Human-First AI to Eliminate Blind Spots, Remove Grunt Work, and Drive Revenue
In this session, we’ll dive into the transformative power of Human-First AI that works for you - illuminating human blind spots, automating the grunt work that bogs down your team, and ultimately leading CS into the new age. You’ll learn how AI can provide real-time insights into customer behavior, streamline repetitive tasks, and empower your team to focus on strategic initiatives that truly impact the bottom line. Join us to explore how adopting a Human-First AI approach can enhance customer relationships, boost efficiency, and unlock new revenue opportunities for your business.
All right, good morning everyone.
Welcome to Pulse Europe 2024.
I hope you all fired up and excited to be here.
This is Track 5, human-first AI in action,
and we have some incredible speakers lined up for you today.
My name is Harry.
I'm part of the Customer Success Team here in Europe.
A couple of reminders.
There is a Q&A at the end of this session in Slido.
So if you can drop your questions for our speaker in the Pulse Europe app
by going to the polls and Q&A area on the homepage, clicking Track 5,
and then we'll take your questions at the end of the session.
Just in case you are wondering, you'll get a copy of the slides
and the audio recording from today.
You can share it with your colleagues.
That would be the Pulse Library.
So don't feel the need to take screenshots and photos throughout.
Right, I'm really excited to introduce our first speaker of the day,
against the OG Tori Jeffcoat, who leads our marketing strategy
for our CS and AI products at Gainsight.
A little fun fact about Tori before we start.
She grew up in Singapore and she's super excited about the Dance Dance
Revolution machine
that we've got here.
You might have seen it out in the hall.
She used to play it all the time, and actually she's pretty good.
So if you fancy a competition with her later, then go check it out.
Tori session is entitled Human First AI to Eliminate Blind spots,
remove grunt work and drive revenue.
So in Tori session, we'll dive into the transformative power of Human First AI
and how it can work for you.
Illuminating Human Blind spots, automating the grunt work
that box you and your team down, and ultimately leading CS into the new age.
You'll learn how AI can provide real-time insights into customer behavior,
streamline repetitive tasks, and empower your team to focus on strategic
initiatives
that truly impacts the bottom line.
We'll explore how adopting a human-first AI approach can enhance customer
relationships
and empower your team to boost efficiency and unlock new revenue opportunities
for you and for your business.
So please give it up for Tori.
[APPLAUSE]
Sorry, it's this week.
[APPLAUSE]
Thank you.
Awesome.
Tori, and welcome everyone.
I am so excited to be the first presenter of the first session on the first
track
that you attend today.
So hopefully this sets a great tone for the rest of the polls.
Super curious how many of you attended our keynote this morning.
Everyone enjoy the keynote, hopefully?
Yes.
Lots of awesome stuff.
A lot of what I'll talk about in today's presentation
dives a little bit deeper into some of the AI that we talked about on stage
with
Ori and Georgia.
So hopefully give you a really good overview, but kind of take it a level
further
to what we covered in the keynote as well.
Also curious, and I know Nick asked this question on stage, but quick show of
hands,
how many of you this is your first poll event?
That's awesome.
That's so many great hands in the room.
How many of you have been to polls before?
Either in Amsterdam or otherwise.
Awesome.
OK.
So this is actually my third poll since my second time speaking with Gainsite.
Super excited to be here.
I think poll says one of my favorite, favorite events that we do every year.
And polls here up I think is a little bit more exciting.
Sorry to the US one.
But super excited to be here and chatting with you all today.
So to kick things off, just a quick overview of today's session, I want to talk
a little
bit about why AI is so popular, breakfast pun club there from this morning's
keynote.
By taking a quick look back at AI over the last year, talk a little bit about
where we
got it right and where we got it wrong.
Spend some time talking about a couple of key ways that AI is reimagining
customer success
and helping us solve some of the challenges that we faced for a decade and just
haven't
really been able to tackle fully yet.
And then end with a couple quick tips on how to think about AI for customer
success.
So I want to start by taking a quick look back.
Not as far back as all of our 80s references in the keynote this morning, but
just a look
back at the last year and AI and how it's kind of grown and evolved and
basically exploded
on the scene in the last 12 months.
So when AI first came out, we all probably felt a little bit like this GIF,
right?
A little bit confused, a little bit behind, a little bit maybe faking it till
we make
it, if you will, with AI.
And it's really hard to, one, keep up with everything that was happening with
AI over
the last year, but also talk really intelligently and deeply about it,
especially if you get
really technical and start bringing in terms like RAG and fuzzy logic and all
the different
LLMs that exist today.
And AI did not make it easy for us to learn either with OpenAI and all the
different chat
JPT models they released with Salesforce and their agent force announcements at
Dreamforce,
lots of forces there, and Microsoft's Co-Pilot.
It's been really challenging for us to not only learn, but also keep up with AI
over
the last year.
So it's probably no surprise that with everything changing and growing with AI,
it's been really
hard for us to learn and actually adopt AI within our CS programs.
This data is from our state of AI and CS report index.
