Subscribe
Share
Share
Embed
A daily podcast delving into the biggest stories of the day throughout the sports betting and igaming sector.
Unless there is a very, very big change
in how these AI and
large language models work.
One thing that they still don't have and
probably won't have in
the near future is judgment
and taste.
Hello everyone and live
from SBC Summit America's 2026.
We are here for another episode of Eye
Gaming Daily brought to
you by OptiMove, the creator
of positionless marketing and number one
play engagement solution
for Eye Gaming and sports
betting operators.
I'm Fernands Nonot, media manager for SBC
and your host for today.
We'll wait a very special guest today,
someone from OptiMove
themselves, Shai Frank.
How are you today?
Shai Frank, SPP of product and GM of
America's for OptiMove.
Yes, that's right.
Thank you.
Hi, good morning.
Happy to be here.
How are you?
All good, all good.
How you find the event?
Yeah, very exciting event.
We've been here for quite a
few years in these events.
We have a big booth.
We enjoy the industry and the people that
we know our clients
and prospective clients,
some of our competitors.
It's an exciting event.
Happy to be here.
And happy to have you
back on Eye Gaming Daily.
It wasn't with me, it was
with Joe Streeter back then.
But ever since you were with us last
time, OptiMove has launched OptiMove AI.
So I want to get into AI because of
course it's a big topic of conversation.
Everyone is talking AI
here at SBC Summit America.
So you launched OptiMove AI.
So how does that impact your positionless
marketing vision going forward?
And what does it mean for OptiMove?
It's a good question.
And maybe a short refresher on what
positionless marketing is
because I'm not sure everyone
is aware or remember.
It's a simple concept.
It means that over the years technology
in general and
obviously AI in particular is
allowing people to extend their range and
be less in a small box.
So think about historically how companies
in general scaled
their processes, marketing
including.
You establish an assembly line with
multiple workstations and
multiple teams or people
that are specializing in specific tasks.
So if it was a car assembly line, you had
the Henry Ford
situation like 100 years ago.
But even with marketing campaign, you had
the people that if you
wanted to do a campaign,
you had to request assistance from IT or
engineering to help
you with the audience.
They would run some SQL
queries to pull up a list.
So you had to wait for that.
And then once you have the list, you need
to hire an analyst to
design an A-B test experiment.
And you go and you wait for creative to
come up with assets.
And everyone is a
specialist in their role.
But the end result is that you have an
assembly line with multiple
weights and multiple people
that have to work in order to get one
campaign out the door.
And it scales well because now you can do
in theory, you can do more of that.
But in practice, the time from idea to
execution was very long
because you had to wait for
all these five or six different
specialists to give you
their input in order to get one
campaign or one journey out the door.
And especially in the iGaming industry
where things are moving so
fast and player expectations
are always growing.
So we cannot afford having such a long
period from idea to execution.
And we believe that as technology has
evolved and AI is
becoming more and more prominent,
you no longer have to have so many
different teams and
specialists involved in one task.
A person that could never run SQL queries
is still able to
analyze data on their own
because technology
surfaces data insights to them.
And technology allows you to design a
Navy test without
being a PhD in statistics.
And you can analyze campaign performance
without understanding the
math behind it because technology
just does it for you.
And clients that we see and brands that
we see that adopt this
philosophy no longer are
moving faster, but they also change their
own processes and their org structures to
facilitate faster, more agile movement.
And they adopt the technology to just
scale and move faster.
And we have a lot of examples like that.
So let's position this.
The inspiration is from
positionless basketball.
In the past, the tall player had one job,
stand under the
basket, catch a rebound and
dunk it in.
Nobody would expect them
to shoot three pointers.
But in the last several years, we see
tall players, very tall
players like Wambaniyama,
are shooting three
pointers, are dribbling.
And this is positionless.
So same thing with
positionless marketing.
And the recent release of, to your actual
question, the recent
release of OOptima AI
accelerates that.
So marketers, the power of data and the
power of creative and
the power of optimization
so they can do more by themselves.
And the system supports them to do that
so they can move even faster.
They can work from within OptiMove
platform to ask questions
about their customer data,
understand hidden gems in customer
segments, design
campaigns, design audiences, design
experiences, including the creative work
itself within the
brand guidelines, within
the compliance and regulatory
constraints, and just move faster.
And we can probably
talk more about that later.
Yeah, absolutely.
And AI in CRM is often related or often
linked to finding the
right campaign for each player
and personalizing the
campaigns that reach the end user.
