Transforming Customer Service with Empathic AI
Join us for an insightful discussion with Andrei Negrau, the Co-founder and CEO of Siena, as we explore the world of empathic AI in customer service. Discover how Siena's autonomous customer service platform is reshaping the CX landscape, enhancing human interactions, and delivering real solutions to customers.
(upbeat music)
- And welcome to our next session
with Andre to Grow.
Andre, nice to meet you.
How are you doing?
- Nice to meet you, Brian.
I'm doing great.
I'm in sunny or not so sunny New York.
I'm actually in a nice small cave here
and we work and I'm doing great.
How about you?
- I'm doing well, but let me turn the attention
to you and your company.
So you are the co-founder of Sienna.ai
who just raised the CX.
- CX.
- CX.
- Okay, I see that.
- It's CNI.
- Yeah.
- It's Sienna and sorry, I want to get this right.
The product is Sienna AI.
But the website is Sienna.CX
and the company is Sienna.CX.
Gotcha.
Okay.
- Exactly.
- So you guys just raised a round of funding
just recently.
Congrats, that's awesome, exciting.
- Thank you.
Yes, yes we did.
Who are the investors and like,
what was that process like in short form?
I assume.
- Do you want to spend the whole meeting?
- I don't, but I'm just curious on like,
has this kind of been a long time coming
or was this like always the goal here?
- Well, I mean, you know, the goal for us
was from the early days of, you know, starting Sienna
last, you know, summer of 2022, we knew that,
we knew that there was going to be a point in time
when we're customer service is going to look different.
And somehow we made this bet that AI
is going to transform customer service.
And keep in mind this was happening,
summer of 2022 is way before charge GPT came to be.
But we were one of the first companies
to bet on the generative AI even back then.
And so we always view customer service
as almost like the last standing,
like the last standing leg of everything being automated
except customer service.
And so you look across the board and, you know, Shopify
took essentially engineering, like they took
like engineering heavy load in it.
They built a beautiful product that, you know,
anyone can use and can set up a store in like a few minutes.
And you can look across the board from logistics
and different areas of the commerce and business space.
But then customer service was obviously already,
you know, you had chatbots, but it was not something
that people really heavily used for automating.
Like it was not like a common standard.
So we bet on the fact that there's going to be
like a moment in time with CNA starting
where you have intelligent,
like an intelligent and pathic agent
that starts to do some of the tasks that agents are doing
and in some ways are doing better
in some ways they need still to learn.
And so for that, from, you know, we knew that
you do need resource for that.
You do need capital.
Although we were bootstrapped
and we had a little bit of capital
and we had some really great angel investors
in the beginning, we started early conversations
for the suit round.
Actually, in the beginning of the year.
And it was, I would say just to sum it up,
it was an interesting process because
chat GPT came out and everyone, like every investor was,
like they didn't really know what to think
because chat GPT came out and they're like,
yeah, well, why would we invest in CNA?
I went chat GPT this year.
But then on the other hand, they saw the traction
and they saw the results that we were bringing in.
So eventually we found the right partners.
We partnered with Sierra Ventures as a leading investor.
They're based in San Francisco, in Valley.
And, but it was an oversubscribe round.
We have to say no to a bunch of really awesome folks,
which I hope maybe we can say yes in the future.
- That's awesome.
Well, congratulations to you and the team.
That's really exciting.
So timing just worked out.
And again, I will say like, this is just such a great fit.
And I'm so excited that you're here.
I mean, when we started the virtual summit planning,
AI, CX, what's coming next?
Honestly, you were the first company I thought about
of like trying to get you into this.
So I'm glad that you accepted here.
So before we get into it,
can you give us a little bit of backstory on yourself
because you've been in the ecosystem for a little bit.
Is that right?
- Yeah, absolutely.
I've been over the last decade.
I've spent the first part of the decade of my decade in,
I would say in business building.
So I've been building e-commerce businesses myself.
And I was able to sell one of them,
which that led me to have some early capital
that I could deploy towards building software business
in the ecosystem.
But most recently before Sienna,
I built a company called Cartup.
