Cheryl Broom [00:00:08]:
Hi, I'm Cheryl Broom, CEO of GradComm and host of higher Ed conversations. Thank you so
much for joining me today. I am thrilled to have on the podcast today, Nick Martin with Direct AI.
Nick reached out to me a couple months ago because he was interested in the podcast and
sharing a little bit about his amazing product, which creates a generated chatbots on websites.
But the different thing about his chatbots, his company's chatbots, is that they're actually two
way. They have personality. They are direct with their customers. You can use them for lead
generation.
Cheryl Broom [00:00:46]:
You can connect questions to events or to other questions to give users a more personalized
experience when they're chatting with your website. So it goes much, much further beyond the
traditional chatbot that we're all used to seeing, which is basically like a replication of your
search bar to really creating an amazing experience for a potential student, for their parents,
and even for your own faculty and staff. So today's podcast is all about using AI, using the chat
bot and improving your search technology to make a better experience for your students. And in
Nick's own words, it's all about being obsessed with your customer, and that's what his chatbots
are. So I took away so many ideas of how to use chatbots and AI technology to improve the
student experience, and I think you're going to be inspired as well. I hope you enjoy the
conversation. Well, Nick, thank you so much for being here today. I'm thrilled to have you on the
podcast.
Nick Martin [00:01:52]:
Thanks for having me. I'm thrilled to be here.
Cheryl Broom [00:01:53]:
You are a true college success story. When we first talked, I was really, really interested in your
success story, how you founded your business, working out of your dorm room, tell our listeners
a little bit about how you got where you are today.
Nick Martin [00:02:08]:
Sure. So when I was in college, what feels like many moons ago now, I had started a direct to
consumer sports brand with a teammate of mine, and we built that business up and had a good
amount of people working for us. And we were selling a lot of product with a lot of great
customers. Duke University, Syracuse, UMC, a bunch of colleges and a bunch of other sports
teams. And then we dropped out of college to continue building that business, did that for a few
years, and then started direct with a couple co founders. And direct is a conversational AI
platform that helps businesses that produce a lot of content, training AI on all of their content
and data so that they can have one to one personalized conversations with all their audiences.
And so we've been doing that for a couple of years about two years ago, this thing called
chatgpt happened and the whole world kind of woke up to the possibilities of AI and having
conversations with an audience. And so it's been a bit of a wild ride over the last 1824 months
for us, but we're really excited about it.Nick Martin [00:03:01]:
We've got 100 great brands on the platform, millions of users, and we think we're doing really
important work and helping a lot of really important companies along the way. So that's a little bit
about what we've done. I'm excited to get into more detail.
Cheryl Broom [00:03:13]:
That's great. And it's funny, I got hung up on you saying it feels like many moons ago, because it
wasn't many moons ago for you.
Nick Martin [00:03:19]:
Well, at my current moment in life, all my friends are having their first quarter life, third life crisis,
and so it's all relative, I think. But I appreciate the sentiment.
Cheryl Broom [00:03:31]:
That's great. So I was really interested when we first started talking about AI's and chatbots, that
yours is a little bit different and it's actually trained to speak in the brand voice that you're
interacting with. Like, how does that work and why is it important?
Nick Martin [00:03:46]:
Well, I think it's important because especially as the Internet continues to move towards a place
where social media algorithms and now these AI search engines really disintermediate brands
from their customers or from their audiences. The importance of having a direct relationship with
your audience, your student, your customer, whoever it might be, is more important than ever
because it's getting harder and harder to reach them. Right. If I go to Google and I search for
something now and I get their new AI overview, I might get the answer right there. I actually
don't have to go to the website from the company or the publisher or the brand. If I'm on social
media, I don't actually see who I follow anymore. I see what's recommended to me. We've
moved from this sort of social graph to the recommendation graph.