We ran this a little bit earlier this year, and we found that of our European
respondents,
only about 47% of CS teams are actually leveraging AI in their day-to-day CS
workflows.
So caveat here, that's AI within CS workflows, so that doesn't count things
like you personally
writing an email and chat JPT, which I do all the time.
But using AI prescriptively for things like success plans for EBRs or quarterly
business
reviews, things of that nature.
We also found that of those that were leveraging AI kind of across the board
when you look
at revenue per company, it was either startups, which are probably a bit more
agile, a little
bit more or less red tape for them to kind of adopt AI, or enterprises where
they had
the resources to kind of invest in that R&D that were typically the first adop
ters of
AI over the last year.
Quick note as well that 47% for Europe was just for the European respondents.
It was 52% globally, so not too far off, but just a couple higher percentage
points there.
So I'm curious based on that data, a quick show of hands, how many of you are
leveraging
AI personally today?
So things like writing that email with chat JPT.
Awesome, that's a good number of hands.
Hopefully mostly AI adopters, given that we're all here on our AI track today,
but if you're
still exploring AI, totally okay.
Follow up question, again, kind of based on our data, curious for this room,
how many
of you are using AI within your CS orgs today?
Okay, that's probably about accurate, right?
Maybe like 40% of folks raising a hand.
I have one more question for you.
How many of you have tried but stopped using AI, either in your CS org or
personally, for
specific use cases too?
So I've used it for some things, but you stopped using it for others.
Curious for a quick show of hands.
All right, we got a few, couple, three, four percent.
I'm not good with math based on hands in a room.
But we have a couple of us that have tried to explore AI and found it either
didn't solve
our problem or it just wasn't really ready for the needs that we maybe had for
it.
And this is actually a trend we're seeing globally.
So that 52% number of CS teams leveraging AI from our state of AI and CS report
, we just
finished our CS index report, not to be confusing, two separate reports.
But this is more of a global report on things like CSM ratios, digital CS, and
we also asked
a couple of AI questions.
And one of the questions was that same one, how many of you are using AI in
your CS programs?
And we actually found that the number globally dropped.
So in April of this year, it was about the 52%.
As of last month of October, it was only 40, I can't remember my slide, 44% of
CS teams.
This was a surprise for me.
I expected that number to go up, right, over the last year, that generally what
I think
most folks would expect.
But as we dove a little bit deeper into the numbers and into the data, it
actually made
a lot of sense.
Because a lot of early adopters of CS and AI were actually missing true value,
right?
We were using AI just because it was the shiny, new, cool thing.
It wasn't necessarily driving results for us.
We also found that a lot of CS use cases weren't really well established.
So leveraging AI for that email and chat GPT, maybe that's still a really
valuable use
case today, and at least helps me to get 70% maybe done with something rather
than spending
that time myself.
But for CS use cases, for our needs, for early warning signals, for insights,
we hadn't quite
gotten there.
And then the last key point, security and privacy hurdles.
I know of several customers who are really excited about AI, but maybe they
have to get
through a lot of that red tape, right, or, you know, legal is a little bit
concerned about
leveraging AI for all of those data privacy reasons, which are really valid
reasons.
But it's also resulted in a couple of AI projects kind of getting shut down or
paused
internally.
So again, numbers I thought at first were really surprising, but actually kind
of makes
sense when you take a dive deeper into it.
And a lot of those challenges are ones that we actually found are the top
barriers to
AI and customer success.
So this is from our state of AI and CS report.
Again, sorry to be confusing with two separate reports.
But from April of this year, we found that the number one concern or barrier to
AI adoption
was lack of internal expertise.
Again, probably not a surprise to anyone in this room.
The next highest was integration complexity, data privacy concerns, output
reliability,
and then resistance from teams.
This is another question we asked again in that CS index report.
And I think it's really interesting to see where these numbers have trended.
So if you compare the two reports in our CS index report, we actually found
that lack
of internal expertise has dropped pretty significantly.
Again, over the last year, we've all been learning exploring AI, so probably
not a surprise
in that drop.
But what's super interesting is that integration complexities jumped up like 20
% or so over
the course of the last five months.
A lot of that kind of goes back to folks who've tried it and maybe found it
doesn't quite
100% work the way they need it to or doesn't integrate with systems.
So such a key call out for leveraging AI effectively for customer success.
Interesting enough, data privacy and output reliability were the exact same
percentage,
like not even 1% different, but does indicate that those are consistent ongoing
challenges.
And then resistance from teams has started to drop as well.
Again, probably no surprise, but just shows the trend of we are getting more
comfortable.
We're still encountering technical roadblocks.
Another super interesting fact is that data privacy was actually significantly
lower, like
10% lower as a concern for European respondents than it was for the US.
So I don't know if you guys are all just more trusting or if that's another
factor there.
But super interesting data point.
So all of this kind of shows that we do have some hurdles to leveraging AI
effectively.
We're still learning, we're still understanding how we can use it.
But the truth is we really, really need AI to solve some of those ongoing
challenges that
we've been facing.
We needed to hard our jobs again.
I know this was a slide that was in the keynote as well.
