Roadmap to Realized Value: The Future of Gainsight AI
Discover how Gainsight is not just envisioning, but actively building a future where AI drives deeper customer insights, automates complex workflows, and unlocks new growth opportunities. We’ll explore the new Copilot, upcoming AI features and enhancements on our roadmap, and share real-world examples of how these innovations are already delivering tangible value to customer success teams. Whether you’re looking to optimize your current operations or plan for future success, this session will provide you with a clear view of how AI in Gainsight is shaping the next era of Customer Success.
Hello, welcome everyone to our second Celeste, I think,
session of the day.
Super excited to welcome Shonton to the stage in a moment.
He works on GainSight's product team,
building all of the amazing AI that,
if you've been a part of this track you've
been hearing about all day.
And we're also really excited about
to help solve some of the key challenges for customer
success into the future.
If you haven't been part of this track,
my name is Tori Jeffcoat.
I lead our marketing team at GainSight around our CS and AI
programs.
Clearly said AI too much today as well.
But super excited to welcome Shonton to the stage.
Two quick housekeeping things before I do.
We do have all of our polls and opportunity
for you to ask questions in Slido.
So if you're in the GainSight Polls app,
if you pull up this session and go to the Q&A and Polls
section, that's where you can leave those questions.
You also have great raffle prizes for breakout session
surveys.
So when this session is done, make
sure you fill in that survey so that you can be entered
into that raffle for a cool price.
But with that, I'm super excited to welcome Shonton to the stage
who's going to tell you a little bit more about the future
of GainSight AI and a little bit more about our roadmap.
So welcome, Shonton.
Thanks, Tori.
Hi, hi, everyone.
Welcome to this deep dive into GainSight's AI roadmap.
I know you've seen a lot of roadmap, AI roadmap,
at the keynote today.
But I thought I'll take a step back
and talk a little bit about how AI is changing the SaaS landscape,
how it's changing CS, and what
GainSight is doing with this technology
and what to expect in the future for CS.
So they say to learn the future, look at the past.
Let's take a look at how customer management, customer
success changed from the '80s to now.
So to capture customer data, people
were basically given a pen and a paper.
And somehow, companies got along without a single digital byte.
And from this world of thick ledgers and filing cabinets,
we now move to these multiple digital systems.
We have your CRM, your CSP, and 20 other digital tools
that basically catalog every aspect of your customer.
And currently, all of these systems
are sort of partially integrated as well.
And in this transformation, we've achieved a magical ability
to store and retrieve information.
It's unbelievable, right?
And we store everything.
We store interactions.
We store customer behavior.
There's data about data.
And what this also did is that it unlocked advanced analytics
for us, and it's made business matter.
But what we've also lost is that it's
put a distance between the user's insight and the user.
You need entire analytics teams to make sense of that data.
And there exists data.
Silos now, because there's so much data.
And so that's what we've lost.
And when it comes to relationship management, again,
back then it was all in-person meetings, phone
call, physical mail.
And from that world where people used to literally
physically mail feedback to their companies,
we now have this digital infrastructure
where most communication is always digital.
It happens via email, video calls,
and all these one to many digital communication channels.
And we have systems that proactively reach out
to people, systems that manage risk.
We've come a long way.
And again, this has unlocked a massive efficiency
and scale over the years.
It helped us be global and serve a global clientele.
And we sort of got all this automation
through standardization of processes.
But what we've lost is that in this standardization,
we've lost nuance.
And we are now flooding customers
with a lot of impersonal communications as well.
Similarly, user journeys, for example,
onboarding was basically a book, if you
are futuristic, it was a CD.
And now from this world of manuals and CDs,
we've bought everything online.
Documentation is online.
Training is online.
Support even communities are now online.
And the scale that this unlocked
must feel like a dream to somebody in the '80s.
You could train millions of people
across the globe with a few digital assets.
And again, what we've lost in all of this
is that personal touch.
It's all the same content being pushed out to people.
We are drowning them in content and spam.
And now, in this journey, we've achieved
a massive amount of scale, but we've lost some things.