Nick Martin [00:04:27]:
And so having a direct relationship with your constituent, your audience, your student, your
customer, your users, it's never been more important. And we believe fundamentally, the only
way to actually have a relationship with someone that's meaningful is to have conversations with
them. Right. If you remove yourselves from sort of the digital world and technology and you just
think practically as a human, well, there's nobody that you have a meaningful relationship with
that you've never actually spoken to or that you talk to. And so for the first time, because of
generative AI and natural language processing, you can have conversations with thousands,
hundreds of thousands or millions of people in a way that's one on one and personalized and
really feels authentic and true and not like this kind of cookie cutter experience I think everybody
got used to. So you asked, what's different if you were to experience the typical chatbot on a
website a couple of years ago and even today on most websites, it's probably in the bottom right
hand corner. It's probably there for customer service primarily. So it's going to try to help you fix
a problem with something you bought, return a package, understand how to get your Internet to
work, or change a flight that you bought some sort of service based use case, which is all fine.Nick Martin [00:05:33]:
Typically, though, those experiences were really rudimentary. Right. There's only a couple
questions they know the answers to. You tend to be begging for a live agent really quickly
because it's not able to solve your problems. And direct takes this perspective that if you
leverage AI in the right ways and you combine that with the actual content that publisher, a
company, a college might produce, and you give it a personality that is aligned with the brand's
personality that already exists today, you can actually engage that user, one, in a way that's
delightful for the end user, but two, you can train these systems to be really effective at actually
helping them get their answers, find the content they need, move on with their day, or dig
deeper if there's something they're trying to engage with you about. And so at a high level, what
we hope we're doing a good job of, and we've got some data to suggest that we are, is making it
so that brands can connect much more deeply in a one on one, personalized way with their
audience, and doing that through conversations so that that audience comes back more often,
spends more time with them, and has a better experience overall.
Cheryl Broom [00:06:35]:
Well, I love this idea of AI having some sort of personality in your interactions. I guess you would
have to really start with having a well defined, like, brand voice on your site, right? I mean, you
have to have done some work to know who you are and what you want your institution to be.
Because I'm guessing your AI could be funny or serious. Sure.
Nick Martin [00:07:00]:
Yeah, it's exactly right. You're much better equipped to get started quickly if you already know
how you want your brand to show up. A lot of brands we'll speak with might not have a really
clean one sheeter on exactly what that's supposed to be. Some do, but some don't. The good
news is almost everybody's been using social in one way or another for about a decade. And so
if you go look at the way a brand is tweeting, the way your university or college is tweeting or
posting on Instagram or the captions of their TikTok posts, all of that is your brand voice. Now
maybe thats not, again, incredibly well documented or structured or thoughtful. A lot of times it
is.
Nick Martin [00:07:37]:
But even if it isnt, you are putting a voice out into the world. And I think if you think youre having
success on those channels or even on your website or in your app, well then you can mimic that
same experience. Now what Rai does is it trains on all of your content that you already have. So
it's going to pick up on how you're talking today. How are you posting content online? What is
the tone of voice you already have? But then you can go in and you can customize the
personality as you see fit. So if you wanted to be sarcastic and funny and cheeky, the way
Wendy's might be on social media making fun of McDonald's, well, you can do that. You can
say, hey, you're sarcastic, you're going to lightly joke with the user, but you're never going to be
too offensive and you're going to use emojis to let them know you're lighthearted and it'll, it'll
really embody those instructions and how it behaves. It's really cool.Nick Martin [00:08:24]:
At the same time, if you were a buttoned up, disciplined, you know, very strict brand for
whatever reason, maybe you're Harvard and you just want to feel like elite and premium, you
can tell it to behave that way as well. But yeah, the better understanding you have of how you
want your voice to resonate with the audience, the better the system's gonna do it. Picking up
on that, we have seen very early indications of companies thinking about these AI's as akin to
like, their mascot, right? So the president of a university has to be polished and professional and
a really good order and all these things. The mascot gets to run out at halftime and do backflips
and pretend to be mean to the opposing fans and can just do a lot of things, right. Mascots have
a little bit of more permission. And so we're starting to see some brands play with this idea on.
Well, is my AI actually going to have its own name, its own logo, its own identity? And does that
give it permission to be a little more free in how it engages with the audience? So it's not
necessarily that it's the Harvard chatbot or AI, which it could be, but maybe it's this new
character they created, which still has all the info, has all the data, but is allowed to be a little
more fun because that's the purpose of it. And that's obvious to the user based on it having its
own visual representation in how it manifests itself to the end user.
