Episode Transcript
[00:00:00] Jim Sterne: It's a completely different kind of technology than we're used to. You
know, we got computers and we automated the work that we do and we called it business
process reengineering. And then the Internet came along and we did digital transformation,
which was taking the work that we do that was automated and putting it on the Internet.
This is not going to be like that. This is more of a metamorphosis, a sea change. This is a
different kind of computing that is unexpected. It's not more calculations and more search and
more database. It's intellectual prowess. It is idea generating. And we're not used to that.
[00:00:46] Jeff Dillon: Welcome to another episode of the EdTechConnect podcast. Today's
guest is one of those rare voices who's been ahead of every digital wave before most of us even
knew it was forming. Jim Stern sold business computers to first time buyers in the 1980s,
consulted and keynoted on online marketing in the 90s, founded a conference in a professional
association on digital analytics in the 2000s. And now he's still running his marketing analytics
summit every year and advises companies on the adoption of generative AI. His 13 books on
Internet marketing and customer service include Artificial Intelligence for Marketing, Practical
Applications, and his latest, the New Science of Customer Relationships. Delivering on the one
to one promise with AI.
Jim, welcome to the show. It is great to have you today.
[00:01:43] Jim Sterne: Thanks very much, Jeff. It is a privilege and honor and just a delight to be
here.
[00:01:47] Jeff Dillon: Well, I have to say I really enjoyed your keynote at Digital Collegium. You
have this rare ability to make the evolution of technology feel accessible and exciting.
And the way you framed marketing has adapted through each wave from the early days of the
web to now with AI. It stuck with me.
[00:02:09] Jim Sterne: Oh, great. Well, thank you so much. I appreciate the kind words and I'm
glad that the work I'm doing is working. I like to introduce myself as a professional explainer and
apparently I don't suck at it.
[00:02:24] Jeff Dillon: Well, it's clear you've not just watched them happen, like you've defined
how people think of them. And you've been a front row witness to every major shift in digital
marketing since the 80s. And I want to ask you, like, what's surprised you the most about this
latest wave of of generative AI?
[00:02:40] Jim Sterne: Wow. It's a completely different kind of technology than we're used to. You
know, we got computers and we automated the work that we do and we called it business
process reengineering.
And then the Internet came along and we did digital transformation, which was taking the work
that we do that was automated and putting it on the Internet.
This is not going to be like that. This is more of a metamorphosis, a sea change. This is a
different kind of computing that is unexpected.
It's not more calculations and more search and more database. It's intellectual prowess. It is idea
generating, and we're not used to that.
[00:03:23] Jeff Dillon: It's changing how we think. Right. I feel like it's changing. Be like I'm
falling right into it. I love it because it's teaching me how to ask the right questions. If you're good
at asking questions, it can really help you.
[00:03:35] Jim Sterne: I'm sorry, but you just summed up my entire keynote speech. Because
there's a big concern that it's also dumbing us down. If I take Ubers or I use gps, I don't learn
how to navigate. I just go where it tells me. And I don't really know where I am, but I know I can
get back. If I have to navigate and orienteer myself, I learn.
So if I just take the answer and run with it, I'm doing myself and my organization a disservice.
But if I'm using it to help me hone my critical thinking skills. Bingo.
[00:04:08] Jeff Dillon: And I've noticed people say AI is going to take our jobs and there's this
fear. Right. We've had this conversation for a couple years now, and sure, I think it will some
jobs, but when midjourney released the prompts that were creating these incredible images, I
was amazed because I'm like, I could do that, but why couldn't I create these same images all
these people were doing? Mine didn't come out the same. And then I looked at the prompts and
I'm like, well, guess what? These people are professional photographers that are saying, use
this camera angle with this aspect ratio and this lens. And I'm like, I could have never written that
prompt. So it's just a whole nother way to think about it. Yeah. If you don't embrace it, sure, it
might take your job, but it's this powerful tool, right?