Now, what does AI bring to the table?
What's the future of CS look like?
So in the AI era, very soon, I expect a world
where there are no CSMs.
There's no CS admin, because the systems can take care of themselves.
And there's no CS middle management as a result of this.
So I'm kidding.
The CS job is too strategic to nuance to be going anywhere.
So pretty sure I gave some of them a heart attack.
My manager's here.
But it's not going anywhere.
And it's very open AI is rapidly expanding the CS team.
So that is not going anywhere.
So what AI can do is it can help us maintain, even accelerate,
the scale we've achieved so far.
But at the same time, it can help recover what we lost,
the human touch that we lost in this scaling process.
We're going to get it back.
So this is how I see customer success in the AI era.
So from information and retrieval at fingertips,
you would have insights at your fingertips.
There'll be very little data entry.
There'll be no more data silos, because systems can seamlessly
talk to each other.
And you'll have analysis on demand.
Similarly, for relationship management,
you'll have all the benefits of automation and scale
without having to standardize things,
because you have systems that can operate with nuance
under a variety of circumstances.
And then when it comes to user journeys,
you get personalization at scale.
You'll have systems that can tailor themselves intelligently
to a given user and a given situation within that user.
So what is it about this technology
and what is all the hype with AI right?
If you look back to the history of human evolution,
the invention of language was a very pivotal step,
because not only did it help you store and transmit knowledge,
it also made us more intelligent.
It made us capable of advanced complex reasoning.
It's the foundation of math, physics, science.
And now you have systems that can understand language.
You have machines that can understand language.
So it means that the machines can now
understand the kind of nuance that language enabled humans
to understand, but more importantly,
humans can now communicate that nuance to the machine.
And these machines are capable of comprehension, reason,
and execution, which is the ability to code.
So you can give them a complex command.
The system can reason to break down that complex command.
And then because it can code, it can also
implement the solution that it has come up with.
And what we are seeing is still the infancy of this technology,
but you can see how powerful something like this can be.
So that's sort of like my setup for what AI can do
toward daily lives.
Now, coming to what Gainsite has been doing and plans
to do with AI.
First, I want to talk about our platform a little bit,
because it's hard to talk about AI without talking
about the platform.
The fact that Gainsite is a mature multiproduct platform
for managing customers is a very big advantage for us.
Because we connect to a vast variety of data systems--
your CRM, your support, your communications,
and all the data like snowflake, et cetera.
We convert this into federated customer data, usable data.
And then we have a lot of tools that AI can leverage.
So access to this broad set of data and a broad set of tools
to go and implement solutions is what
makes Gainsite special and Gainsite AI more powerful.
I also wanted to talk a little bit about what
I call the trust layer, which is a set of data security
and privacy mechanisms we have in place
to make sure that your data is just as secure now
as it was prior to the LLM error, because you
have some well-founded doubts about your security and privacy.
First thing, we continue to be completely GDPR compliant.
The second thing is for the EU customers,
the data remains in the EU.
It does not leave EU.
And just for staircase--
I mean, you've heard all about staircase.
It's still working progress.
Very soon we'll get there as well.
Your data will not be used for training in LLMs.
We have that explicit contracts with our vendors.
We also have a zero data retention policy with our vendors.
So that means outside of the transaction of the prompt,
your data will not be stored outside of Gainsite.
And we do a data privacy impact assessment for every feature.
We make sure that every feature we ship adheres
to all the protocols that we need to.
And we also mask PII across most of our platform
with the exception of staircase, which again,
is working progress.
So all of this means that your data is just as secure
as it was prior to the LLM error.
Now I'll talk about the CSAI capabilities.
I'll begin with what we currently have,
because there's a decent amount of AI already
that you can use.
The first is the-- I'll give you a very quick overview.
I'm not going to detail about what we have.
We have the AI cheat sheet, which
is an AI-generated XX summary derived from timeline.
More sources coming soon.
AI follow-up is automated summaries, action items,
et cetera, coming into timeline from meeting transcripts.