Nick Martin [00:09:44]:
So a couple of thoughts there on just like, the different ways you can do it, but we certainly think
personality is critical, because otherwise you don't have anything that differentiates you. In a
world where anyone can chat with information anywhere, what's going to make your experience
stand out? And one of it, one part, is utility. Do you give them what they need? Are you really
good at that? I think you should assume that's like table stakes. And then the differentiator is
always, well, what is unique about your brand voice and how you show up in the market?
Cheryl Broom [00:10:10]:
Now, we have a client, it's a consortium of ten colleges, and they're competitors, but they've
decided that for certain career education programs, they're going to market holistically, because
they're not competing with each other on these programs, they're complementing each other,
and it gives them more money to buy media in the Los Angeles area. And our website we built
for them, it does have a chatbot, right? And our number one question, and our chatbot's not, it
doesn't have a name. That's. Now I'm like, huh, we should name it. Like, maybe that would,
maybe that would help people. But our number one question, our first question is, are you a
person?
Nick Martin [00:10:47]:
It's amazing.
Cheryl Broom [00:10:48]:
And I wonder, from your experience working with other brands, is having a personality, is having
a name, does that make people more comfortable in interacting or more patients, or what have
you seen from that one?
Nick Martin [00:11:02]:We see more engagement overall. People just interact with it more often and more deeply. One
of the reasons we believe that is, without that, it's not obvious to the user if it's a chatbot, if it's a
customer service bot, or if it's just customer service. So a lot of experiences across websites,
just in general, they all look the same. But sometimes if I'm on making this up at and t.com, and
I type into the little chatbot, it's literally just a person the whole time. Then if I go to Verizon,
again making this up, it's a chatbot, there's no person. So the user always, if they are trying to
talk to a person, they're like, the first thing I have to ask is this is a person, I don't have a way of
knowing. And when you change it to be a character, a Persona, a mascot, whatever it might be,
well, it's really obvious it's not a person.
Nick Martin [00:11:51]:
So it does solve for maybe that initial confusion. I think the more, maybe more important thing to
focus on is if users are asking that question often, it's probably because they've been having a
lot of experiences where they aren't able to get the information they're looking for and they need
to talk to a human, right? People are asking, are you a human? Because they're used to the AI
or the lack of AI in the chatbot not doing a good enough job, giving them what they're looking
for. And so you get frustrated, you send four messages and you're like, please, live agent, live
agent, human person. Talk to a real person. Youre sending all these messages to get in.
Cheryl Broom [00:12:22]:
I do that. Im like real person.
Nick Martin [00:12:24]:
Me too, all the time. And fortunately, in our experiences, we power, we dont tend to see that.
And theres tons of reasons why, one of which is the engagement and the types of experiences
were trying to deliver. Part of that is the branding and the way its integrated into the site. And
another part of that is the use case. I think if your use case is customer service and people are
trying to return packages, well, that's a reasonable thing to want to talk to a human about, but
you should be able to automate that and do a really good job of that with AI. I think if you're
getting that question too much, you might want to look at, well, how successful are you at
fulfilling what users are asking you for? Because if they're looking for a human, it's an indicator
that maybe you're not delivering on the promise of this chat experience that's supposed to solve
your problems or deepen your engagement, whatever it might be, to the extent that maybe you
should be or could be, and that could be an indicator of a moment where maybe it's worth
double clicking into just the quality of conversations you're delivering today.
Cheryl Broom [00:13:21]:
Now for our chatbots with customer service, I feel like a lot of the colleges I work with, that is
what their chatbot has become. It's almost like a auxiliary search function. Nobody can find
anything on search because there's 40,000 pages on the website that are outdated. So they've
decided to put a chatbot on. And now the chatbot just is pulling up the same crap that the
search for is pulling up. Like, how do you solve for that? How do you make it something more
than just a search function.Nick Martin [00:13:52]:
Yeah, that's a great point. So this is something we're seeing a lot of. So direct powers, these
conversational AI, so these assistants or chatbots and a lot of our customers early in the
deployment, or the first conversation or two will be like, oh, this could replace site search, the on
site search engine for their website. And we're like, yeah, of course it's going to do a better job
of that. It marries the best of what already worked in search, but fills in all the problems with
what didn't with really good AI. And that's a great thing about just having a chatbot that really
does know and understand all of your content. So just if a user has a question, you can give
them the right answer. One of the interesting things that we started to notice is like, well, it
doesn't make a lot of sense to have a search bar that takes a user to one set of results and a
chatbot that doesn't have the same result.