So at the same time, operators have
different goals, different objectives.
So can AI, like OptiMove AI specifically,
help arbitrate between objectives and not
just campaigns?
Yeah, that's a very good question.
You know, I think intuitively when you
think about how AI can help us
personalize our customer
experience further, right?
I think the recent, it's funny to say
intuition because it's all
like very new, but the recent
intuition would be, okay, I want to
target my customers or
my players with a specific
context or a specific offer, but I want
to give the AI a bank of
potential options, right?
Different levels of generosity, different
reward types, right?
And I want AI to decide
what's the best offer per player.
That seems to be very
intuitive of how AI can help you.
And that is perfectly correct, right?
And part of the OptiMove AI suite is what
we call offer decisioning.
But here's the thing, offer decisioning,
like the one that I
mentioned, or even content
variations, right?
It works within the context of a specific
campaign or a specific journey, right?
I now want to cross sell
players from sports to casino.
That's the context.
And within that, I can offer multiple
levels of, you know, free
bets, free spins, deposit
matches, each of them with different
levels of generosity and
the content creative changes,
but it's still within the context of one
campaign on one journey.
But what happens when a player is
eligible to different
campaigns or different journeys
at the same time, right?
You have a life cycle journey for risk of
churn and you have
some seasonal campaign,
right?
The World Cup, right?
And maybe someone happens to be, it's
their birthday week or
their birthday month.
And we also want to cross sell whatever
players from sports to casino.
And now all of a sudden, CHI is eligible
to four different campaigns.
But the underlying reasoning behind the
campaigns are
different objectives, right?
If I'm a big sports fan, an operator may
want to leverage my
sports affinity to get
me more and more
engaged with World Cup, right?
So I have, so my World Cup campaign is
trying to engage
customers and players who are more
into sports.
At the same time, the same operator in
general wants to cross
sell from sports to casino,
right?
I might be in a high risk of churn and
there is a life cycle that goes on.
So we have different objectives, save
customers from churning
and usually you're willing to
sacrifice some margins.
So you're willing to be
more generous when you do that.
And I want to exploit
more of my sports affinity.
So I'm more engaged with World Cup.
At the same time, the operator wants to
get me hooked into casino.
Now I'm eligible to all
of these three campaigns.
Each one of them has different offers and
different levels of generosity.
Now how we decide?
And this is where I think to your
question, OptiMove can, AI
can arbitrate between different
objectives, right?
So this is what we
call journey decisioning.
So when a player or a customer is
eligible to multiple
different campaigns or journeys
at the same time, typically those
campaigns or journeys are serving
different objectives.
Then OptiMove AI needs
to arbitrate between that.
So it's no longer within a specific
campaign I'm trying to
maximize casino bet amount because
I'm cross selling.
Because the other campaign is trying to
maximize retention rates in
general and the other campaign
wants to maximize NGR and not GGR.
So how do I arbitrate between all of
these different objectives?
And OptiMove AI in that case looks at
kind of like the
primary metric that determines
the player's lifetime value.
Beat GGR, beat NGR, doesn't matter.
Each operator with their own definition
of what a player lifetime value might be.
And then OptiMove AI can arbitrate within
that and say, okay, shy
is eligible to a cross
sale to a World Cup offer to a risk of
churn their birthday.
What is the campaign today that puts shy
on the path to maximize
lifetime value in general,
regardless of the specific objectives of
each and every campaign?
And then data changes, sports calendar
changes, customer
preferences changes, and we rinse
and repeat and we
personalize for each customer.
Learn how OptiMove's positionless
marketing is changing
how iGaming teams operate.
Discover how operators are using
OptiMove's positionless
marketing platform to launch
personalized CRM campaigns, dynamically
change casino lobbies and
bet slips, and create engaging
game-of-life experiences.
Learn more at OptiMove.com.
And, of course, AI tools have already
made it into the life of
most people and most marketers
actually use already Claude
or ChatTPT into their work.
So how does that change the way iGaming
teams approach or
strategize around campaigns, segments
and personalized player experiences?
Yeah, I think it's funny.
As a product person, for many years, when
you wanted to build a
product or you wanted
to decide how your organization needs to
work, there were always
best practices, right?
The world has figured out how to do
software development 20
years ago with agile and things
like that.
So you didn't have to reinvent the wheel.
It's funny in these times, right?
There are no best practices, right?
Everyone is trying something and every
day someone posts
something on Twitter on LinkedIn,
"Hey, look at this framework I created
and look at the way
that I'm doing things."