But AI does a really good job at doing all the stuff we really don't like
anyway.
The data entry, the opportunity to analyze data, spend time looking at spread
sheets, spend
time looking at emails, right?
Things we probably don't enjoy unless you just really, really like spreadsheets
So taking a lot of that stuff off our plate, so we can be more strategic.
So we can build better relationships and be a better partner to our customers.
We also need AI to tackle some of those, again, challenges and use cases that
we haven't been
able to solve.
Things like identifying at risk customers, being able to prep for meetings
without us having
to spend again all that time and grunt work kind of doing that.
Being able to plan our workloads better and engage our customers better at
scale as well
with the same personalization and expectation of a good customer experience.
Because ultimately we need AI to stop the diving saves, which might be very
exciting
in things like soccer, football, football, but less exciting when it comes to
customer
success and saving our customers from turn risk.
So during today's talk, I want to unpack some of those use cases a little bit
deeper, how
we can better understand and capture early warning signals and insights for our
customers,
how we can remove a lot of that low value grunt work, and then ultimately how
we can
create more strategic business level opportunities to drive revenue and save
revenue with AI.
So let's start by talking about better early warning signals.
Again, hopefully you all caught the keynote and learned a little bit about
staircase and
some of what it brings to the table, but I want to unpack that a little bit
more and
what that really means for customer success.
There we go.
Okay.
Any of these phrases sound familiar?
Not aware of a support ticket and a customer is super upset, not able to
understand how
your teams are doing without either micromanaging them or making maybe risk
assumptions, or
not knowing when a customer is about to turn and having to explain that to your
CEO.
Anyone familiar with any of these phrases?
Experienced sending them in the course of your career?
A lot of these are the things that we really can't solve when we don't have the
right signals.
I can get my clicker to work.
There we go.
A lot of these are things we really can't solve if we don't have all those
signals under
the hood, right, under the surface, all the things that happen in each
communication,
each support ticket, each conversation with our teams, even on chat, right,
things like
people using Slack, internal Slack channels with customers.
It's really, really hard to capture all of that data when you aren't leveraging
something
like AI and can look at all of those signals at scale.
Okay.
I think I clicked too fast.
I think that's it.
So I want to talk a little bit deeper about real time always on sentiment.
Or you talked in the keynote about how staircase captures sentiment
automatically from communications,
but I want to talk a little bit about why it's so important that it captures it
instantly
at the time that that sentiment is happening.
If you're a CSM and you're logging these engagements, right, there's probably a
delay between when
the conversation happens and when it gets into your system.
There's also challenges with how much you log, how frequently you log it, right
, we're
all human.
I definitely forget emails on occasion.
I'm sure our CSMs do as well.
But when AI can capture all of these things and provide that real time trend of
sentiment
over time, it helps us to really, really get a constant, continuous, and solid
signal about
our customers.
It's a little bit like the difference between getting a daily checkup, right,
if you take
your pulse every morning versus going to a doctor for a once annual visit.
When we rely on things like NPS, when we rely on things like a single meeting
with our
customers that happens every six months, it's a lot more like that doctor's
visit, right,
a full checkup, versus getting our vitals on that ongoing regular cadence to
know how
we're trending.
And without AI, this is really hard for us to do at scale.
The other element of what staircase brings to the table and what AI, when it
can analyze
these signals, provides us, is that unbiased customer health.
When CS started, health scores were mostly manual, right?
CSMs are inputting how they felt a customer account was.
And maybe you're really, really positive and think, "All of your accounts are
amazing.
You're super pessimistic and think everyone is a turn risk."
So it's really hard for us to get that unbiased sentiment when we're leveraging
things like
manual health scores or health scores based on systems that may not be
optimized for our
true customer needs.
And so what AI helps us do is to really unpack all those signals without bias
based on an
algorithm that understands not only where your customers are, where they are in
comparison
to other similar customers in a similar segment.
So it helps us to analyze and understand those risk signals based on those
comparisons to
other customers, which is also really important for AI to understand our
accounts and our
customer needs for our particular business model and our particular product
versus kind
of a generic use case.
And then the last thing I want to highlight is the multi-threaded relationships
Ori talked on stage about how this helps you to understand who is talking to
your customers,
right?
That relationship radar that he showed.
What's really important about AI providing us with this relationship score and
understanding
of strength is that it helps you to know who has really strong relationships,
who are kind
of the key linchpins to each conversation with your customers, and where are
the risks
that if a CSM leaves, right?
If a customer maybe has a one key contact who's out of office for three months
on vacation,
right?
Are you going to be able to have a strong relationship with that account to
carry it forward?
And so AI helps you to understand every single touch point and the strength of
those relationships.
It does that based on understanding how quick someone is to respond, how long
their delays
are between conversations.
One really cool thing about staircase chatting with a staircase team as they
became part
of GainSight is that it'll actually track trends over time as well.
So something a human would never be able to capture, right?
Like say someone takes a day longer every time they respond to you.
Over time with a number of customers that you have to manage, that's probably
something
a CSM would never realize, right?
Never capture, never understand.