Nick Martin [00:14:42]:
Why would you have two input boxes on the same site? When is the user supposed to use
either one? How are they supposed to know? And if your search experience isn't working well
but your chatbot is, why even happen? All those questions started to emerge, and so direct is
actually now rolling out functionality so we can intercept searches and handle them with the AI.
The user can use the same search engine that they're already used to, but they'll get the
answers they're actually looking for. And so you're delivering a better experience. It's really a lot
of utility there, but it's also a way to onboard them, to activate them into this new conversational
experience that we believe is better for users, more delightful, more helpful across a whole
spectrum of use cases. It's an interesting Venn diagram. A lot of what chat bots do broadly is
search, but they can do a whole lot more than that. And so we're seeing kind of this moment in
time where they're coming together a little bit and chat is becoming the new way you search.
Right.
Nick Martin [00:15:35]:
I might ask a question, but I want to have a follow up question. Well, now I'm just in a
conversation. Is there a difference? Maybe when the user searching, certainly, you know, they
have an intent for a specific answer or piece of content. And so Rai will handle that a little
differently than if they're just chatting, where maybe they're trying to discover what they're
interested in, or maybe they have a different motivation. But certainly there's a lot of overlap and
broadly in the market, we're seeing that as well. So if you look at chat GPT, they just rolled out
search GPT, and they've announced plans that'll eventually be integrated into the core chat GPT
experience, which I think is evidence of, like, they can be different things, but they really should
be homogenous. And for the end user, I think the experience, any business, any college, any
publisher wants to deliver, is if your audience has a question, you should be able to answer it
really quickly, really delightfully, and then that should be an opportunity to build a direct
relationship. Maybe you ask them for their phone number, their email address, and you say you
can follow up with more information about the topic they're searching.
Nick Martin [00:16:32]:
Maybe you convert that into a conversation to better understand what's important to them so
you can meet their needs better on their next visit, and then a whole host of other thingsunderneath that as well. But we think of them as related concepts that a singular AI that truly
understands all of the information about your business should be able to handle really well.
Cheryl Broom [00:16:50]:
I love this idea of using your chatbot for lead generation, because that's something that a lot of
our clients are trying to do, is to connect, you know, one on one. Maybe connect an advisor to a
potential applicant. But nobody ever writes down anybody's names. Like, call the college, and
we'll come in with secret shoppers and see if anybody ever asked our shopper, like, their name
or their email or phone number. And it's like 99 out of 100 times. No, like. And there was never
follow up. Nobody ever called.
Cheryl Broom [00:17:20]:
So I love this idea of having the chat, like, if you were doing searches where it's clear that
somebody is trying to become a student. Is there a way then to ask them, like, hey, would you
like to actually talk to an advisor? If you give us your name and number, we can connect you
with someone. Have you seen that done before?
Nick Martin [00:17:37]:
Yeah, we have. And the data we've seen suggests it's incredibly effective. Right. One of the
really, like, analog, for lack of better term analogies here, is the sales rep in the beauty store or
the fashion store. Right? Like, that is effectively what the AI can turn into in the right context. I
think that's like, the really elegant difference. If you can engage people, let them engage how
they want to engage, and then find the right moments to say, oh, Cheryl is clearly potentially
interested in applying to be a student here. For her child to be a student here, that intent is
obvious in the conversation.
Nick Martin [00:18:13]:
Let's ask them if they want us to follow up with more information. Maybe somebody comes to a
college's website and they say, what time is the week three football game? And you can say, oh,
the week three football games at 12:00. I noticed you're not in the local area because maybe
you can look at the IP address and have a general frame for where they are. Are you coming in
from out of town to watch the game? Oh, yes, I am. Oh, interesting. Now, the AI also knows that
that's like parents weekend, so they can assume. Are you a parent of a student? Yes, I am.