And it's really exciting
to be a part of that time.
And we see different
patterns emerging, right?
We see people that they would prefer
working with their own tools, right?
They are already working within Slack or
within Claude or within
ChatTPT in their own life,
right?
They want to meet them where they live.
And on the other hand, we have
specialized products and
tools that have their own AI,
right?
Optimal AI within the optimal platform
that has the benefits
of creating seamless user
experience between AI and point and click
interfaces and an
advantage with the context
the platform can provide.
So I think it's not either or, right?
There are different use cases and each
use case you may choose
the right path for you
as a marketer or as a user to use.
So when we built and released Optimal AI,
we came up with this
idea that says Optimal
AI is inside Optimal.
So you can use our own AI assistants and
AI agents from within our user interface.
They know everything about your data.
They know everything about your campaign.
You build stuff with AI and you can
seamlessly go from the
output of what AI generates to
the more traditional interfaces of, okay,
okay, now let's tweak this with the mouse
and keyboard and point and
click and let's approve it.
And you can go back and forth between AI
and more traditional interfaces.
And at the same time, we have Optimal AI
outside of Optimal where
you can use your own ChatTPT
cloud, Gemini, whatever agent harness of
choice to interact
with Optimal's platform
behind the scene.
So you can ask questions about your data
and get results both
from your own cloud skills
that go to your own
snowflake data warehouse.
And at the same time from the data you
have in Optimal and
cloud in that case would be
able to course reference between them.
And we think this is powerful.
So we've released a new version of our
Optimal MCP server who is
highly, highly capable of
answering questions about your data,
creating audiences, creating
campaigns, creating journeys,
creating content and message templates.
And from there you move into Optimal UI
to approve and tweak, right?
Nothing gets sent
without human in the loop.
And there's a third layer
on top of Optimal, right?
Where now our professional services team
can go and help
clients with customizing very
unique solutions and use cases that
before SaaS companies
like Optimal didn't really,
couldn't really and
didn't really want to do.
You wouldn't customize software for
individual customers, right?
The SaaS industry always frowned upon
things like that, right?
We never do that.
But now we have the infrastructure and we
have the tools for our
teams to go to a client.
They have a very unique use case, a
compliance, a planning
process, some content generation.
And now our teams can help with and build
custom solutions with AI.
So for example, we help one of our
clients create this
planning app that fetches the
sports calendar automatically on a daily
basis, analyzes data
from Optimal, finds the
customers that have the affinity to each
team or each match
automatically, shows it in a
nice custom application front end.
And in the click of a button, you create
an entire marketing
plan inside of Optimal with
dozens of audiences and campaigns that
are tailor made for that
specific sports calendar
of the week.
This is something we
couldn't have done before.
And every client has their own thing, so
it's hard to create a
generic software like that.
But now our teams can do it with the help
of AI and the
infrastructure that we build
to support agentic workflows like that.
So the bottom line is that we are seeing
different patterns of usage emerging.
And we are building our product in a way
that supports all of them.
So you can work inside of Optimal UI with
the context that you have.
You can use your own agent of choice from
outside of Optimal and
interact with our platform.
And you can get help from our teams on
top of everything you
have to customize solutions
for you.
And what has been the feedback from your
partners to Optimal AI?
You mentioned the patterns just now.
So I'm interested in that.
So how does that work?
And how do you work when you get the
feedback from your
partners and you identify those
patterns that they use?
Yeah, it's a great question.
I think there's a lot of excitement in
the world in general
and with our clients and
partners in particular.
When we released this new version of our
MCP server, it
included the option of working
with Optimal Gamify, which is our
loyalty, gamification
and mini games product.
So our MCP server supports our entire
product portfolio, right?
So you can use AI to interact with all
pieces of our product portfolio.
But within less than one day, less than
24 hours after we
released this, we had a client
that used it to build an entirely
customized loyalty front end
that lives in their website
and app within less than one day, right?
Before we even completed
the release notes, right?
We just released the thing and a day
later a client comes to
us, "Hey, look what I built."
An entire webpage, which is not just
demo, it lives in their
actual website now that
shows each player their loyalty points,
their tiers, their
progression across missions and
tournaments and leaderboards and badges.
And it's all customized, fully customized
for their brand and
the way they wanted it
to look like and the
mechanics of their program.
And it's all based on our Optimal AI MCP
server in that case.
And I gave you the example before of how
our teams helped
clients build this app for
sports calendar planning.
