation of chat and Found drift back in 2017 and it really came out of me just needing to build drift for myself so that my bot could answer any questions someone could ask right at those, that 80/20 rule those top 80% of questions and then have some look without having to talk to me, and eventually and we’ll talk about that later I ended up crushing that and then eventually started doing it for the balloon companies and tour companies and venture I was like  what the real models B2B, I did sales four years and running sales teams and operations let’s just building for B2B and then that’s kind of where we went then this last year we want to partner the or drift, so now we mostly build them for from small organizations all the way up to fortune 5,000 companies now for both b2c and B2B

Eliav: Awesome! So, and I will get into your bot sleeve bilks they’re pretty impressive but before we get into that I want to ask you, so if we’re going to look up Eliav on social, Facebook, LinkedIn, Instagram and try to figure out who you are, what’s one thing that we would not know about you and be able to figure out?

Billy: You would not know that when I was 17-years old I was a professional street performer doing juggling and magic and there’s probably one photo still on social from many years ago where I am able to bend silverware with my mind

Billy: Wow!

Eliav: And now is what I became no. for was basically having one point I had about 50-dozen Forks in my car that were bent and my girlfriend at the time was like why do you have all these pieces of silver bent all over your car and I was like just trying to learn how to bend over with my mind.

Billy: That’s awesome! So, where were you doing that?

Eliav: I was in Boulder Colorado on Pearl Street Mall, it was funny was you most people in high school they get mentors and most people pick like lawyers doctors yeah one who’s successful and since I was the age of four I saw this guy named Johnny Fox was a sword swallower and magician and this guy David who knew every zip code in the world. I asked them to be my mentors so David the zip code guy became my mentor and taught me a whole bunch of bad jokes, how to do crowd control how to really and the performance side of the tricks and how to make decent amount of money pretty quickly from random people in the middle of the store so…

Billy: Awesome, awesome man, that was probably the last thing I expected so I love it. Hey, telling bad jokes to employees in a crowd it sets you up for success. I worked as a river guide when I was in high school in college and that’s it’s the same jokes over and over again but it’s awesome, so well let’s get into drift. So, interestingly enough before we connected I had actually found your bot you built for Seattle air ballooning and brought it to my team and showed it to him a few months ago and said like this is amazing, we need to strive to like build Bots as good as this and especially interested because  most people use drift for B2B and this is a b2c used case.

So, I mean break it down how this bot is one of the most complex bots on drift and like how did you start building it out what would what was your process for building this?

Eliav: Yeah so this was actually pretty funny because I remember  getting drift and being really excited about building my first bot and I set it up and I use and I crate some keywords and I literally put it out on Facebook, right and I say guys I built this chatbot that answers every question, go test it and it failed every single question anyone asked, and I was like damn it so I went back to the drawing board and I added more keywords and I added different buttons and different pieces because I really wanted it, because what I found was that people ask the same questions over and over again and I would spend 20-minutes on the phone with them and really for someone to buy a product or even book a demo right really they just have to get one or two questions answered, if you can answer those one or two questions they’ll either a book a demo or they’ll buy a product right

Billy: Totally

Eliav: The problem was there was a lot of questions people were asking, like how high do you go, can my four-year-old come, what if we don’t drink alcohol, I’m handicapped can I come, my mom a hundred years old is she going to be able to get in the basket- and some as I started answering those questions I put it together again and I spend another like week two weeks building this output on Facebook again and I go guys test out my amazing bot and again every question people asked it pretty much failed, it gave the wrong answer.

Now, here is the funniest answer it gave. So, someone asked “can I have sex in the balloon?” as a test and it literally said, “yes, do you want to share her private flight?” I was like nooo and this is where I learned about keywords. I started to understand what is the difference between of white keyword fires off a certain answer, so I started looking there’s contains which means it contains that word there similar means that word is similar to another word and you’ve exact phrases right and so anyway so it’s funny for that one killer sex in the balloon, it literally looked at it was looking at the word balloon or can and instead of sex as the actual keyword and so in the end I was I sent it to drift as like hey drift fail, I thought it’s hilarious

And then I ended up changing it so now someone uses that word or the smoke or  weed or something like that’s an odd question ask, it then says you probably want to book your private flight JK, I think we all know we’re not that kind of comfort but again you if you’re answering questions you’re trying to handle questions that are weird or rare and be cheeky about it, you to make sure it gives the right answer and that comes back to even a general question someone to ask is if let’s say someone asks the question to your bot and it gives the wrong answer, well what does that experience like to the prospect, you basically say like you guys suck, you’re giving you the wrong answer I no longer of trust and I believe that your bot is your entire user experience, as soon as you put a bot on your site

And so if you can answer the right question you win and give the right answer if you and you give the wrong answer it’s confusing, that almost puts you off in the wrong foot with a prospect, that’s he got to be really careful with keywords to do it properly as you launch it so anyways as the story continued eventually  I continue to build key words out and this was in the days before they had keyword groups, you couldn’t name them so just had number one through like 40 and I had to do each question and I spent I don’t know, how many thousands of hours but in the end now this bot has over 2,500 keywords to answer pretty much any question anyone asked.