Great.
Nick Martin [00:18:43]:
Well, just so you know, we have a lot of special events we're organizing for parents weekend.
Do you want us to send you an email when those are announced? Yeah, but here's my email.
So it can be really thoughtful and really contextual to who the user is, what they're interested in.
And if you think about the way you tend to communicate today, like if you get emails from any
business, college brand, retail, airline, whatever it might be, typically the bottom says do not
reply.
Cheryl Broom [00:19:08]:Right?
Nick Martin [00:19:09]:
Right. At the bottom of the automated emails, do not reply to me.
Cheryl Broom [00:19:11]:
Or you hit reply and the address literally is no reply.
Nick Martin [00:19:15]:
Right. And you get, you either get a bounce back or it's like, we can't help you. Not the best
experience. We think you should be able to reply. And so we imagine not that far off into the
future, not only would you be able to like collect emails or phone numbers to send text
messages or emails to people in a really thoughtful, contextual, relevant way, but if they then get
the email and they have another question, they can just hit reply and type the email and that
same AI assistant can respond to them. So they can send the list of events and then go, oh, do I
need tickets for the luncheon at 12:00 and the AI can actually just write the email back and be
like, yeah, here's the link for tickets. And so you can just deliver a much more connected
experience as if in the best version, you had this human assistant that knew everything about
the university available. Twenty four seven at your beck and call.
Cheryl Broom [00:19:58]:
I love that idea. That is so great. And I was, as you were talking, I was thinking about one of our
clients is a large community college district with a number of campuses and they're having what
they call Super Saturday event. And I think it might have actually been this past Saturday. We're
in a recording this in August, but it's coming out in October. So it was for fall semester. And what
an amazing thing. Like if people were clearly searching on like, how do I apply? Or when does
fall semester start? Or how do I get financial aid? That would have been such a great
opportunity to say to these people, hey, it looks like you're interested in coming this fall
semester.
Cheryl Broom [00:20:33]:
Did you know we have these great events going on? Like, you can talk to staff, you can visit the
campus. So that is not integrated in any of the AI's that I've seen at any of our clients. So how
do you get that information in there? Because is it just crawling the site or is there still a human
element behind it?
Nick Martin [00:20:52]:
Theres a couple different ways that you can build what we call the knowledge base, which is the
brain of your AI. Again, at a high level, we think every business, every institution, every
university, every publisher will have their own AI that knows everything about their business,
their data, their content, their customers, to train the knowledge, the brain of that AI. The first
thing we make really easy is just pulling in every piece of content youve ever put online. Were
going to just ingest everything on your website thats going to form the base, and thats really
easy. It takes no effort from the publisher or the brand, the university. You just click a button andit magically happens. Now you might have information thats not on your website. So what do
you do there? What do you do if you have some form that isnt publicly available that somebody
at your desk is emailing to people when they write in and say, how do I change my classes for
the second semester? The deadlines pass, and theres this exception paperwork.
Nick Martin [00:21:45]:
You can add all that types of external data to any knowledge base and then use that so that
there's awareness of it from the AI. Then you can also supplement sort of popular common
questions manually as well. And so one of the things that we'll see is we work with a lot of media
companies, they'll get a ton of questions, right? They have all this scale, they reach hundreds of
millions, billions of people, and we'll actually look at some of the data and go, huh, like 8% of
people were all curious about this one thing, and you've never written a story about this, or you
didn't have a story about this specific heart that they were interested in. Maybe that's an
opportunity to produce another piece of content. And so you can build this feedback loop where
you're really, as an organization, able to get smarter and smarter and better serve your
customers because you can learn what they really care about by what they talk to you about.
You can identify gaps in your knowledge base, so gaps in the information you have available to
fill in. And then over time, that experience can also just become much more personalized for the
user so it can remember, okay, this was Cheryl. She came here on the third week of the fall
semester for parents weekend for students, a sophomore for students majoring in this.