Takeaways is your one-stop shop to understand
all customer feedback directly from service
and directly from timeline.
AI scorecards is an admin tool to help understand
how effective your current scorecards are
and how to improve them.
Right with AI is a convenience feature for all CS users
to prompt AI to draft first versions of emails
so they can be more productive.
We also have the renewal likelihood score for those of you
who use the renewal center product,
where you get an opportunity conversion score of sorts
for every renewal opportunity based on the data available
within GainSight.
So this is all currently available and GA within GainSight.
GainSight CS.
Now coming to the roadmap, the first one
is AI follow-up, but for all vendors.
This has been one of our most requested popular features.
So we only support GONG as the source of transcripts,
but even if you're using Zoom or Microsoft Teams
or Google Meet or Clary or Chorus,
we'll start to support all of these vendors
for the AI follow-up feature, or coming very soon.
And as I said, you get summaries, action items,
risks and a sentiment score for all your meetings,
automatically pushed to timeline.
And you can convert those with a click of a button.
Those action items can be converted to tasks.
And an enhancement to this coming later
would be the ability to draft a follow-up email
based on the transcript and the summary
that you can send to the customer.
Also, the AI will look for open questions
that the customer might have asked during the call.
And it will respond back or attach within the customer email
and answer to that question.
The answer is derived from Gainsite Connect
to knowledge bases and other sources of knowledge.
So this is all AI follow-up and all vendors.
The next one-- and again, this is a very big investment
going forward for Gainsite--
is the Gainsite Co-Pilot.
The very first version of Gainsite Co-Pilot
is going to be Co-Pilot Ask.
The idea here is whenever you have a question
about a given customer or across your set of customers,
we want you to ask the Co-Pilot.
And the Co-Pilot will answer based on all the information
available within the C360.
So that includes the timeline, scorecard, cockpit, success
plan, and opportunity data for that account.
And for a given question, for example,
you could ask it to help you prepare for an upcoming EBR.
Or you could ask it to summarize risks.
Or summarize the value realization
for a given customer before the QBR.
Or across customers, you could ask for questions like,
which customers have implemented instructor
training in the financial services domain
when you're looking for references.
The Co-Pilot will find these customers for you.
You can ask for product feedback.
You can look for systemic issues across segments.
All of that becomes possible through Co-Pilot Ask.
The next feature is Co-Pilot Analyze.
Now, Ask was dealing with customers and C360s.
The idea with Analyze is we want to be able to answer any question
across all gainside data.
Think of Co-Pilot Analyze like an AI analyst
available for you on demand.
So instead of having to go to the admin for them
to create the reports and insights that you need,
you can ask the Co-Pilot.
And Co-Pilot real time will build the necessary visualizations
or reports or analysis.
And so this is us basically disrupting our reporting
and dashboards.
And this is the future of analytics, essentially.
The next one-- so what you saw, Co-Pilot Ask and Analyze
is sort of milestone one for Co-Pilot,
where the goal being if you have a question on gainside data,
instead of having to hunt for it, you come ask the Co-Pilot.
The next evolution of Co-Pilot would be workflow agents.
The idea here is we want Co-Pilot to play a more proactive role
in the day to day of a CST, and not just answer when a CSM
asks for questions.
So it could do things like looks like you have an upcoming
meeting with this customer.
Here is a meeting prep that I generated for you.
Or I can help you prioritize the tasks
that you have this week.
Or looks like based on the email you sent to the customer,
I think one of the CTS can be closed.
Do you want me to close that for you?
Or you have a lot of pending CTS.
Do you want me to autoclose them because they seem stale?
Co-Pilot will play a more proactive role
in the day to day lives and assist in key CS workflows.
So that would be the next evolution of Co-Pilot.
And as I said, this is-- Co-Pilot in general
is going to be a continued big investment as far as gainside AI
goes.
The next one, called NARA Analyzer,
is another interesting feature.
So more and more, we are getting an ask for ROI
of various CS motions.
What is working?
What is not working?
How do I improve my CS function?