[00:04:49] Jim Sterne: Yeah, yeah, yeah.
[00:04:50] Jeff Dillon: What really stuck with me was how you combine clarity with
pragmaticism. How do you approach translating complex tech shifts into actionable insights?
[00:05:01] Jim Sterne: Well, thank you. Don't stop now. Tell me more about me.
[00:05:06] Jeff Dillon: You coined the role fractional evangelist. What does that mean in
practice? What kinds of organizations benefit most from that kind of guidance today?
[00:05:13] Jim Sterne: Well, a couple of things. First of all, in a large organization, the peak of the
consulting pyramid is the evangelist who comes in and says, let me explain what it is so that you
understand it. So you understand why you need to hire the next guy to help you build a strategy
and then the next person to help you build a plan and the next person to execute the plan, I'm
there waving the flag saying, hey, look over here, this is important, this is new, it's not what you
expect. And here's how to approach it. So that's evangelizing within the organization.
The other side, my other clients are vendors, they're startups. We have this great tool. How do
we let people know how great it is? It's like, well, you got to explain it in terms they can
understand because your website is just one big jargon word salad. And I actually think I know
what I mean. In fact, quick short story.
Had a client in Europe wanting to bring their product to the US and they sent me 50 megabytes
of PDFs at my request explaining what the software did.
And I threw it into ChatGPT and had a conversation with the documentation until after 30
minutes I emailed back and said okay. The hour long meetings I have with the cto, the CMO and
the founder, I don't need those anymore because I now I understand what the software does.
[00:06:37] Jeff Dillon: Well, when it comes to generative AI, many leaders feel overwhelmed.
Where do you suggest they start? Especially if they're in marketing or education.
[00:06:45] Jim Sterne: Yeah, two things at the same time. Number one, play with it. Dedicate an
hour a day. And it's a lot, I know, but this is, it's important. It's a big business changing tool, so
you need to try it and learn what it, what's good at, what it's bad at. And active learning means
you're working with other people, not just take a course online or not just play with it, but actively
learn it for an hour a day. The second part is bring people together. So the folks who are
interested in it, in your organization should come together and form a council who will be
responsible for figuring out where the company is, what is the risk tolerance culturally in order to
create policy. So I need the IT department and the legal department and, and the operations
people to help put together a policy. Then we can put together a formal training program. Here
are the rules and regs, here are the guidelines. Now we're going to walk you through how to use
this for your job.
[00:07:44] Jeff Dillon: One of the basic things I tell universities when I consult for them is make
sure your people are sharing the prompt library. Don't shame them for using AI. I think we're
getting there to where we all know we're using some prompts. Like, let's share those. Right. Your
person leaves, they're really good.
I lost all that. So I get some resistance though. There's a lot of people in the trenches that are
good this, that are reluctant to share their prompts. And so what I'm getting at is this is often
writing, it's often in a creative mode. I'm creating some sort of writing piece on the coding side.
I see a lot of people proud that I coded this with AI. Look at this. I'm proud. Like, this is amazing.
There's this dichotomy between using AI for code and writing. Why do you think that is?
[00:08:27] Jim Sterne: Code is engineering, it's architecture, it's mechanical. And writing is
creative and fluffy and emotional. And so when I come up with an elegant way to make the
computer do what I want, I want to show off how skilled I am. But if I come up with some flowery
language, well, wait a minute, because my language is flowery.
That's me, that's my identity that I created that. And if I used a tool to be creative that way, no, I
want to take credit for it. But a coder using a tool to write code for a tool that is their identity. I
was able to use this higher level thinking in order to create more powerful software. I think it's
the right brain, left brain thing.
[00:09:16] Jeff Dillon: Yeah, I think it does come down to tapping into our, our identity. You know,
it's, it's scary.
[00:09:22] Jim Sterne: Well, this is really interesting. I've spent the last 25 years diving into
marketing analytics. I run the Marketing Analytics Summit every year.