Nick Martin [00:22:52]:
And you can hold on to that. So when you come back the next time and you're like, how do I
change this class for my, my student that you're going to know, well, I remember who the
student is, you can actually just respond and be like, are you trying to get them out of
chemistry? Right? What's. And you can just really deliver an experience that to the end user
feels as if though they're talking to a person, even though they know they aren't. And so there's
a lot of opportunities around it and those are just a couple of them.
Cheryl Broom [00:23:18]:
So I'm going to take a quick break so we can hear about our sponsor. But when we come back,
I want to talk to you more about this concept of building the AI brain. I haven't heard it put that
way and I have a lot of questions about that. So we'll be right back.
Cheryl Broom [00:23:33]:
How do higher education decision makers find the right solution when technology evolves at
light speed? Well, we usually start with our network. Edtech Connect is the network that's
democratizing the higher ed technology conversation. Edtech Connect is free, so anyone with a
email address can sign up and list the software and services they use and their role at their
school. Once you're in, you can find out what solutions similar schools are doing over the
country. Whether you're looking to find a hot new AI tool or maybe learn options, you have to
upgrade your campus search engine or even get to your short list of marketing solution vendors,
Edtech Connect is the place to go. So visit edtechconnect.com and set up your free profile to
get a pulse for what's happening with higher ed technology today.Cheryl Broom [00:24:25]:
All right, Nick, so before we took a break, you were talking about two things that I thought were
so interesting. A was the AI brain and the other one was this idea of a feedback loop, which I
love. But before we talk about the feedback loop, this idea of building an AI brain, I mean, I think
we've always thought, at least I have, of AI as this like, computer or robot, not that we are
actually building like interconnectivity. Like a brain is not just like a pile of mush. It's like stuff like
moving around and moving around. So we were giving examples about like inviting somebody
like noticing through their interaction with AI that they're interested in being a student, and now
we're going to invite them to an event. Does the AI make that connection? Does it say, oh, I saw
on the event calendar that there's this event on Saturday, or does a human still have to tell the
AI to do that? How smart are your chat bots? How deep can they go in these connections?
Nick Martin [00:25:21]:
Yeah, it's a great question. So they're quite smart. They're getting smarter every day. So some
capabilities on this front that already exist, or we're going to remember who you are and what
you talked about. We're going to use that to inform how the chatbot behaves going forward. And
so that level of sort of personalization and continuous improvement is something we already do.
One of the other things we'll do is we'll actually run these. We think of them as like infinite
optimizations.
Nick Martin [00:25:43]:
So these real time, 24/7 dynamic tests. So let's say that your college has 10,000 people who
come to the website every day or every month. We will test 24 7365 different suggested
prompts or questions they can talk to the AI about. We might also start the conversations by
asking them a question. So maybe you want to learn something. Hey, are you excited for this?
Are you a prospective student? What's your favorite class? Maybe you want to collect some
insight on your audience for some purpose. We'll also run questions to the user dynamically all
the time, and what we're doing is we're looking at, well, what questions are users asking the
most, and what are they answering the most? And then, well, take that signal and we'll go, okay,
this looks like a meaningful cohort of people. Really care to talk about this? They really need to
understand this one problem.
Nick Martin [00:26:32]:
Then we'll recommend that out to a broader audience. And so you're delivering a better
experience to the majority just by always measuring what people are interacting with you about.
The other thing we can do is proactive recommendations in general. So maybe you're
navigating a site and you've injected a little experience where going to pre populate an FAQ
based on knowing where you're at in your journey. So, okay, this person is on a page about
tuition costs. They've been to the website twice in the last 30 days. They've never applied. They
keep searching for things around the idea of FAFSA or student aid of some form, depending on
the state they're in.
Nick Martin [00:27:10]:Maybe we want to proactively have the chat bot be like how important is qualifying for student
aid for your decision to apply? And the person go completely. It's the only thing that matters.
And then maybe that to come back to the previous point is an opportunity to go, well, do you
want us to email you with specific programs available based on your student? And you go,
yeah, that'd be great, and then go, well, tell me about your student. Where is she based? What
are her grip? And you could maybe tailor the follow up communication for them. Now, that last
part is a little bit of a sophisticated process. And the point where the human comes into this to
part of your question is getting the right data into the knowledge base, into the brain. So if you
just create a chatbot that's called college Chatbot 101, and it has no access to any information,
it can't do anything. If you let it scrape all of the content on the specific college's website, it's
going to know a lot.