So, this point it breaks maybe once or twice per week where someone asks a question it can’t answer, it’s actually about ballooning. You asked about catering remodels, it’ll answer question right so we still bar back up there but the idea is in structures, A. can answer any question and then B. when someone asks the question I looked at what is the intent behind the question of if I answer it what should happen next because if you’re just answering questions over and over again that’s like an FAQ page who cares, you’re not moving them forward and so it’s how do you move them forward from that question based on what they’re asking, should you then go to try and book community should you go and try to book in with a schedule writer ask another closing question or do you just go hey what other questions do you have and that’s the kind of way that it was designed using intent, so it’s kind of like the early case of drift AI but using both sales and tent based questions with choices of two positives at the right time smooth someone forward in the sales conversation

Billy: Yeah,  I mean this but it’s awesome, you’re right, I doubt it breaks very often unless you just totally go out go off-topic but if somebody is like building those keyword Bots and drift it’s a little bit daunting if you haven’t done it before and what advice would you have like if I’m starting to build I’ve got drift or an intercom or some bot platform, I’m going to build a keyword-based bot rather than then use predefined answers and just buttons that they go down, what’s your advice to them?

Eliav: The best advice that I have for some star teeing keywords and I’m just doing this right now for a site it’s called www.visitrainier.com and visit Rainier is this big like four people are visiting Mount Rainier right and we, although we have a lot of buttons that go through what type of, do they want accommodation are they looking for activities and this has over 200-branches, doesn’t mean or the biggest bot I’ve ever built 200-branches right let’s sort of just the buttons and at one point we have them where they can do open-ended question about sling specific about Mount Rainier National Park and so what we did was instead of we created some initial keywords that we knew the answers and then basically we’re in testing phase, so if it gives the wrong answer it then goes to a live person, where if it doesn’t have the right keyword it goes to a live person and gives them that opportunity and then what you use is you look through those and go is this a common question we can answer?

Okay, like one that we can send them to another page on the site, is it the same answer or is it one that changes over time that’s going to need to be updated or is it one that like no one else would ever ask this question? And you start to then slowly add keywords in and one question because if you try to do it in every single question it’s not possible so you have to go through a certain amount of users to find out the types of keywords they’re using and you can use out even live search and drifts I’m not sure with intercom but you can put in like and see what, if that keyword is used many-many times or how many times then go is this a question we should answer and then you also look at well if someone asked a question and gave the wrong answer but should have given the right one, you start to look at keywords

So, let me give you an example I care about which side I was doing before but oh I think it was actually for Seattle ballooning and someone asked one time, can my dog come. Now I had in a category with cat, dog, animal, like in a couple other keywords someone said the word category in in a in a question that they asked because I was using the word contains for cats, it thought cat was the same as category give the answer about animal, makes sense?

Billy: Yeah, yeah.

Eliav: There was another one where — so you start to look at, there’s another one on a visit Maori near this last week, I’m visiting here where it said someone said, asked a question about what is the best area to go snowshoeing, now one of the keywords I had was around what should people wear? It’s a common question like what closely wear and one of the keywords was hot melt hot is as a as one of those keywords the problem is hot is part of the word what so I was like oh why did it give that answer and I wouldn’t deleted the word hot and change that to a similar keyword because it didn’t contain age 18 in a word, does that make sense?

She start to really go; how do ends up getting the right answers is really challenging but I’d say the most challenging keyword I’ve ever had to deal with is the word time and this is with Seattle building now, most of not all  companies going to deal with time but let’s say it’s like how much time does it take to build drift or how long what’s the timeline right, on getting swing complete like good master on time those all ended up having to be exact keyword matches because it was the only way to answer that question properly, like what time should we meet there how far ahead of time should we book.