So NARA Analyzer is an advanced correlation engine
that will correlate various customer attributes.
And CSM activities done for a given account
with the eventual renewal or expansion of that account.
So you would know what attributes matter.
You would also know what CS activities work
and what do not work.
And in the process, you'll also be
able to establish a very cohesive narrative
on the ROI that your CS team brings to the table, which
is a question that is asked of a lot of CS leaders.
Does CS really matter?
Are you moving the NARA needle?
And this will help you provide a very cohesive narrative
on how CS is adding value in improving your NARA.
Again, this is a more traditional AI, not gen AI,
but much, much asked for feature.
Now, I'll move on to the staircase AI capabilities.
Staircase, as you know, is our brand new AI acquisition.
We are very excited with what staircase brings to the table,
to our ecosystem.
Again, a quick recap.
What staircase does is it connects
to your conversational data sources.
So it connects to the meetings, emails, and support tickets,
and events like conversations you have about your customers.
And from this data, it unearths a lot of insights and value.
And it does all of this in about a day.
So it connects to the systems and almost builds a C360
of its own for you with a lot of insights about customers.
And these are some examples of the insights
that staircase provides.
It's only a small list, but there's more here than just this.
So you get a sentiment that you can break down
by customer and person and segments.
And you can get alerts for risks and expansion.
We also have AI health scoring, which is all auto-generated
from customer data and the context that
is available in customer data.
So you get sentiment and engagement,
and you can also add it with usage data.
You get event detection, another very interesting feature,
where staircase sort of surfaces interesting moments
with a given customer.
So here was an extremely negative support ticket.
Or here is an executive who seems to be very unhappy,
either over an email or a survey.
Or here is a very positive event that happened recently.
And all of that is automatically
surfaced from customer conversations
and also pushed to execs as needed.
Insights is staircase taking all of this data
and giving you a very important set of alerts
about a given account.
So all the sentiment and engagement alerts,
but also the fact that you maybe did not
have a QBR with a given customer,
or the fact that you're not multi-threaded,
or not engaging all the right executives.
All of these would automatically
surfaced by staircase.
You also get all manner of summaries,
account summaries from all this conversational data.
You get churn analysis.
You get renewal analysis.
So those are available.
Staircase also gives you a very rich view
of the relationship that your team has with your customers.
So which people in your organization
have what kind of a relationship with the people
at your customers?
The frequency of interactions, the sentiment,
and the general strength of relationship
is automatically surfaced to you for every account
and every person.
It also shows the various topics that are being talked
about across customers or with a given customer.
And finally, you also get a customer effort score.
So this is a matrix of how much time and money
you are spending with a given customer
compared to the revenue that they're giving you.
And this is a very insightful metric for a lot of leaders,
because there's a lot of time there's
a few surprises on where the money and effort is going.
So this is all currently available and impressive set
of AI capabilities.
And this acquisition has accelerated our AI roadmap
by quarters.
So this is very, very excited to have this.
If you talk about what's next with Staircase,
the biggest priority by far currently
is the integration between Staircase and CS.
So think of the integration as happening in three phases.
The first phase is to give you the experience
of a single product.
So you'll have single sign on so that a user, once they log
into one of the products, is automatically logged in
to the other.
And then all of the insights about a customer
that I just talked about, they will now
be available on the customer 360 within GainSight.
So that will be available within the C360.
And also all the interesting reports and exec dashboards
that I talked about, they will also
be available within the GainSight CS.
So that is phase one of the integration.
And this is something that will be available in a matter
of a few weeks from now.
Next, we are going to enhance this integration
with a more deeper embedding of the two products.
So Staircase has automated--
Staircase connects to your email sources.
So you'll have automated email logging coming directly
into Timeline so that CSMs no longer
have to actually check a box for the emails to flow
into Timeline.
Staircase does the deduplication.
It routes the email to the right customer.
And they'll start to be in Timeline.
And all the lifecycle events that I talked about,
the bad emails, et cetera, all of them
will start to flow into the CS customer journey widget as well.