But with AI tracking kind of that over time cadence, it's really easy to
understand when
things are starting to drop off and when you're starting to see those early
risk signals.
So ultimately, when you're able to really understand and capture all those
signals at
scale, when AI can help you to analyze all of that content, you can be a better
listener
to your customers, right?
With zero data entry, which is a massive, massive part of it.
But being able to get that better signal, being able to understand your
customers and
then take better action more effectively really helps to solve a lot of those
key challenges
that we in CS have just continued to face right over the years.
So we can use AI to get better signals, but how do we use it to take better
action?
How do we remove a lot of that grunt work really, truly in our day-to-day that,
again,
we don't want to be doing anyway?
I want to talk about a couple of key ways that AI helps us to take back our
time.
First, eliminating that need to log everything, eliminating the need to comb
through data
for insights and eliminating the struggle of finding the information that you
need quickly
and effectively when you need it.
So I want to talk first about how we can stop logging everything manually.
Anyone hate logging things?
Administrative work?
Yes.
How much do we all spend on admin work?
Is that like 50% of our time?
Is that probably an accurate statement?
When we're thinking about all of that time that we could be spending with our
customers,
we could be building those relationships, this is one of the key opportunities
for AI.
So a key way for us to solve a lot of that basic manual effort that, again,
just takes
so much of our time is to let AI automatically log all of our different data
sources.
I think this is one of the key challenges, and I know Ori talked about it with
our customer
OS, is that we have so much customer data in so many different places.
And it's really hard to get all of that data together without having to set up
extensive
API connections between different systems.
AI really helps us to do all of that automatically and a lot more effectively
so that it knows
what we need to capture and log and bring it all together.
So probably a basic use case, but a really important one.
The second one I want to talk about is combing through data for insights.
We talked about cheat sheet on stage as well, but tools like cheat sheet help
us to instantly
summarize accounts, highlight all of their key risks, understand when renewal
discussions
are happening, what the strategic priorities are, and if they've shifted over
time, which
really helps us to not only stay current on an account, but also get caught up.
Great example use cases for things like cheat sheet are if you're a CSM who's
just changed
to a new book of business, for example, or if you're a brand new CSM who's come
in to
take over an account.
Things like cheat sheet help us get up to speed so much faster so we can be a
lot more
effective in our roles and in that new account.
It's also great for things like executives who are maybe jumping on a call, for
example.
This is actually a feature that's replaced for some of our customers' meeting
prep.
They said they used to spend 45 minutes creating a document for an executive
before every
business review.
45 minutes of combing through data, 45 minutes of double checking the most
recent conversations,
45 minutes of making sure that they had examples of those to share with the
execs so they knew
what was happening.
They replaced that entire process with cheat sheet.
And so one of the ways that they leverage this is inside gain site, sending an
with cheat sheet, which you can do with a click of a button, helps you to get
that information
in front of your execs, avoids that 45 minutes for every single call they were
spending,
and gives their team a ton more time back as well.
And then the last one I want to talk about is eliminating the need to navigate
around
a product for answers.
Even the best products still have a couple screens, reports, dashboards that
you probably
have to visit to find what you're looking for.
And with things like co-pilot, which George had demoed on stage this morning,
you're really
able to get a lot faster answers about your accounts in a lot more screens, a
lot more
clicks, a lot more time back for you to actually leverage those data.
Georgia, sorry, words, Georgia talked on stage, about a couple of key use cases
in terms of
being able to understand reference customers, strategic priorities, being able
to link into
timeline entries and understand where that data is coming from.
But co-pilot can help you also to do so much more from a strategic perspective,
right?
If you're an executive, you can use co-pilot to understand where my top risk
accounts were
I need to spend my time.
If you're a product person and you're trying to understand what your CS teams
need, being
able to ask questions around what are the top features that my teams are asking
for,
what are the top product risks that are showing up in every account?
Being able to summarize and provide that data with something like co-pilot is a
really key
way for us to, again, leverage AI to collapse those insights into a really
consumable format
for our teams.
Another great example with co-pilot is besides doing meeting prep like Georgia
demoed on stage,
being able to understand your entire book of business a lot faster.
Two key things that are built into co-pilot with gain site that are really
important.
One is co-pilot only looks at accounts that you have access to.
So if you're a CSM and you have a specific book of business and that's all that
you should
have access to based on your business, it'll pull answers from similar accounts
that you
can actually engage with.
So it's not providing you information you can't really take part in, right, or
where you
have to contact another CSM to talk to their customer and that whole spiel
which doesn't
actually save you time at the end of the day, right?
But it provides you with the right data that you can take immediate action on.
The other key thing with co-pilot is that it is available right inside Gain
site CS.
So I'll talk a little bit about this in a minute, but one of the great ways for
AI
to save us time and enhance our workflows is when it's readily available to our
teams.
When it takes us so much time to navigate out of the product to go to a
separate point
solution to find that data, it's really hard to save time ultimately at the end
of the
day.
So navigating around or avoiding navigating around for answers is another key
way AI can
help us.