Nick Martin [00:28:02]:
It's going to work a lot better than that. Search engine works because it's using modern
technology and better understanding the semantic breakdown of the text within the content, all
these little details. And then if you go to step three and you start layering in user behavior and
user data and holding onto that and having some sense of memory of what you've talked to
somebody about, it's even better. And then again, you go and you just keep moving sort of down
the rabbit hole of all the ways that you can enable the AI to be smarter as really an extension of
your brand to deliver the best student experience, parent experience, customer experience
possible.
Cheryl Broom [00:28:33]:
I just absolutely love this idea of the AI having a two way conversation, because when I, when I,
we first talked and I, we talked about your company, I was like, how is this different than every
other chat? You know, like, it's. But now that I've talked to you, I just see so many possibilities,
you know, like how great to ask questions that people don't even haven't even thought of to ask
themselves. Because maybe it knows that follow up questions from like 10,000 previous users
have been this. But now this person hasn't asked that question, so it's going to offer it. So now
it's becoming helpful. You know, it's, it's guiding a student rather than just me coming and trying
to figure it out on my own.
Nick Martin [00:29:14]:
Yeah. And I think if you want to, you know, look at the business impact of that type of
experience, like, it's really easy to imagine. Not only does it help the user ask the right question,
but it also learns, okay, not only were these most popular follow up questions, but of the six,
these three had a much higher conversion rate to getting an email. Right. And then this
particular question then had the best open rate on the emails. So you can also start layering
data over time to really get clever to understand that, oh, there's a knowledge gap when the,
when the parent or the student doesn't know the answer to this question, they hesitate to take
the next step. So if we can proactively recommend that and then give them the right answer
now that they feel more comfortable moving forward. And so we, again, one way to think about
this is like, if you had a humanity who knew literally everything about your university and was
available all the time, how much better could they make? Every single person's experience? Ithink that's this general North Star that's actually a good framework or proxy for what your AI
should be able to do over time.
Nick Martin [00:30:16]:
And then it's like, all right, well, obviously Suzanne is not available to talk to 10,000 people at a
time and to do all these things. But, man, if we could replicate Suzanne so everybody could
have the Suzanne experience, where Suzanne is the stand in the, for an incredible
administrative executive, maybe at a university, thats really where were going with it. And the
fun thing is, none of this was really possible up until recently. And so I think everybody that has
a big audience or a lot of information and content that their audience cares about. Right. And
theres plenty of that when it comes to colleges and universities. Youre a really unique position to
deliver an experience thats really differentiated. And I especially think if youre an institution thats
maybe leaning into technology, leaning into education around technology or AI, and thats
something thats important for you as a brand, then having your own experience is also a great
way to demonstrate, like were forward leaning.
Nick Martin [00:31:10]:
If youre going to come be a part of our institution, were going to educate you, get you ready for
the world, were showing you that were also a practitioner. We have our own AI. Were deploying
it. You can see that we're eating our own dog food. Again, to the extent that that maps to the
vision or the strategy or the thinking of any given education institution or business in general.
Cheryl Broom [00:31:30]:
Well, I think what's really important with what you said, there is so many colleges have brought
in their chatbot in order to solve, like, a personnel problem. So they're looking at it from an
institutional perspective. We don't have enough. What was the name the, my switchboard
operator, my college was Susan. We don't have enough susans we get 8000 calls a day. I don't
have enough Susan's. Let's stick this chatbot on and then maybe people won't call Susan all the
time. But really you need to be thinking about the student experience.
Cheryl Broom [00:31:57]:
How can this piece of technology enhance a student's experience, not solve our own
institutional problem? It's great if it can solve our institutional problem, of course, like your CFO
is going to love that, right? Your human resources going to like it, but really it's a tool to help a
student find what they need and to enhance their experience with your institution. So I think
that's a really important thing that colleges need to keep in mind when they start integrating
technology. It really comes down to your customer and your student and supporting them in
every way you can.