The sentiment and engagement scores
that Staircase auto-generates, they
will be available in the GainSight health scores
as out of the box metrics within that scorecard.
You'll also have all the insights that Staircase provides
available right within GSO man C360 in CS.
So deeper integration where all of the Staircase insights
are now available for CS users and within the CS product.
Going forward, we are going to have a joint motion
on expanding the AI capabilities of the GainSight CS
platform in general.
So that would mean the ability to embed various Staircase
widgets into the layouts within CS.
Both of our AIs working well with each other.
So Staircase AI insights being available for our co-pilot
or to the GainSight cheat sheet and the other way around.
So all of these deeper integration and the co-evaluation
of CS is going to happen past the expand phase.
So this is the GainSight and CS integration.
A lot of effort and urgency is being put
into accelerating the integration.
Now the next very interesting feature
that Staircase team is working on is the idea of AI tasks
and next best action.
So Staircase already contains the ability
to detect open items.
But what AI tasks is, it's more strategic in nature.
It's not that, oh, you said you'll send an email,
you did not send it.
But based on all the conversation and the context
I have about this customer, here is some tasks
that I think you should be taking up.
And we want to evolve this into essentially the next best
action for a given customer based on all the information
available to Staircase and CS.
So that's the next-- that's in the roadmap for Staircase.
And we also want to bring Staircase to where you are at.
A lot of users, not necessarily CS teams,
but a lot of other users also.
They're not necessarily logging into GainSight regularly.
And we want Staircase-- so Staircase
will automatically start to push insights
into specific customer channels.
You can invoke Staircase from within those channels.
You can even perform certain actions
from within those Slack channels.
Similarly, the ability to seamlessly share the insights
and the data that you're getting outside of GainSight
is also something that we're working on with Staircase.
And the next is teammate efficiency.
Staircase, as I said, the customer effort score,
which tells you how much you are spending
compared to the revenue.
Teammate efficiency is essentially
Staircase giving you insights on how you can optimize
this customer effort score.
Is your effort in proportion to your revenue?
If not, what can you as a user do
to make it more aligned with the revenue?
Or what can you as a manager do to the allocation
of portfolio of your CSMs to make sure
that the efforts are aligned with the revenue
that you're getting?
And you're maximizing the hours that your CSMs have,
basically.
That's the next evolution, essentially,
of the customer effort score.
And we're also going to show a lot, many more new kinds
of signals that Staircase currently does not show.
So the first one would be expansion.
Staircase currently is focused more on risk,
but it will also now start to look for expansion opportunities
that potentially with CS, you could auto generate CSQLs even.
And also, it's going to surface more advocacy moments
so that you can drive your advocacy processes
and use that during your EBRs or annuals.
Also, product feedback is something
that's going to become surfaced by Staircase
that you could possibly push to your product teams as needed.
So new kinds of signals apart from just what risk is--
apart from risk.
And again, as I said, we want it to be omnichannel, very
contextual, and pushed outside of K-Insight
as needed as well.
Now, I'll finally talk about our digital/scale products, which
is our CCC and PX products.
So what you have seen so far essentially
are three vehicles of Gainsight AI.
So the CS, the flagship product, is the workflow engine.
You are using it-- the CS team is using it
to drive their workflows.
Staircase is the insights engine which
takes all of the raw data and generates valuable insights
from it.
And finally, you have our scale engine,
which is this set of digital capabilities,
a lot of which are actually for your end users, right?
Not just for your internal teams, but for your end users,
for better digital experiences and self-service.
If you look at what we have currently, within CCC,
our community's product, you have right with AI for content.
So where an admin, when he's wanting to create posts,
can prompt the AI to write first versions of the drafts,
and then they can edit it and then post within community.
We also have AI-generated post summaries,
which include all the original posts and the comments.
And you can quickly understand what went on in a post.
You also have AI recap, which is an AI-generated summary
of recent community activity across posts and groups.
In our customer education products,
you also now have auto-generated captions for course videos.
This is, again, all GA available to you right now.