And the last key thing I want to touch on is how AI can help us from a business
strategic
perspective to create more opportunities to drive revenue and ultimately solve
some of
the key business challenges that we face today.
Couple key things I want to talk about briefly.
Then using AI to help us better understand and eliminate churn at scale,
helping us to
identify time waste and inefficiencies, and then helping us support our
customer needs
with a better data set for other teams to help drive actual change.
The first key way AI can help us improve our bottom line is to understand and
eliminate
churn.
Hopefully we're all doing this today, I hope, in some form or fashion.
But it's often a manual process, often anecdotal based on what our teams are
saying or kind
of the common churn risks or needs for our customers.
And it doesn't always tie the data into what's actually happening within our
business.
So a key way for AI to help us to one, gather all those insights, but then make
them actionable
is things like takeaways where you can understand what are the key topics and
what is the general
sentiment about those topics.
As well as understanding things like churn analysis, this is one of the reports
that's
readily available in Staircase, but understanding what customers are turning
for what top level
reason.
And again, it's not the reasons that our CSMs might be suggesting, which might
be exactly
the right reasons, but it's supporting all of that with actual data and
understanding
where those reasons were the ones that immediately instigated the churn risk,
right?
And getting a little bit earlier in the cycle as well to identify those risks
at the right
time where we can take action on them.
The second use case is actually, I think, the most valuable one, but
identifying misspent
resources with AI.
This report is, I think, one of the ones that Ori showed on Staircase as well,
but understanding
team efficiency and effort scores helps you to understand where your teams are
spending
way too much of their time or where they're not spending enough.
One of the key things that AI lets us do is not only understand and aggregate
that data,
it also helps us tie it to revenue.
So we know exactly how much time and money we're spending on each of these
accounts over
time and where, again, we're miss spending those resources.
One thing to report like this would be really valuable for your team.
I think this is one of the things that every time I mention it, I get a ton of
head nods
because it is such a critical need for customer success.
Really interesting stat from TSIA at an event earlier this year, they said that
something
like 40% of CS teams can't actually correlate action to results.
So it's really hard for us to say, my teams are engaging, they're having these
EBRs or
QBR conversations, they're chatting with our team on a daily basis.
I have no idea if that's actually driving them to be retained or if it's that
they just love
the product or if it's something else because they have a relationship with an
executive.
So things like this really help us to not only understand where our resources
are being
spent, but really prove our value at the end of the day as a department.
The last use case I want to touch on really briefly is supporting customer
needs with
data.
Again, probably something you're doing today, flagging what are your key risks,
understanding
and working with product to say these are the top challenges we hear all the
time from
our customers.
But what AI lets us really do is support that with data.
So understanding what are the key topics that are coming up on a regular basis,
which accounts
in what amount of annual recurring revenue is tied to those risks.
It makes it a lot easier to have a conversation with product when you can give
them that data
and give them kind of the full picture of how it's really impacting their
customer experience.
So a key use case for us to leverage AI to do something, hopefully we're
already doing,
but do it a lot better, faster and more efficiently.
The last thing I want to touch on really briefly is how we kind of think about
AI for the future.
These are a couple of tips for whether you're using AI really effectively today
or whether
you're just kind of getting started and here to kind of learn a little bit more
about what's
possible.
These are really important things to think about when it comes to implementing
AI for
customer success.
The first is that AI needs to drive real ROI.
We talked in the beginning of the presentation about how one of the key
challenges and reasons
people aren't necessarily continuing with AI is that it's not actually driving
value.
And a lot of the early use cases, a lot of the gimmicky AI was really in that
bucket
pretty squarely.
But now that we've gotten to a point where we understand our use cases, we
understand
was needed for customer success in particular, but also in general.
AI can actually translate to $2 saved and hours saved.
GainSight saves about 10 hours per week per CSM with AI by doing things like
automating
data entry, by doing things like helping us to prep for meetings.
Our customers collectively have saved over $36 million, not Euro, sorry, didn't
do the
conversion there, in labor cost savings, as well as taking that savings from AR
R saved,
expansion opportunities identified that we could take advantage of because of
AI.
And then the last point here is AI being strategic.
82%, this is one of the stats from our state of AI report, think that AI will
have a significant
impact on company strategy over the next year, NCS strategy in particular.
And a lot of that is because we can get more prescriptive, we can't understand
why we're
turning, we can't understand where our team effort is being misspenned and
leverage that
data to make a lot more effective decisions as a CS business.
The second point is that AI needs to be fine-tuned for customer success.
A lot of early AI was pretty generic, and things like this example here, the R
FP email
where it's been great working with the team, excited to include you, but
procurement has
started an RFP which is a request for proposal which means they're looking at
different products,
right?
If this is sales, that's a great email, that's a happy email.
For customer success and for post sales, RFPs are generally not a good sign.
So AI that understands our industry and knows our needs can flag this as a risk
as opposed
to maybe thinking it's a positive email, again if it's not fine-tuned for post
sales.
Another key point AI needs to be within workflows.
I mentioned this briefly earlier, but when we have to navigate between systems,
we're
losing a lot of that time savings that we're gaining with AI.