And it's really tough because people who are doing the job, their identity is in the task. I am a
software or a data engineer. I know how to get the data and build a pipeline. I am a data
cleaning expert. I am a SQL query query expert.
And now if the machine can write queries for me, what am I? Who am I? And that's the elevation
of, okay, you were down here in the trenches. Now you're going to come up to become the
trusted advisor and have an opinion about what the data results are and have recommendations
for the business based on the two most important things that humans bring to the table are
context and taste.
Now we know that these machines hallucinate. Hallucination has become a technical term. Taste
is becoming a technical term. It's the thing that the computer doesn't have. It is an opinion about
the recommendation. And I'm going to Be emphatic that I believe this is the right way, because
the data suggests this is what we should do is a trusted advisor rather than.
Here's a dashboard.
[00:10:43] JEFF: And now a word from our sponsor.
[00:10:47] AD: How can your next campaign soar?
With experience helping colleges and universities raise billions of dollars, Mackie Strategies
delivers communications, fundraising and tech expertise that your campaign can take to the
bank. Mackie Strategies, build your breakthrough.
[00:11:10] Jeff Dillon: I want to go back because some of the stories you told in that keynote
were really starting this evolution where we could see how everything kind of got going. You
were a pioneer in web analytics before most of us even had websites. How do you see the
lessons from that era shaping how we measure AI powered interactions today?
[00:11:30] Jim Sterne: The difference, first of all, is that in the beginning there was the
webmaster and there was one person in the whole enterprise that knew what a website was and
created one. Then they had to evangelize that within the company to get people to understand
this. Everybody has access to it all at once and everybody can use it all at once.
So what used to be here is this little thing called a website. Let me explain how it's going to
change everything to everybody has it all at once and everybody has misconceptions about it.
And it's not up to the webmaster to explain it to everybody. It's up to everybody to work together.
This is what's different. There's still going to be the spectrum of, on the one hand, oh, no, can't
touch it, regulatory reasons, we don't understand it, legal reasons, no, no, nobody can use it all
the way to the other side, which is run fast and break things. And both of these are dangerous.
So we have to be somewhere in the middle there.
[00:12:27] Jeff Dillon: You said webmaster and you, you jogged a memory. I have to date myself
here and tell you a story about the webmaster life back in the day, because that was my first job
in university was the webmaster. They took the job out of the IT team. It was some guy doing it
on the side that was like a developer. And they said, we're putting this in marketing. A lot of
schools are doing that, but it was not that elegant looking even for back in the day. So I took it,
but I was the enemy of IT at that point. So I was this webmaster and I said, okay, I'll, I'll handle
the webmaster email. I was getting all these emails like, how much does tuition? What chemistry
degrees you offer, athletics programs? And I'd go to the marketing team, I'd say, I think we're
Onto something here. I think, I think we should do something more with this website.
And what I did was no one would really listen. They're like, I just keep answering them. I'm like, I
don't have time for this. So I created a pull down menu that had all the different areas and I got
email addresses for everybody. I'm like, okay, I'll take the leftovers, but you guys got to handle all
this. That was my right. What can I do? But it was pretty funny that it hadn't been discovered that
this could be our primary channel of recruiting students.
[00:13:29] Jim Sterne: You know, a parallel story is 1995, 96. I'm in the IT department, IT
marketing meeting at Hewlett Packard. They're meeting for the first time and there is a half an
hour long argument over whether or not they should put their 800 number on the website.
And I said, why would you not do that? It's like, we don't want to be overwhelmed with phone
calls. It's like, yes, you do. Well, who's going to pay for it? Okay, that's on you. I don't know. I'm
backing out of that argument.
[00:14:00] Jeff Dillon: Yeah, it's crazy to see how far we've come. In your latest writing, you
emphasized the value of trust in AI. How do organizations like universities build that trust
internally and with their audiences?