Coming to the roadmap, the first one in communities
is AI moderation.
So this is going to be a big help for community admins,
where the AI will sort of move content to the right places.
It will give you sentiment scores.
It will close outdated content.
It will give you automated tagging.
And basically help with moderating a community.
And this is a tool for the moderator, essentially.
What is also coming very soon is real-time translation.
So against any post, the user can go in and ask
it to be translated to any language of his choosing.
And the translation happens real-time right within that page.
The next one, again, very interesting and powerful use
case, is AI search.
So within the community product, your customers, their end-users,
can come in and ask questions about your products and services.
And the AI will answer those questions
based on all the information available within the community,
but also all the sources that the community connects to.
So our community connects to knowledge bases.
It connects to support systems, et cetera.
And the AI will answer based on all of this information.
So this is just better self-service for your customers.
And you'll also get links to the specific sources
from which the answer was generated.
And very soon, we'll also bring this capability in-app
through our PX product.
So your end-users wouldn't even have to leave the product
to get their product questions answered.
And they can serve it right there within the product,
through the KCPot.
The next one is the AI tutor.
I mean, we've all sat through these long training videos,
which is not really gripping content.
What you really need is the ability
to tailor your own learning path and get answers to questions
as and when they arise.
And that's where the tutor comes in.
So the tutor is this interactive experience
in a education product, where you can ask questions
to the tutor, get your answers, and sort of go
on your own learning path as opposed
to sitting through this videos about products.
And again, this is going to be the future of customer
education, definitely.
And finally, you have here sort of like what we have now,
what we are actively working on.
Available is what we currently have within the product.
Now is stuff that we are actively working on.
So we expect to deliver in the Q4, Q1, type and time frame.
And finally, what we plan to work on post that.
And this is just-- I've categorized what we've just
all talked about into this slide here.
And yeah, so that was my presentation.
[APPLAUSE]
Thank you so much, Shanton.
Such great features, such great content.
I think we have a couple of really good questions
coming about everything you presented.
I'll go ahead and read through them if we can get--
yep, there we go.
So first question, when is Co-pilot going to be available?
We had, again, Safe Harbor.
Don't hold me to this.
They're targeting the end of Q4, end of January for this.
So the Co-pilot asked, B1, targeted for end of Q4.
I realize we're asking a product person to commit to decline.
It's not recorded, right?
[LAUGHTER]
Second question, our core challenge
is to have proper CRM admin work from CS on gain site.
Without proper data feed, how effective can AI be?
So I think the question is asking,
when you don't have all the right data,
how effective can AI be, and how will we address data
cleanliness problems in gain site?
So this is a bit of a two-step problem.
And I know that a lot of our admins
have this longstanding issue with us
that we're not giving them the tools for data cleanliness.
I'll tell you what our current thinking is on this.
First thing is, there is a lot of data that we--
what staircase does, right?
You don't need to do anything.
We pull all the raw data and surface all the relevant insights.
And that should answer a huge portion of--
and add a lot of value from directly that.
But obviously, especially for our enterprise customers,
there's an entirely different layer of data complexity
that exists when you're connecting with CRM
and a lot of the other systems.
Why we are not putting a ton of energy
into solving those problems is that it's very possible
that there's a very AI-first answer to this problem.
So instead of having to make enhancements to how our rules
work, for example, the future of administration
could be you configuring AI on how to identify the right things.
So think of the Copilot Analyze use case, right?
So what we would need from you is object definitions.
And we would need maybe what specific terms
within your domain mean so that the AI knows
what to do with these kinds of questions
and what fields to look at based on the descriptions.
So it's a very different kind of admin work
compared to you necessarily having
to build all the rules for the AI to necessarily work
and build the right reports.
So yes, we will provide tools for you
to maintain cleanliness and make it better.
But I also think it's going to change a lot in the coming years.
What that means to maintain data cleanliness.
Awesome.
I'm going to ask these next two together
because they kind of go hand in hand.
So there is one question around our Copilot and staircase
paid add-ons.
And then another question around is Copilot included?