So, I have over 248-exact match phrases just for the word time because I need to give the correct answer or you look at like; how do we create a more general answer that encompasses all the words about time as an answer? So, you start to look at strategy around and if we answer this correctly what happens next, does that makes sense? So, I know it’s challenging the keyword

Billy: Yeah keywords, I mean we’ve done keyword work it is it is not easy, it’s not and  the more complex especially with anything b2c facing, it’s just like it seems like you got a deal with a lot more things  somebody comes to buy a SAS solution for something, they’re pretty like down to business generally so a little easier but one of the things I’ve found and tell me if you what you found is with anything keywords it’s just getting people through the bots, like you have your best guest when you first build it but you just got to get people through that and then just start working.

Eliav: Correct, it’s all just about failing and  I think one of the my favorite quotes was someone talking about you just have to fail over and over again but don’t fail in the big ways, make the smallest amount of small failures you can learn from then quickly and is the same thing with ballooning, like you want to learn a lot and you’re only going to learn by failing but don’t make a failure where you hurt yourself, like and so I think failures are really positive things with bots because it’s the only way that the bit you can learn and even looking at  putting your Bots out there for people to test them, putting it out for your customers and it’s setting up that bot in a way that when it does fail it doesn’t give a terrible customer experience, I mean that’s I think the bigger challenge is setting you up properly so you can learn but not creating that for experience and a UI standpoint

Billy: Totally, totally, awesome. I wanted to dig into one other thing with you well I’ve got you so you you’re good you guys are an amazing drift partner and I wanted to ask you okay, if somebody gets drift what do you think are the best  one to two use cases for a B2B business that’s using drift, what strategies should they employ to really get the return on their investment?

Eliav: I would say that there are in twofold. I would say number one is drift has laid out the entire plan of how to actually use drift effectively and one of the problems and challenges I believe with people trying to implement drift is that they look at it only as lead-gen from someone who is going to be interacting with a bot that fires on a page and if you only do, like if you’re still going to keep forms on your site like Marketo forms, HubSpot forms, any type of contact form you’re missing out on it on the opportunity to be able to actually have drift do what it does best which is converting people that are interested and ready into an immediate demo, better than any other way and so I would say that that would be number one is making sure that you connect all the places on your site where you can completely get rid of forms.

The number two thing I would say is making sure that the bot is something that that opener actually provides something that they’re interested in and then be the bot gives actual value because there’s nothing worse than having a bot that just basically is a form and so if you look at the bots that we built or if you’re old school that I owe and you go through our bot explains about, like how, we give a lot of like educational videos about you acts or about your brand and why you’d want to do that we’re really showcasing how you would best to use drift right or how you’d use it with drift video or conversational content.

And so, I think making sure your body has some real value and can teach them something or get them to other places on your site, that are actually entrant helping them what they need it’s not always that when someone clicks on it that they’re ready to chat, and so I think that’s about you so when I look at Seattle ballooning I pulled up our stats and our that homepage bot has a 29% engagement rate, that’s the highest I’ve ever seen but it’s because it’s so damn contextual,  saying like we had an epic drone video on last week want to check it out, it’s epic, like if you were to put an opener on any site that said want to see something fascinating, probably people are going to click on it, now you better back it up with something fascinating right?

Billy: Yea, yea

Eliav: You’re going to say it you got to follow it up, you can’t, it’s like clickbait doesn’t always work right no so that’s the what that would be the second thing and then I think the last thing would be to be able to connect all of the different places where someone is coming to your site within the bot you’ve built, so whether that’s coming from an outbound email or average Taiyo or wherever they’re connected and connecting all those things is challenging for one reason, the challenging part about it is other people own those specific parts and don’t always want to use drift and so being able to get everyone in and that’s sometimes helpful using your partner signs it’s helpful  to be able to have those other conversations of why it’s important and I find this with a lot of these large companies,  we were talking about earlier you and I for this about  drift AI and three other pieces you can implement, if you haven’t built the core and got rid of your forms 100% you shouldn’t be doing anything cool drift yet, don’t you get the core then you get to do the fun stuff like for webinars or for different  using it in ad campaigns and the other piece but you really need to build out that core bot to sling that actually works and is powerful in is connect all your MTO’s to STO cause all those pieces are so important and the operation is perfect before you start trying to do the creative stuff because otherwise you’re not going to get the reason, you don’t still get good results but they’re not can be epic results, I think that’s what you and I try to do is how do we have someone go from okay results to epic results

Billy: Totally

Eliav: And sometimes that requires redoing a lot of the basics before we can do anything fun

Billy: Agree man

Eliav: Sexy, right?