So it was in a similar space
where we were both conversational commerce
for text marketing and a lot of the early insights.
And a lot of the early hypotheses from Sienna
actually stemmed from Cartup.
And Cartup was heavily using humans
as a way to personalize conversations.
And we just, I think it was a great timing for us
and it was a great opportunity for us building Cartup
'cause I think if Cartup didn't exist,
we Sienna wouldn't exist today.
So all those insights, all that knowledge,
we basically just put it back into Sienna.
And I'm a builder at heart.
I'm really passionate about building.
And specifically, I love building things
that solve really hard complex problems.
In Sienna, it's probably the hardest thing.
It's probably the biggest thing that I've ever built.
And it's one of the hardest challenges
that I've ever tackled.
Because when we can segue into what is empathic
and what are we building,
but building a really robust AI agent
that works seamlessly across every single channel.
And it gets the job done and it speaks into voice
that you wanted to speak in.
And basically checks every single checkpoint on your list.
It's really, really hard.
It's much harder than, you know,
most people when you think about AI,
I think it's just chat, JupyT,
and you know, you give it some data
and all of a sudden it can't do this.
So, no, no, this is a whole different beast
because Sienna and RAI is basically embedded in the help desk.
So just imagine everything has to work perfectly
around a clock.
Like the way it changes tickets,
like as simple as changing the tickets,
the tickets status, as simple as tagging.
I'm not even talking about the core AI engine
that has to work flawlessly
and provide all the contextual information for the users.
Everything has to work perfectly.
And that took us a really, like, you know,
it took us about a year to develop,
but it was, it was really intense.
Like it was a period of nonstop feedback and iteration.
And it's been really hard because there's not a straight path.
You know, when you look back into the old days,
when you build a system,
you have systems that are quite easy.
Like, you know, you look at the typical CRM
and you have a set of features that every single CRM has.
Now, when you think about building Sienna
and you think about building for the future,
you have to let go of most of those paradigms.
Whatever has been built before,
it's not gonna apply to what they're building today.
And many of the things that exist today
as paradigms in help desk, we're not even considering,
like we're looking at them in a totally different way
because we believe the future is gonna look different.
And we think that customer support and customer success,
customer, what we call customer experience,
it's gonna have a different meaning in five years.
It's not gonna be the same thing that we look at
through the lines of today.
- Okay.
So let's first take a step back and briefly explain
or just summarize the company for us in the product.
- Sure.
- Who and what is Sienna?
- Sure.
So Sienna AI is an empathic agent that works with
commerce brands, we'd help them automate their customer
service volume all across channels.
So think about it as an agent that you train once
and once you do that, it's able to handle questions,
answer, I mean, answer questions, take actions
and then route the conversation at the right time
when that's needed.
So the best way to think about it is like an agent
that's embedded in your help desk.
And it has all the right tools and it has all the right
knowledge to provide the perfect response
on the right channel at the right time.
- So it's not a replacement for the help desk
and I'm just clarifying for the audience.
But it is a seat in AI seat, honestly,
that can be used in any, well, actually not any,
you integrate with who right now?
- So we integrate with all the top help desks.
So we integrate with customer, we integrate with Zendesk,
we integrate with Cordjose, we integrate with some other
help desks, yeah.
And Sienna indeed, it takes a seat and it works
as a seat, it works alongside a team.
It does not replace help desks.
We're 110% focused on building gag products
that drive automation and drive retention.
And we don't have any plans on building
like help desk systems and ticketing systems.
- So you talked about, I went through the website,
I hope everyone sees this as well too.
You talk about AMP-happec AI, describe that for us.
- AMP-happec AI goes beyond the, let's call it
like the front facing empathy for the user, right?
That's almost natural, it's what Sienna has been doing
for a long time, meaning it connects with the user
and understands the broader context of a conversation.
That's just the first pillar, which is obvious,
that is happening, it's obvious that we're doing that.
But AMP-happec AI, it also looks at the broader,
like at the broader customer service landscape,
it includes all their stakeholders.