[00:14:15] Jim Sterne: Well, that's two very different questions. So the first one, internally, it's
training, training, training, practice, sharing. It is understanding how you can trust it. You can
trust it to be creative, you can trust it to have ideas and opinions. Do not trust it for facts. The
foundation, the large language model builders are working hard to make it figure out how to do
math. So that's a problem they recognize that they are solving.
But anytime you get an output that says this is the truth, it's on you to validate and verify before
you publish. And so trust is, do we trust ourselves to know how to use it? And how much do you
trust the machine? This goes back to the fact that it is not a calculator. It's not a database, it's
not a search engine. It is a generative thing. So my wife is a superior court judge. And when I
explained to her how it works and how it uses word vector scoring to figure out what the next
word should be, she said, wait a minute. What you're telling me is you ask it something and it
gives you what it thinks is the best answer you want to hear rather than the truth? I said, yeah,
that's right. She said, oh, so it's just like witnesses in my courtroom.
[00:15:29] Jeff Dillon: Exactly, exactly.
[00:15:31] Jim Sterne: Humans. I will give you an answer that I think sounds right without, you
know, checking it five times in advance.
But if I'm writing a peer reviewed paper, I'm going to check all of my sources. So that's the
internal. The external is, how do your students, your donors, your faculty member, how can they
trust your use of AI? I just read this morning a big pushback. Where was it? I think it was BBC
had an article where somebody said, I applied for a job. And they said, okay, well the first
interview is going to be on Zoom with AI. And he said, well, if it's not worth your time, it's not
worth my time for having the job. So no. Bye. So there's. I don't trust you. If you can't be
bothered, then go away.
So there is this.
Where do I, as a consumer want to use this tool?
I go to the website, there's a dropdown menu, I find what I want. Great.
Actually, I have a question that's a little bit more complex. It could take me a half an hour to find
it on your website.
I could call the call center and get put on hold and then the line is dropped and they don't know
the answer and somebody will get back to me. Or I could talk to the chatbots of old. The little
thing in the corner that I click on and, and I ask it a question, it says, I'm sorry, I don't understand
your question. It's like, why did I bother? Now they understand the question.
This morning I called Apple to find out about upgrading my iPad, that's ancient.
And I had a chatbot, verbal chatbot, saying, I understand complete sentences. How can I help
you? Well, I have this product, I want to upgrade it. I want to know what the options are. I want
to know if I can pick it up in the store and can I make an appointment for tomorrow. And it
responded, yes, all of that is doable. And let me turn you over to a human.
Yes, it's exactly what I want.
[00:17:31] Jeff Dillon: I want to throw one more example of trust out there that I've seen
universities starting to experiment with. There's tools out there like hey Gen. And yes, I think
there's many. Hey gen is one of these translators. You can upload your video and make it look
like you're speaking a different language. I've used it, it's great.
But you want your president of your university to deliver a speech in the native language of
these international students.
You have to Tell them this is AI.
[00:18:02] Jim Sterne: Oh yeah.
[00:18:03] Jeff Dillon: Because if you don't, the promise is people could think, I have a president
that's speaking multi languages. A great promise is to say we care enough about you to make
sure you can hear this in your native language. That's a great promise.
[00:18:15] Jim Sterne: Yes.
And with the caveat that I need a native speaker to listen to that speech, to validate and verify it.
Because translation and interpretation are different.
[00:18:30] Jeff Dillon: Right. And don't take that as the final result.
[00:18:32] Jim Sterne: Yeah, yeah. Because that becomes a question of making promises. That
is advertising and you're on the hook. If the president said, we are often free for lunch on
Thursdays and it gets translated as we offer free lunches on Thursday, that is an advertised
promise that now you have to live up to.
[00:18:51] Jeff Dillon: I think we're going to see that coming soon. Some sort of translation error
that's going to cause chaos. But hopefully not in higher, I don't know what industry, but from your
consulting work, what's the biggest misconception marketing leaders have about generative AI
right now?