Billy: I couldn’t agree with you more, you’ve got to use drift first to like take that low-hanging fruit to engage them and get them booked for or demos or whatever you do, before you go and do all of the fun really creative stuff and get organizational buy-in, so they can already see like oh  you took drift from booking ten meetings to fifty meetings a month for us how can I leverage you in my ABM campaigns or email marketing for my display ads totally

Eliav: I think that’s one of the biggest challenge most people have when building drift, is that I’m sure you hear clients say this a lot too but I have clients at the beginning they’ll say things like oh we just want to walk before we run, we just want to like let’s just like one piece and I think that’s the wrong way to go with drift and or with intercom or any of these things, hit dip your toe in this swimming pool and expect to have a lot of fun  same amount of fun is jumping full in and swimming and enjoying yourself and I think that’s where you can’t, there’s no way to test like is this working better than a forum then literally going flat and testing it and seeing the impact now you and I both know that it’ll work better and a lot of people are concerned there and go will this work better will we be able to use as soon as our SQL’s we’ll be able to put that into all of our operations on our keto and our smart list all these different things and I think that’s what stops people from being effective with drift initially and that’s a lot of the conversation we have to have with people is like okay like if you’re going to do this you have to do it full-out otherwise you’re you’ll figure results but it’s not going to be like mind-blowing

Billy: Yeah, you’ll get some results but unless you really commit, like once you commit you’ll see it is worth it and it works with all of your other systems and most of the other bought platforms,  what if you’re not using drift they’ll get you similar results if you do it the right way.

Eliav: I agree with you too, I mean there’s a lot of folks that are on  intercom and HubSpot and all these  different chat ones, I think one of the one of the things that I like about drift I’ve looked a lot of platforms there are ones that I do like you I better write about the ones that if they’ve that’s been there sole focus. I think being able to segment customer groups to give specific conversations and then being able to use conversational content of turning any PDF into a conversation and using drift video and combining the three of those, that’s what has me used drift over some of the other platforms but there are other people that  use intercom for different pieces also, they’re just wildly different products and I think it depends on what you’re trying to use it for but I I believe that  drift ends up being one of those almost like Salesforce first HubSpot, they both work but you can build anything in Salesforce and it takes more time and it’s more challenging but you can still build a basic, even with drift you can have three basic bot, like you can have a homepage bot, pricing in a second net bot that just on the forum says hey, want to skip the forum, you’ll still do better than not having a bot no matter what’s  platform you’re using

So, when you start to get into like how do I integrate this into segmenting users with Marketo versus hard odd versus connecting all your systems, that’s where you’re almost forced to have a little more complicated system be a little more strategic in the way you’re using it and building it and that’s where helps to have some someone who’s an expert

Billy: Yeah and I agree with you I think at the moment drift is the most complete option for sales and marketing using a bot but the other ones they work pretty well depending on what you’re trying to do with them so.

Eliav: Yeah, I think that there’s a lot of— the only ones that I think are not effective are ones where it always asks for an email first before information and I really like it I got to come explain it’s the same way with like gating all your content  like some people just want to ask a question before giving you information, so you’ll just want to read an article doing research before having you send them a ton of emails and so I think that there’s a lot and there are some really cool there are other systems and from a strategy standpoint, it’s a lot of the same strategy, it just depends as you start connecting your systems and all your Marv TEKsystems, if you can connect them all in a powerful way depending on what you’re trying to do and how to complicate your sales process is and how specific contextual and cohesive and per size you need to be in those bot conversation, that’s where a lot of times push people a little bit more towards drift but there are some very cheap alternative options just build some basic Bots that will move it along for sure

Billy: Yeah I agree with you you’re right on that so before we wrap up though I just wanted to give people a chance if they want to get a hold of you to want to continue the conversation and learn more about you and ultra-cool, what’s the best way for them to get to contact you?

Eliav: I would I would say just go to www.ultracool.io and go through the bots and check it out, we mostly put out a lot of videos that are on there that are just basic information  of walking through putting branding into a bot and why that’s important or the user experience of how to make the bots not annoying and then also walks through like all the information about using personalization in Bots and then you’re welcome to go to the bottom book at times chat or people are interested in checking out the differences, we’re always happy to chat with people and for the most part we work with people who have a minimum of 10,000 people a month they come to the site for the most part or you’ve amassed a huge outbound process to integrate it into and otherwise we have a lot of them but I think that’s the best way and then you can always walk we do a lot of stuff with drift too, we’re doing an upcoming insider course I’m talking about advanced bot building of some of stuff we were talking about today in strategy, you can always just get more information on drift insider too, at least some great videos on there, we’ll be doing some of those too and do some webinars to drift to just general information, people just want to learn and do it themselves

Billy: Yeah, awesome Eliav. Thank you so much for joining us, sharing the knowledge, really good stuff and we’ll chat later

Eliav: Absolutely, awesome Billy.

Billy: Okay.