So when we say AMP-happec AI,
we also account for the agents that are involved
in the process, we also account for the business,
that's operating that the customer service department.
And we're also accounting for the team
that's managing the AI tool.
Empathy meaning you have to connect
with all those stakeholders.
And you have to build, you have to design a platform
that seamlessly integrates with all of those stakeholders.
It's not just simply responding very,
in a very personalized way that it's nice and fluffy
and like warm with the customer,
but it also connects with the broader team.
Because ultimately, what I believe,
where we're gonna end up in the next few years,
is you'll have AI at the core,
but then you're gonna have teams of people managing the AI,
and you're gonna have teams that are gonna be able
to derive insights from the AI.
So we're shifting from a place where AI is,
or from a place where customer service
is just the fact of sending a message closing it tickets,
we're actually getting into a place
where customer service is gonna be insights,
it's gonna be automation efficiency,
and it's gonna drive business growth.
Like for the first time, you're gonna have it,
even with CNM, I'm having to provide some insights.
We look at the stats, track through help tests,
for example, and we see the AI driving meaningful revenue
for our customers, right?
And all that to say that EmpathyKI encapsulates
the relationship between all these stakeholders.
That's what we're optimizing for.
And obviously the first stakeholder that we care about
is the consumer, how do they feel?
Because from one of our core thesis is like,
if we're doing a great job with the consumer,
and the consumer is happy, and we solve their problem,
the business is gonna be happy,
and the business is gonna grow,
and down the chain is gonna be happy.
So that's our first pillar.
But then it's a broader conversation,
like how do you go from just being empathic
with the consumer, which I think everyone,
when you think about EmpathyK,
everyone thinks about, yep,
providing the right answer at their right time,
at the right messaging and whatnot.
But then you have to go downstream
to actually think about how do you translate that
into a touch points with the rest of the stakeholders.
So it's almost like, just to sum up,
Empathic AI, it's almost like the method that we follow,
when it comes to building the AI ecosystem
that we have set up to build,
because it's much more than just purely
the fact of automating a conversation.
That's just almost like a commodity.
You just send a message that that's been moderated.
- Has that always been,
I mean, it just sounds to me that you are,
like you're starting with Empathy,
like from the very, very beginning,
and you're building your product around that,
has it always been that case?
And was that taken from your previous experience?
Like how do you know that Empathy should be our first
principle and we'll build around that
when it comes to customer service?
- You know, when we started Siena,
all of the founding team had prior experience with e-commerce.
And I think all of us were really,
like we had a really good sounding board,
knowing that if we're gonna build this,
the system, the AI that we're about to build,
has to be orders that then meant to better
than anything else that was built before.
And for us, if we couldn't pass that bar,
we might as well not build anything
because there was already a million,
or like a gazillion different tools
that enabled some level of automation.
For us, the litmus test is,
if I have a conversation with the system,
would I enjoy it?
Would I feel like at least I'm listened to and I'm heard?
And obviously, it is my problem being resolved
because if you couldn't check those initial boxes,
we might as well not do anything.
So for us, it was important for the system
to understand you and your blog,
like just think about what was the core problem
with all the previous technology?
It's like, you just talk through something
and you know it's just kind of respond with it,
kind of respond.
It doesn't listen to you,
it doesn't really understand what you say,
it's just some sort of keywords that are being triggered
and you get a response back.
And that's why people really like never really like,
like consumers never really liked
and businesses never really saw that much value
in the previous technology because it's almost like
you talked to a wall for us.
It was important to build something that if you,
let's say you're sending an email
and you get a response back for us,
the litmus test from the beginning was,
do you know that, would you be able to tell
that's an AI or that's a human?
And if you'd be able to tell that's an AI,
that's a problem.
It means that we didn't do our job.
And so that was like a non-negotiable for us.
It was really, really important for this to feel real
and to feel like you're talking to someone
that listens and cares because, you know,
at the end of the day, people when they reach out
for whatever reason, be it for guidance using a product
or they wanna cancel like an order
or they wanna make amendments or something,
they have a problem.
And if they feel like they're talking to a wall,
they're only, the problem is only gonna get de-saturated.