[00:19:07] Jim Sterne: Boy, it's two extremes. This doesn't work. It's going to take my job.
I'm terrified. I'm losing my identity. And on the other side, oh, I can fire half my people. It can do
the work, it cannot do the job. It can do tasks and it can do tasks as well as it can given who's
asking it to do what.
I also cannot get wonderful pictures out of mid journey. Not a graphic artist, not a photographer,
not a fine art painter. I don't have the vocabulary.
So if I ask for artwork for my marketing, I'm going to get really lame results.
So I need to not fire the people who have context and have knowledge and understand the
bigger picture.
On the other hand, the naysayers and the Luddites are just going to get left behind. I mean, I
said that for 10 years about the world wide web. You have to have a website and it took 10 years
for people to go, yeah, I guess you're right. This is 10 minutes.
[00:20:11] Jeff Dillon: You've launched conferences, chaired boards, written bestsellers. How do
you stay grounded and forward looking in such a fast moving field?
[00:20:22] Jim Sterne: It is more of a challenge. You know, 25 years in digital analytics and it's
like, oh, we're answering the same questions and solving the same problems again. Okay, fine,
I'll write another article about that.
And then this comes along. It's like, oh, wait, hold my beer. Something new to learn. And it's
very exciting. It lights me up. And I think the Reason that I'm so lucky that way is that my job is
to understand it so that I might explain it to others instead of my job is to crank out yet another
email and create yet another Persona and measure yet another open rate, in which case, great,
I'm going to automate things. That's fine, but it's not fun.
The human brain does not like to have to change, and it is biological. I am trying to conserve
calories. I don't want to burn up all my calories in thinking hard about something, but I think it's
fun. So rather, I mean, I've got physical energy, I've got emotional energy, and I've got
intellectual energy. And if at the end of the day, I've used all three of those, I've had a good day
and I will sleep well. It.
[00:21:31] Jeff Dillon: I've found that it's really made me super productive, almost to the point
where I expect so much of myself now. Like, I love it so much, I can't stop. And my wife has to
tell me, like, hey, what are you thinking about, Renee? Are you gonna stop working? Are you?
I'm like, oh, yeah, yeah. But I'm the same. Like, what could I be doing with this tool? How can I
connect this tool with what this school wants? But it's kind of maddening, you know.
[00:21:54] Jim Sterne: Okay, well, you know, boundaries.
Give yourself a schedule.
And the danger. I mean, the first week that I had access to deep research, I had 10 tabs open.
With each one, had 25 pages of results of things that had found out that I want to know about.
But I don't have time to read 10, 25 page documents today and not just read them and not just
understand them, but vet the validity of the content.
That's a big job. So, yeah, be careful.
[00:22:29] Jeff Dillon: What excites you the most about the near future of AI Marketing and what
worries you?
[00:22:35] Jim Sterne: What excites me is the possibilities. The ability to get out of the toil and
into the creativity. Be less of the direct marketing catalog mailing machine and more of the Don
Draper looking up at the ceiling and thinking about humanity. And. And our role is to identify what
resonates. Here's our new product. Here's our target audience. Here are our competitors. Here's
a whole new department that we're starting at university.
Well, how does this look competitively? And give me 25 ideas about a billboard I can put up on
the highway to tell everybody.
My job is to look through 25 of them and say, 23 of these are really bad.
These two are interesting. Let's work on those. That's the fun part.
What worries me is job displacement is people thinking, oh, we have a marketing department
with five people, we can do it with two people, it's like, no, you can't. But with five people, you
can do the work of 10. Right, right. So that mind shift is what concerns me because some
organizations just think, oh, we have to cut costs. Whatever we do, we have to cut costs. Well,
okay, but don't cut the people. They have the intent, the desire, the taste, and the context that
the machine just doesn't know about yet.
[00:24:03] Jeff Dillon: You come from an analytics background. How do you see the role of
analytics evolving? As generative AI becomes more integrated into.