So a really key use case for AI and post sales is to have AI where your teams
are, where
they need it, where they're able to adopt it as well.
One of the key things we're seeing around AI adoption is that learning a new
tool is
still learning a new tool, whether it's AI or whether it's any other new
technology.
And when you're not able to leverage it within systems your teams are already
familiar with,
it takes a lot more effort for them to be able to adopt and use it effectively.
So things like right with AI within gain site, which is just using AI to help
you write a
faster email.
Yes, it's tuned for post sales, it has some CS use cases you can send it from.
It's no different than writing an email with chat GPT ultimately at the end of
the day.
The value is it's right within where you are already, where you're writing that
email,
and so you can leverage it quickly and effectively rather than navigating out
of your system,
creating that email, copy-pacing it back into your system, sending the email.
And then the last quick tip is that AI does need training, right?
It is something that we still need to educate our teams on.
There's a lot of change management involved with AI and changing our processes
from how
we've traditionally done things to leveraging this new tool and technology.
And so it is something that we really have a huge opportunity to upskill our
own teams
on.
It was actually at a women of CS event last month, and one of the quotes from
one of the
panelists really stuck with me.
She said that AI won't take our jobs, but CS professionals that know how to use
AI will.
Probably true, right?
When we think about the next 10 years, I know Nick talked about Sam Altman's
quote, right?
A baby's an AI being more intelligent than babies ever will be, which is sad.
I have a five and a three year old, so hopefully they're all so smart.
But leveraging AI and helping our teams to get ahead with AI will be really
important
for us to be able to stay ahead of our competition as well as keep our industry
growing and thriving
over the next couple of years and far beyond.
But very important for us to make sure we're providing that training for our
teams so they're
confident in AI, as well as AI in the specific use cases we have.
One other quick anecdote, Gainsite has an AI for all program where every Friday
, it's
like an hour, hour and a half, where a couple of folks on our team just walk
through using
AI in general.
Not necessarily for CS, not necessarily within our tools, but just how do you
ask a good
prompt?
What are good ways to use AI that maybe we haven't thought of before?
What are all the new tools out there like cloud that we should start to explore
and use for
different use cases, right?
And so it's a great opportunity for our teams just to get comfortable with AI,
not necessarily
get trained on our own specific AI technology.
So a great way again for us to provide that training.
But ultimately AI is only going to accelerate.
I don't think that's a surprise for anyone.
I think it's going to continue to grow within our industry.
It'll also continue to grow at Gainsite.
So we have a great roadmap session later on this track if you're interested to
learn how
we're leveraging AI in the future.
But we'll continue to grow within the Gainsite product within probably every
other tool that
we use.
Another great quote I heard recently, AI is the new UI.
So using more chat interfaces, using more co-pilot type experiences, is
becoming more
and more than norm when it comes to products in SAS in general.
And so quick summary, key takeaways to close us out.
AI can do a lot.
It can help us to eliminate a lot of those blind spots that we traditionally
had in customer
success.
It can help us to remove a lot of the grunt work that takes up so much of that
administrative
time.
It can also help us to be more strategic and find new opportunities to drive
revenue.
And ultimately when it's fine tuned for post sales, it can deliver so much more
value for
us and our teams and our specific needs.
And I think that's it.
So thank you all so much for sticking with me.
I think we have some Q&A and Slido to go through.
All right, please give it up for Tori.
Amazing.
Thank you, Tori, for those insights.
I'm sure you've all got some food for thought leaving the session today.
If you haven't done so already, please go into the app, submit your questions
via Slido.
We're going to take something down.
Okay, so the first one, if possible, could we get the questions up on the
screen as well?
Tori, how does the AI sentiment respond to a multi-language environment?
This client communication is mainly in Dutch, English, French, Portuguese and
Italian.
Yeah, it's a great question.
One of the things AI is really good at is translation.
So it does a pretty good job of understanding communications in any language
and then bringing
those all into the system.
I will say most of our systems are built in English.
So as far as like the interface being in English, that's kind of the output.
But it will analyze all those different communications regardless of language
that they're provided
in.
It's a great question.
Okay, thanks, Hannah, for your question.
Okay, so, Maria, how do you drive adoption of AI features within CS?
It's a good one.
Is this driving adoption within CS teams or driving adoption of our own AI
features?
Within CS teams, thanks.
So I think there's two kind of ways that we do this ourselves against site.
Thank you for the clarification, Maria.
So one is we include AI within a lot of our own workflows where it's kind of
not optional
for team-celebrage it.
So in terms of creating specific tasks, maybe there's something that we require
them to use
AI for, a great way to drive adoption if you kind of have to use it, right?
We also do a lot of internal training on our AI features.
So for us to be able to use them, I guess one thing about Gainsite is we're our
own best
customer zero, right?
We leverage our AI to be able to tell our customers how to leverage our AI
effectively.
So I think that really works in our favor as well.
But one of the key ways that we do a lot of that internal training as well.
One other, I think really good example I've heard from other customers and
different companies
on this is just encouraging AI within specific kind of extra credit assignments
for their
team.