It's not gonna be like, oh yeah, you know,
I got a message from this random bot
that doesn't really understand what I'm saying.
It's just gonna want even more to talk to a human like,
okay, I'm really angry now.
Like help me here connect with someone.
So we want it even though in the beginning,
Xena was not that capable.
'Cause we, you know, it had its limitations
in like the first few months.
It's still what was incredibly,
incredibly insightful for us to see
is that even the first touch points
where it might have said something like, okay, hey,
you know, hey, Brian, I hear that you wanna,
you wanna get your order shipped faster.
Okay, I'm not able to do this,
but let me connect with my team,
which is gonna get you in the next 24 hours.
Even that would drive tremendous value.
So again, it didn't fully resolve a conversation
at that point, but the fact that they see that, okay,
I'm getting a message from us, you know, almost like a human.
It gives me that the sense of trust that, okay,
I'm being heard.
I know that someone has actually reviewed my message
and it feels much better than sending an email,
not getting a response.
I feel like not even like 24 hours.
So even that was a big insight for us in the beginning
because it just tells us that the empathy in this
plays a much bigger impact than we initially thought.
- So just to clarify, like,
from the consumer point of view,
when dealing with your product,
like they truly think that they were dealing with a human
and that's the goal from your side, or is it not?
Or is it up to the brand really to decide that?
And I just kind of wanna push you on this.
Like from a consumer perspective,
do you want it to be that they're still thinking
that they're talking to a human,
but it's AI behind the scenes?
Or is it okay to get to the point that they're interacting
with an AI bot and the consumer knows that,
but it's still a great experience
and empathetic experience?
- The way, so the way the majority of our customers
have currently set up,
DNR is that they don't necessarily advertise
the fact that it's an AI.
Like it is not necessarily saying,
hey, I'm a CNI AI, I'm like an AI thing.
But if someone asks, are you a bot?
It's gonna say, hey, I'm an intelligent agent,
I'm gonna help you with your problem.
So that's what you can expect.
- Yeah.
- So that being said, there are customers
who given their branding and given their overall
brand positioning and messaging,
they might wanna say upfront that this is an AI
that has responded to a specific query.
So that's perfectly, like that's something
that you can do from just from the settings,
you can add like a signature or something.
- Yeah.
- So that's possible.
But for the most part,
some of the benefits of using AI,
some of the benefits of using CNI is that it's so
and having that, you know,
we've automated millions of conversations.
And probably the number of conversations
where people really felt like there's just my BNAI,
it's probably like below 0.1%.
So it's like incredibly small.
- That's amazing.
- Yeah.
- Now you talk about solving so many tickets already
and the NSCANA solving complex problems.
Is there anything atop of mind,
and again, it started to just like put you on the spot here,
any complex problems that you've seen,
your platform solve that you can share with us,
that's like, wow, that's quite amazing
that they solve something that complex.
- You know, obviously complexity is subjective.
What do you call subjective?
I mean, for an agent, most of the queries are,
and at the end of the day, rather basic.
Like let's say we're talking about a refund, right?
You analyze, you look at the policy,
and you compute like, okay,
I should probably give them a refund.
And that's like for a human, even that is pretty basic.
As long as you know what you need to do, right?
For a machine, for an AI,
that becomes pretty complex.
That whole process of each one refund
can become pretty complex because,
well, let's just take a step back and understand
what actually happens when someone sends you an email,
like, hey, I need a refund, right?
Well, first of all, did they provide any reason for that?
Like, just imagine people might say,
hey, I just need a refund, right?
You need to understand what's the reason for that?
You know, okay, then once you know the reason,
you need to look at, okay, what's the order?
Like, okay, what's the reason?
Is this reason eligible for a refund?
Is it something that we offer refunds for?
If not, then you have to have a counter strategy
to tell them, well, what can do that?
But here's another alternative,
like I'm gonna give you a discount, right?
Well, let's say the topic or the criteria, it is eligible,
then you have to understand,
let's take a look at the order.