And we see a lot of excitement and a lot
of people that are
taking advantage of these
new capabilities, as well as continuing
adoption of features and
capabilities that we had before,
like decisioning agents,
which are not entirely new.
We just keep improving them all the time.
So we have a lot of clients that are
using our journey
decisioning, as I mentioned before,
and are offered decisioning to optimize
their level of generosity.
And as capabilities and machine learning
and AI models are
improving all the time, every
once in a while you get a step function
on the capability and the adoption.
And it's really exciting to see that.
All right.
Yes, sounds exciting.
And of course, in the future, it's also
exciting because this
technology keeps moving forward
and advancing at a
pace that is unbelievable.
So how can AI in the future help iGaming
operators with player
experience and all that?
Look, it's a great question.
And the reality is that the planning and
vision horizon has become very short.
It's really hard.
In the past, we were
planning like a year ahead.
We had a roadmap and a vision for like
two, three years forward.
Now you can plan six months ahead.
And by the time you get to month six,
everything is changing.
And Tropic just released less than two
days ago their new frontier model, Fable.
And with early experimentation, we seek a
step function in its
ability to reason and
its ability to do things that you
couldn't do before, right?
And long-term things, which we didn't
really know was possible
less than 48 hours ago.
So with the pace of progress, it's hard
to predict what will be the future.
But I think it's safe to say that unless
there is a very, very
big change in how these AI
and large language models work, one thing
that they still don't
have and probably won't
have in the near future is
judgment and taste, right?
So AI is really good at getting an order
from you and producing stuff.
It can produce content, it can give you a
summary and analysis.
You can produce audiences
and journeys and campaigns.
But still, all of us are using AI in our
daily basis or most of us.
And you can see that many times you have
to, a human being has to
steer the model to make
good judgment calls, right?
The model can do
things very quickly, right?
Like, "Oh, give me an
analysis with my players."
It will give me like a three-pager within
seconds, like a
three-page of all kinds of
analysis and insights.
But what out of these insights really
matters for my business
is something that AI hasn't
figured out just yet.
So I think what we think is going to
happen and already
starting to happen is marketers
or marketing organizations evolve from
execution of plan, right?
Build the audiences, build the tests,
analyze the reports,
create the journeys, analyze
the journeys day in and day out.
They evolve from that into becoming more
of orchestrators and managers of agents.
I have an AI agent that produces insights
far faster and far
more efficient than the
human being go into
querying some databases.
And I have an AI agent that can produce
dozens and hundreds of
audiences and campaigns within
minutes.
My job as a marketer is to apply the
judgment that says, "What
objectives do I need to make
and what campaigns matter and which
insights actually matter
for my business right now
so I can steer those agents into
producing the right stuff?"
All right, and when I get an email
template produced by AI,
the human has to apply their
taste and say, "This is
almost great, but you know what?
I need to change some layout here.
I need to change some wording there
because my brand in this
region, we want to save this
thing."
It's becoming really good
in following instructions.
And I give it context and it produces
email templates that are almost there.
Right a year ago, they
were like, "Great demoware.
I can produce an HTML email.
Looks great, but it's not really useful."
And AI has evolved and
it's now almost there.
But this last mile is
really hard for AI to cover.
And I think this is the place where
marketers still need to
apply their judgment and taste,
but they have more leverage.
They can move faster and they can
orchestrate agents that do
most of the grunt work for
them.
And we are building our products to
facilitate this idea of
you have all these AI agents,
you need to understand what they're
doing, give them the rules,
give them the instructions,
give them the guardrails, and see the
results, apply judgment,
and use your taste as a human
being to iterate on that.
And this is kind of how we built AI
Decisioning Studio and the
way that we're building the
interface for our AI assistants to help
facilitate the production
of stuff while you still have
to apply your judgment
and taste on top of that.
Cai, it all sounds very exciting, but
unfortunately, we've run out of time.
We can geek out on these
things forever, I know.
Yeah, absolutely.
But we will continue to be looking at
Optimum and Optimum AI
and the evolution of AI in
general.
Thank you very much, Cai Frank, SVP of
Product and GM of Americas for Optimum.
So that's all for today's episode.
Thank you very much, Cai.
Thank you very much for having me.
Yeah, of course.
Thank you very much, Naeem McDonald, for
producing this episode and
the team here at SVC Summit
Americas.
I'm Fernando Nott, and to our listeners
out there, we'll see you in the next one.
Goodbye.