[00:24:11] Jim Sterne: Content and campaigns, it is becoming that trusted advisor I talked about
before. It's getting out of the weeds. And instead of crafting the perfect SQL query or creating
the best R package or Python routine, I turn to AI and say, what's new? What are the
anomalies?
What should I be looking at? What seems to be working better? And I always fall back on. Never
ask for questions of fact, ask for opinion.
So instead of how many people opened this email, clicked on this link, and downloaded that
PDF, the question is, where do we seem to be making the same problem over and over again?
And it'll tell you something.
Is it true?
That's on you? Is it helpful? Does it make you think? Then it's helpful.
[00:25:04] Jeff Dillon: Well, for teams starting to pilot AI tools, what's one mind shift that they
need to embrace to succeed, celebrate failure.
[00:25:12] Jim Sterne: And this is antithetical to humans. Like, look at all the stupid things I did is
really helpful.
I had one individual come to one of my conferences and say they were just starting to push, put
in a very expensive web analytics tool and they were there to learn what they could. They just
got hired. And I said, oh, do me a favor, Keep a journal and everything that you're doing and
lessons learned, and come back next year and present your experience to my audience.
And he did. He came back the next year and he got up and his presentation was the 10 stupid
mistakes I made.
One of the best presentations. Everybody was just like, oh, my God, I was just about to do that.
Thank you so much.
[00:26:00] Jeff Dillon: I love that. That's definitely clickbait. I would go to that.
I gotta name a session that at one of these conferences, maybe.
So as we close, your next book is titled the New Science of Customer Relationships. Can you
give us a preview? What's the one idea you hope readers will walk away with?
[00:26:18] Jim Sterne: Well, it's delivering on the one to one promise, which was something that
Don Peppers and Martha Rogers wrote about in, in 1994 in their book the One to One Future,
which was about database marketing. Now that we have computers and databases for
marketing, we can give people what they want. We can remember them and give them what.
And this has been an amazing uphill battle ever since. But with generative AI, the ability to
understand not just the Boolean search, but the context of what they're asking means that we
can respond.
Like my experience with Apple on the phone this morning, they can understand the full
sentence, parse it out, and say, oh, well, in that case, I know who to send you to, or I know what
answer to give you, or I know what question to ask next.
And so it is this idea that we have more data. Right. We went through the big data phase. Yay.
We've got all this data and we can do anything. Yeah, no, it's harder than that.
But now we have a new tool to let us swim through the data, parse the data, have a
conversation with the data in order to identify Personas. So, you know, we've got a Persona of
an alumni, a donor. Well, actually, you can have seven different donors now because you can
handle it and identify, given the signals that you're getting in from them, you can identify not just
what message to send and at what time, but using what tone.
And that can be automated.
[00:27:56] Jeff Dillon: Well, I loved this conversation. Jim, if someone wants to get a hold of you,
what's the best way to contact you?
[00:28:03] Jim Sterne: I'm on LinkedIn all day, every day. That's the easiest. And then my
website is targeting.com, which is my consulting. And I'll point again to the Marketing Analytics
Summit coming up in April in Santa Barbara, California.
[00:28:17] Jeff Dillon: Excellent. We'll put those links in the show notes. And again, thanks for
being on. Loved it.
[00:28:21] Jim Sterne: Thank you, Jeff. Pleasure.
[00:28:24] Speaker C: As we wrap up this episode, remember EdTech Connect is your trusted
companion on your journey to enhance education through technology.
Whether you're looking to spark students student engagement, refine edtech implementation
strategies, or stay ahead of the curve in emerging technologies, EdTech Connect brings you the
insights you need. Be sure to subscribe on your favorite podcast platform so you never miss an
inspiring and informative episode. And while you're there, please leave us a review. Your
feedback fuels us to keep bringing you valuable content. For even more resources and
connections, head over to EdTech and your hub for EdTech reviews, trends and solutions. Until
next time, thanks for tuning in.