Is the entire order subject to providing a refund
or it's only part of, it's like a partial refund
that we need to issue?
Do we have a receipt?
Like, for example, if someone says,
I don't have a receipt, sorry,
would we have evidence for the fact
that someone claims the order never arrived
or the product is damaged on arrival
or something like that?
There's all those things.
And now, again, as an agent, it's quite easy
'cause you can just say, yeah, send me an image.
I wanna see if the product is actually damaged
and they send you this image
and then you take the right action.
The challenge here is for the business to sit down
and to find what is the process.
This is one of the challenges that we've seen
is most of the times,
unless we're talking about 100 plus agent teams,
most of the mid-market agents,
teams of customer success teams,
they don't have that much of a structure
around how to deal with these processes.
So then what they have to do is,
they have to sit down and understand
what is the process for dealing with refunds
on the grant stream, I think.
And then having all that,
Sienna is able to take that
and he's able to perform, like apply reasoning,
what we call our core engines, our core AI engine,
he's applying reasoning,
and it takes an account every single thing,
including also the customer sentiment, right?
Because sometimes you wanna take a look at,
hey, if this customer is incredibly angry
and if they're high value customer,
I know that there's a policy for no refunds.
We all know, but every customer support team
knows that if we're talking about a high value customer
that's really really angry,
you might just wanna say,
you know what, just give them a refund, right?
All those things, the beauty of the end is that
you can literally put all those data points.
Sienna takes all that information
and it's a fine reasoning to determine
the next set of actions.
So it might determine in real time
that, okay, for this particular customer,
given the fact that they have a lifetime value of,
you know, $20,000 and they've been with us for like 10 years,
I'm gonna go ahead and provide everything.
So that's where the complexity is,
but everything is happening in real time.
So I'm not sure if I answer your question,
what kind of complex problem is it solving,
but it's something that wasn't possible before.
There was no way, ship a forum,
when you're thinking about the kind of the branching tree
systems where you define those flows.
Because of the complexity and the nuances
in customer service, there was no way, ship a forum
that you could define in the design system like this.
Now you can't wait Sienna.
And that's why customers,
I mean, both consumers and businesses
are absolutely loving it because it gives you both
the structure that you need to put in systems.
But also you can essentially tell,
hey, here's an exception,
this is something that you can do as an exception, right?
And it's so amazing because this all basically
plays out in an infinite scale.
Like you don't really know ahead of time
what kind of customers are gonna reach out.
You don't know, but you know that if it's a really
high value customer, you wanna be in a position
to give them a refund and not tell them,
I'm sorry, we don't have this policy and whatnot.
So it's really important.
And the forefront, like at the core,
we integrate deeply with all the main tools
that a business, specifically now we're into
retail and commerce based needs.
So we're gonna think of it with like CRM's
order management system, subscription platform.
So Sienna has all that knowledge.
So Sienna knows, okay, when was the last order placed?
What is the total value of orders?
Oh, so it knows everything.
And that's why it's able to take those actions.
- And from just really briefly, from a high level,
most likely you're connecting, obviously if you have,
let's say a CRM, a help desk, something like customer,
you're connecting on Clavio and recharge to that.
In this case, are you also then connecting recharge
to Sienna or is Sienna already connected to?
- Okay, okay.
So you're connecting these right to Sienna as well too.
- Exactly, yeah, exactly.
So on one hand, we integrate with the CRM with the help desk
and that's the main channel.
Like that's where it has to just flow in and flow out.
And then on the other hand, we have the data platform
that we're doing.
So that's where the data resides.
That's how we basically make use of the customer data
and the business data.
Obviously here, it's important to know that,
you know, from a business standpoint,
you do wanna connect, like you do wanna have
as many connectors as possible.
You mentioned like even Clavio, right?
For example, you do want the AI to know that,
hey, this customer, they received an SMS campaign 24 hours
ago that, you know, mentioned this, this and that.
And guess what, that information,
it's gonna be invaluable for the AI because,
wait, you know, if you don't connect Clavio,
if you don't have that information available,
then someone might say, hey, just imagine like getting
a ticket which says something kind of like,
seeing when you're running like, oh, this doesn't work for me.