So sort of treating it as like a prize to be one for different kind of programs
and different
adoption initiatives that they'll run internally is also a really great way to
drive that adoption.
Thank you, Maria.
Okay, so it looks like we've got a Gainsite customer here.
They've got the CS, education and PX products, but not staircase and won't get
the budget
for it right now.
So Tori, can you talk to what functionality around AI is available within Gains
ite CS
that doesn't require staircase?
Yeah, it's a great question.
I wish we all had unlimited budget.
I totally understand.
I think one of the big things within Gainsite CS right now is a lot of our
workflow tools.
So things like co-pilot, things like cheat sheet, all of those are directly
available
within Gainsite CS.
Co-pilot is again where you kind of ask all of those different questions and
still provide
some really great insights.
So even if you don't have staircase, you still have a great opportunity to kind
of collect
some of those and summarize and capture all of those for your team.
The one thing you won't get is staircase brings in a lot of channels that aren
't necessarily
native integrations to Gainsite.
You can integrate Gainsite CS with a ton of different systems.
Staircase is what kind of eliminates a lot of those connections being necessary
and a
lot of that data entry.
So it's really kind of the difference between the two.
Staircase is almost the insights engine, if you will.
Gainsite CS is still where you leverage a lot of those workflows and AI to help
eliminate
some of that groundwork.
Amazing.
Okay, so next question.
English isn't the first language for most customers.
And so conversations are often direct and perhaps could be perceived as bad,
but they're
not.
So how does Gainsite AI manage this and does it process other languages?
Yeah, it's a great question.
This I think speaks to that kind of fine-tuning need for AI.
Both within customer success, comparative to other customers.
So one way that RAI kind of understands if a conversation is bad, quote unquote
, right,
is compared to all of your other customer base.
If everyone is kind of talking in the same way, the AI can understand that that
's like
the normal mode of conversation.
That is not something that we treat it as a negative sign.
And it'll capture and understand that pretty effectively.
One thing we also do with some of our AI is we have kind of the human element
of cross-checking
that information.
So if AI is 98% on par, there's still that 2% where maybe it needs some tweaks
based on
every business, right?
A great example is if your company name is something that might indicate a turn
risk,
right?
Every time someone uses your company name in an email, probably not a good
thing to flag
as a turn risk, right?
So we do have some kind of human elements in there that help support, hey, this
should
not be flagged or this should be, just to make sure it's really fine-tuned well
for all
of our customers.
Okay, next question.
We get this a lot and you'll be pleased to hear the answer.
So any plans for integration with Microsoft Teams and other Microsoft solutions
Yeah.
So Staircase integrates with Microsoft Teams today.
There's no delay on that.
That exists.
We also have very, very soon on our roadmap and I think it might already be in
beta potentially.
Microsoft Teams within AI follow up in Gainsite.
So leveraging that Teams data to use a lot of those AI features is on the very,
very
near-term roadmap.
Yeah, if you want to find out more about that, please go to the Gainsite booth
and speak
with our team.
Okay, Timo asks, "How about AI driving upsell opportunities and not just
reducing turn risk?"
Yeah, it's a great question.
One of the big ways for AI to help with that is by identifying expansion
signals in your
customer base.
There's a great example I think Ori gave on stage this morning of a support
ticket where
maybe they said, "Hey, I want to add new users in my account."
Great opportunity, right, to upgrade someone to a new package.
And that support ticket is only addressed by your support team whose goal
should be, right,
to close as many support tickets as quickly as possible.
It doesn't always get flagged to your account management team, right, to be
able to take
advantage of.
So AI helps us to grab a lot of those signals and flag those four expansion
opportunities.
It also helps you to understand which accounts are primed and ready for
expansion based on
your customer base.
So by understanding your overall signals of if these three things happen, this
account
is more likely to expand.
AI can flag those for you so that you know where your expansion opportunities
could potentially
be, even if you haven't necessarily had that first paying right or sign of
commercial discussion
yet.
Absolutely.
Okay, next one.
Can AI handle when customer data is spread over different systems that are not
integrated
with each other and without a unique customer ID?
So I think for AI to work effectively, it has to be able to grab the data just
to make
sure I'm understanding the question correctly.
So when all of those different systems are bringing your data together, whether
that's
in staircase or a different system, and it can analyze all of that data, that's
where
it kind of makes the magic happen, right, to integrate all of that and
understand those
signals at scale.
You don't necessarily need unique customer identification in something like
staircase
to grab all that information.
So staircase knows, AI is smart enough to identify when you have, say, multiple
accounts
that are super similar in names, but maybe it's the same.
One of the key things that AI does really, really well is help you actually
clean up
data.
So with staircase, when you bring in unclean data from, say, your CRM, for
example, it
can actually suggest contact cleanup for contact management and then help you
feed that information
back to your CRM so you can clean up which customers are turned maybe and no
longer active.
So you can clean up, maybe you have two people where someone type owed their
last name.
I'm sure that never happens.
But type owed a last name and it's actually the same person.
AI is really, really good at suggesting those so that you can make those
updates in an account.
Awesome.