Nick Martin [00:32:30]:
And that that always actually comes back around. Right. Like if you're obsessed with your
customer experience, that's better for business. It ends up solving. People do call Suzanne less
or Susan less, right? Like when they know they can just text this AI and get every answer they
need, well, they dont have to call her. And so it works both ways. But to your point, I think theright north star for any business is who is our customer? What do they care about and how do
we give them the best experience possible? And then everything else sort of takes care of itself.
And not to be trite about it, but that compared to the inverse, its really obvious which one is
better in the long term.
Cheryl Broom [00:33:04]:
And the colleges ive seen that are the most successful are those ones that are obsessed with
their student experience from the minute that they visit their website to post graduation. And
they want to create the best experience possible.
Nick Martin [00:33:17]:
Well, and think of how important the testimonial is when you're picking a college, right? Like how
does a student pick a college? Maybe they watch a bunch of YouTube videos and tiktoks and
stuff that is in a way, testimonials and referrals and word of mouth in the modern day. But one of
the huge factors is who from their hometown went there? And when they go, did you like going
there? Says, yeah, it was great. And there's a lot of things outside of your control as the
university or the college in how that's going to go, right. Social dynamics and all these things.
But there's a lot that's in your control and anything you can do to enlighten or, excuse me,
enhance the experience that any given student has. There is that trickle down, that halo effect
that obviously happens where somebody goes, what was it like going to the school and they go,
it was great. The professors were awesome. The class sizes were good.
Nick Martin [00:34:05]:
And whenever I had a problem, you can actually send a text message and they get it fixed right
away. It feels incredible. All of those little things add up. Whereas if you're like, well, I was trying
to go to the website, and I couldn't change my class in time, and there was a deadline, and the
website crashed and it didn't work. And I was trying to search for what to do, but there was no
answer. And I tried to call, but they weren't working at those hours. So I got stuck in chinese and
I failed chinese, and I got set back a semester. I hate the school.
Nick Martin [00:34:29]:
Like, that's a crazy example, but that probably happens.
Cheryl Broom [00:34:32]:
Oh, it happens a lot.
Nick Martin [00:34:33]:
And why does it happen? It happens because somebody couldn't find the information that they
needed to find at the right moment in time. So there's a real impact to delivering this kind of
always available, really intelligent experience.
Cheryl Broom [00:34:44]:
And it's not a hard problem to solve.Nick Martin [00:34:46]:
Right?
Cheryl Broom [00:34:47]:
I mean, you can solve it. You can solve it with your company.
Nick Martin [00:34:51]:
Yeah.
Cheryl Broom [00:34:51]:
Well, this has been such a great. I was worried about our conversation being boring, but it's
been anything but boring. I am, like, full of ideas. I've written, like, two pages of ideas as we
talked, and I know that there's going to be some of our listeners that are interested in learning
more about your business and company. So how can they find you? How can they search for
you?
Nick Martin [00:35:10]:
Yeah, you can go to our website direct AI. Direct is d I r e q t. So d I r e q t a I. You can reach out
to us there. You can find me on LinkedIn. Nick Martin, you can send me an email and happy to
chat anytime.
Cheryl Broom [00:35:23]:
Wonderful. Well, thank you, Nick, so much. This has been such a great conversation. And Nick's
got, like, us, Rolodex of, like, every Fortune 500 company on this website. And maybe someday
we'll see some community colleges and four year universities rolling through there as well.
Nick Martin [00:35:38]:
We'd love to talk to them. Thanks for having me, Cheryl. This was really fun.
Cheryl Broom [00:35:41]:
Thank you.
Cheryl Broom [00:35:43]:
And that wraps up this episode of the Higher Education Conversations podcast. I'm host and
GradComm CEO Cheryl Broom. A big thank you to our sponsor, Edtech Connect. Edtech
Connect is free, so anyone with the email address can sign up and list the software and
services they use and their role at their school. So visit edtechconnect.com and set up your free
profile to get a pulse for what's happening. With higher ed technology today. And while you're
online, take a few minutes to leave our podcast, a five star review. It will help other colleges and
universities find us and learn from the great.
Cheryl Broom [00:36:19]:
Experts we have on the show.Cheryl Broom [00:36:21]:
That's it for now.
Cheryl Broom [00:36:23]:
Until next time.