And you're like, if the system is not really integrated,
if the AI system is not deeply integrated,
it's gonna be like, okay, what is not working?
You don't know what's going on.
- Exactly.
- But because you have the connection and you know
what's the last campaign that was triggered
in Clavio and all that, then you'd cycle 360
the preview which is what's currently important.
- Yeah.
Now, I want to kind of move the conversation to outcomes
because I think that what you've shared
across your website and myself,
that there's been a lot of success stories
that have been just amazing so far
in such a little short of time.
Can you share some of those with us?
- Yeah, absolutely.
So, you know, every single company defines customer service,
success, differently so everyone has different indicators
of success.
We look at three things, mainly at Sienna.
We look at automation rate, we look at router rate,
how many of the conversations were routed to the team
and what is the CSAT score?
What is the actual quality of those conversations?
Without, you know, without giving names,
you can obviously just check it out on Sienna.Cx
and we have a great customer section
where you see case that is and all the numbers
and everything.
We have customers ranging from 50%,
we're half customers that are ranging
in the automation rate from 50% to 80%.
Even we have customers that surpassed 80%.
I think the best one was like 92% automation rate.
Those are some incredible success stories
because it just tells us that with those customers,
they started using Sienna from a place
of always firefighting and always just doing the same.
Like you have to just have to run on their hamster wheel,
like just answer data and sort of take it,
but there's no real outcome besides that.
Now what they're in a position to do now
is basically take a step back and date,
just redesign their whole Sienna department.
Like they're looking at how can they not just do support,
customer support, but how can we increase
the customer experience overall?
And the impact is like, is absolutely like tremendous.
We're talking about companies that are currently,
if previously they had to maintain
and they had to answer 100% of their take as now,
they only have to worry about 20% to 30% of those.
That's a dramatic difference.
It's almost like overnight you have 70% of traffic
and you can take that time to think about
how do I improve the overall customer experience
for my team.
So that's only on the impact on the business side.
That's just right on the business side.
Now when you look at the customer facing impact,
one of the big things that we saw
and we're looking at ways to track this better,
but it's a huge increase in CSAT.
And when I say CSAT, I'm speaking about reviews.
So we're looking at companies that leverage Sienna.
And again, this is public information.
You can look it up on our website
and then just look for reviews.
And you'll see that consistently,
they're able to garner really good reviews
because of the impact CNI is providing
within their customer service department.
So if previously you would see reviews like,
"Oh, yeah, this is not a good company
"because I send them an email, they couldn't help me
"and I had an issue and no one got back to me."
Now you have reviews that say, "Wow,
"I just sent them an email and it was immediately resolved."
And it's just incredible to see that
because it doesn't get more tangible than that.
Ultimately, I think business efficiency is great.
Like everyone wants to improve their efficiency
and customer service leaders.
They look at reducing costs and all that.
But I think the greater impact is far more than that.
I think it's the actual consumer impact
that just completely changes the way
consumer interacts with the business.
And again, some things are subjective, like a trust.
Like you cannot really measure a trust.
I'm not able as a CEO of CNI to tell you,
we've increased the trust of the customer.
Like over the consumer, that would be a stretch.
But we do see that customers are way more empathic
on their own coming back to the business
and the kind of conversations they have
because guess what?
Like who doesn't love to work with a business
or to buy from a business that gets back to you in time,
mean like seconds or like minutes
as opposed to like waiting one, two, three days.
And I just wanna give you,
I shared this example 'cause it's like so trivial,
but like it has such a huge impact.
I don't know if it ever happened to you,
but even myself recently, I placed an order with a company.
And for whatever reason, I paid with Apple Pay
and it was just, it was the wrong shipping address.
So you know, like Apple Pay has that shipping address
and they wanted a different one.
And I reached out to them and I, hey guys,
I just made an error.
No one reached out to me, obviously.
Like I waited like two days
and the order was shipped to the wrong address, right?
And then I had to go through like this whole process
like multiple days of like showing them like,
hey, I don't live here.