Okay.
Next question.
Collecting data from Zoom, Teams, and so on with customers, how do you manage
the data
privacy concerns with customers?
It's quite a common question.
Yeah, it is a common question.
I think there's a couple of different ways that data privacy kind of comes into
play.
One, everything with gain site, whether it's staircase or an organ site CS
system, has
all the same security and privacy protocols that anything else within gain site
does.
That is actually one of the advantages I think we hear from our customers a lot
is that gain
sites are already passed all those security and privacy and legal requirements.
So there's confidence in the AI within those systems to kind of have the same
stuff applied.
In terms of collecting it from customers and kind of getting customers sign off
on that
information, it's no different than if you have gone or any other call
recording system
right where you're already analyzing that.
So there aren't a ton of differences between leveraging that data for AI versus
in any
of those other systems if you already have any of those in place.
I think it is still a concern though.
We hear again a lot from our customers around data privacy and security.
So it will continue to be something when it comes to AI that I think the
industry is
trying to figure out the best path forward over the next year or so.
Okay.
Next question.
Obviously, at gain site, we've been using AI before we acquired staircase.
What's the key differences between the two-tour in your opinion?
It's a great question.
I highly encourage whoever asked this question to join our roadmap session a
little bit later
where we dive so deeply into this.
But in general, kind of quick hits, gain site AI is within our existing gain
site tools
like gain site CS, gain site customer communities, both have AI elements
available today and
there in education where you can do auto captions on videos.
Staircase AI is a separate product.
So it's not an add-on or kind of just our AI within our existing systems.
It's a totally net new product.
It is AI first and was built kind of from the ground up with AI.
So it is a little bit more AI than some of our other existing legacy solutions,
if you
will.
But that's kind of the main difference between the two.
Staircase separate product, gain site AI is all of our AI within gain site
products and
workflows.
Okay.
Next one.
How does AI calculate which accounts should take up more or less time?
What are their values and what does AI take these values from?
Yeah, it's a great question.
So the way that Staircase does this is you can actually input how much time you
think
it takes per email and per meeting.
You can also input how much salary, for example, your average CSM costs, right?
So if you calculate out how much labor costs that is for every hour that you
have, say,
10 people joining a customer call because they have a massive issue, that's a
lot of
manpower.
It's a lot of hours.
It's a lot of salary, right, from a revenue perspective.
So what it does is it takes all of that data to understand if you're spending
10 hours
with a certain customer because they have a red flag every couple weeks and it
's a
massive issue but maybe it really isn't.
It's a lot of headcount and manpower again that you're spending.
It leverages that in addition to the ARR for every account and that's how it
kind of
calculates the team efficiency.
So if you have an account that's like millions of dollars but you're spending a
lot of time,
that's probably okay, right?
That's probably the right high touch model for your business.
If you have a lot of small accounts where you're not really having as much ARR
for
those accounts and you're spending all of those hours and labor dollars, that's
where
it starts to show you that your efficiency rates are pretty off.
Okay, I think we've got time for one more question.
Is staircase going to eventually replace the CS health, health growing gain
site and why
isn't the health school using more AI to automatically determine weightings?
Yeah.
So to answer to this question, in Gainsite CS, we actually have an AI optimizer
tool.
Scorecard optimizer within our existing scorecards where you can optimize your
weights and measures
using AI.
No staircase required for that, that's totally within Gainsite CS and is
available today.
So what that feature does is it'll actually say, based on your entire customer
base,
based on your trends of who's churned and who has renewed, these are the
weights that
we suggest you weighing more or less based on that historical data.
So it'll not only input or output what weights you should have, what measures
you should eliminate
or include and what those weights should be for those corresponding measures,
you can
also create scorecards that way.
So if you want to create a new scorecard, you can leverage the AI scorecard
functionality
to again create, here's what we think your weights and measures should be.
We actually have a customer who did this and they actually had their business
intelligence
team run what they think the scorecard should be and they were almost an exact
match.
So it's a really good validation of the scorecard optimizer actually looking at
the data correctly
to suggest those weights and measures.
In terms of will staircase eventually replace CS health scores, as we kind of
look at our
roadmap into the future, and again, highly encourage attending that session
later, we
have I think a couple different ways that staircase and CS will come together.
One of those is leveraging staircase data to inform health scores.
So I think that's definitely on the long term roadmap, but today you can use
that AI
scorecard tool to create scorecards now if you wanted.
Awesome.
Okay.
If you didn't get your question answered today, Troy's going to be in this room
for most
of the day.
But if you've got any other questions, please come and grab Tory or you can go
see the
Gainsite team at the booth.
Just a quick reminder, please do go ahead and complete the survey from this
session.
We're going to have a price draw and a raffle for $50 or €50 Amazon gift card
Go visit our sponsors.
You've now got 15 minutes to get to your next track unless you're in this room,
which case
just stay comfortable.
But please give it up for Tory.
Thank you.
Thank you.
Thanks, everyone.
Thank you.
Thanks, everyone.
for watching.
for watching.
for watching.
for watching.
[BLANK_AUDIO]