And like, and it was like multiple days of back and forth
for like a very trivial thing, just pay in.
I just need to change my address, is this right?
That's something that Ciena's currently doing
like without any problem.
So for us, having, sorry, for our customers,
having the possibility to set up Ciena
in a way that catches all those emails in real time.
So every time someone places an order
and it's, for example, the wrong email,
the wrong shipping address, Ciena makes amendments
in real time.
So you can, you know, if it doesn't have the address,
it's gonna ask the customer, hey, where do you want,
you know, where do you want to have it shipped?
And it's gonna make, it's gonna make that action
both in the CR RAM, it's gonna make it into like a orders
platform, it's gonna make it in the subscription platform.
And that's a very, again, it's a trivial example,
but the beauty it's like it's happening in real time.
So the customer doesn't have to wait over the weekend
if it's on Friday, like it's happening right then and there.
And because most of our customers use like third party,
like, Pokemon centers,
they order to get some end of just in time.
So it reaches the, it reaches the Pokemon center.
So that's where the power of AI,
I mean, it doesn't really matter like
how many agents you have on your team.
It's really hard to beat that
just because it's real time, it's 24/7.
And that's just a simple example of this,
something that I've been through recently.
And I was like, man, I'm just wasting hours of my life
with just something as trivial as that.
Now it's Ciena, you don't have to worry about it.
- Yeah.
So let's end on one thing.
I wanna give a takeaway to the audience members.
I think what would be really interesting is your take on,
let's say, a customer experience,
customer support, managers of teams of agents, right?
If they actually want to take that step
to start experimentation with AI agents,
what's the narrative that you would propose to them?
Like how should they be talking to their upper management
VPCOs on why they should take these next steps
to kind of experiment with the AI agent
with their current support?
- Yeah.
Yeah.
I think, you know, as any business decision,
everything has to stem from a need
or from a desire to do something differently or better
or improve a system or something.
I think the biggest pitfall here,
I'm just gonna like, never for a second,
is just like trying to do AI,
like implementing AI just because it's AI.
I don't think that's the best approach.
My approach would be try to sit down, understand
what needs to be improved.
Is it, you know, are we trying to improve efficiency?
Are we trying to improve the over experience
that we provide for our customers?
What is exactly, what are the top two pain points
that we're trying to solve here?
AI is just a technology.
AI should always be applied towards the business,
the broader business goals.
That's what I tell everyone.
Once you define that, I would assign,
obviously, you know, that all the tools out there
in the market, that's really important to understand
if that's the best approach.
It's important to understand if the tools
that you're assessing match the goals
that you have set up for that initial project,
be it like, you know, getting efficiency or CSAT
or just overall better experience, you know,
significantly reducing, waiting times and all that.
And then the third point is assigning an owner
or like a directly responsible individual with KPIs.
So just the way, you know, these days
with every single, like for example, in marketing teams,
you have a person that's owning email marketing channel
or like maybe you have a team, but you typically have someone
who at the end of the day will look and say,
okay, this month we've generated $5 million through email
like as a channel.
That's a KPI, that's a KPI
that is known by the whole company, it's tracked by everyone.
I would apply the same, and this is what I see,
what we've seen great success with some of our customers,
where they apply the same mindset for leveraging AI.
So what they do, they set hard KPIs around AI performance.
So they would do something like, okay, by Q2,
we want to be at 50% automation rate,
we want to be at below 10% handoff or router rate,
and we want to be at a CSAT score of X on a scale of one to five,
with CNA or with AI.
That is simply, that is the process that I would follow.
So first assess, understand what are the top three priorities
that you would want to transform
or you want to add transformation into your team.
Second, obviously find a solution that can fit
with then can check all those boxes.
And third, put those KPIs behind the initiative
and have someone own those KPIs.
That would be it.
- That's awesome.
Well again, Andre, thank you so much for your time.
This has been really quick, 30 minutes,
but I could have lasted much longer.
I wish you all the best.
And I think the audience here, please check out CNA.
They're doing really incredible things.
- CNA, that's CX guys.
Thank you.
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)