Or is it an additional add-on to the system?
So a Copilot and all the AI features
that I mentioned in the CS bucket,
they are going to be free.
We will not be charging for them.
Staircase is a paid add-on.
[APPLAUSE]
Thank you.
Staircase is a paid add-on.
Awesome.
Good rounds of applause to that answer.
The next question around staircase
and staircase is specific health score.
Can you program your own weightings in that health score?
Or is it a set standard?
Yes.
So staircase comes up with its own default set of weights.
But just like CS, you can go and adjust those weightings
as you see fit.
Can you add just to add to that question, Shanton,
what are some of the differences between health scores
and gain site CS versus the staircase health score?
Right.
So the structure of it is the same in the sense
that the final score for staircase and CS
is a weighted average of certain metrics.
What the main difference between staircase and CS
is that in staircase, all of those metrics are auto-generated,
and they're all from conversational data, which
is very rich in context about a given customer.
What you have in gain site is that you
have to pull in the data yourself and create the rules
and the definitions for those metrics.
And a lot of times our customers struggle to define them.
They don't know exactly what metrics there
need to be, exactly what should be the weight for those metrics.
That's why we had the scorecard optimizer, the admin tool
to improve the scorecards.
But staircase sort of circumvents that entire process
and looks at an entirely different set of metrics.
And we were competitors once, but the staircase methodology
really does work.
And the good thing about this is they complement each other
very well, because we capture very different sites
of customer health.
And once we do our phase two of integration,
where staircase and CS scorecards coexist,
it's going to be a really good health monitoring framework.
I'm going to add to that that even prior to the acquisition,
a lot of folks had two health scores.
They would have the staircase health score
and the gain site size score side by side.
Because like San Chantan said, they don't really overlap much.
Even sentiment usually in gain site is a manual value
versus the one that staircase generates.
So you can have them side by side,
and now we'll work on merging them.
We have Ori in the audience who just spoke, if anyone didn't
catch that.
I think just to add on to what Ori kind of said,
if you couldn't hear essentially the difference between,
I think staircase is described as kind of vitals
versus your annual checkup.
Staircase is a little bit more of that real time,
always on health versus the kind of annual checkup
and gain sites a little bit deeper.
There's another question about integration in roadmap,
which Chantan, I know you mentioned the roadmap integrating
staircase and gain site CS.
So I'll go ahead and ask what's the second question
on the screen here.
Will the existing AI available in gain site come at the cost,
come at a cost with the integration of staircase,
meaning how will these two things come together
and how will that change in our roadmap in the future?
We are still defining that boundary.
What is staircase?
What is CSAI?
But as I said, our general thought process
is anything to do with driving workflow or automating
would be the CS domain.
And a lot of work that has to do with insights
would be under the staircase bucket.
But a lot of times those boundaries do get blurred
where insight ends and action begins.
And we will be defining that boundary as and when
we decide on feature items.
I mean, one thing also to consider with AI is--
yeah, thanks.
Sorry.
One thing to consider with AI is the more you start using AI,
you also incur more processing costs.
So just think about-- so if today you're analyzing maybe 100
meetings, suddenly you'll be processing thousands of meetings.
And that is a cost, obviously, that we just
need to balance any business.
And we're monitoring that and understanding
where does that line drop.
So that's from a cost feature perspective.
And I think all of you or those of you
that are software providers, you understand exactly what we mean.
So I think that's kind of the trade off right now
that we're trying to understand.
We analyze sometimes tens of thousands of emails a month.
Sometimes it could be in a year.
It could be a million or two million emails.
And we just want to make sure that we can balance--
deliver you the value at the same time.
Also run the business in a healthy way.
So that's a little bit of the considerations there.
Just want to make sure that--
you don't think that there's a pay wall because we want
to have a pay wall.
We really want to give you the value.
But we also want to run the business healthily.
All right.
Awesome.
Next question-- just I think around validating
the way that staircase works.
So in order for staircase to generate the analytics
that we've kind of mentioned in the presentation
and kind of throughout today as well,
it sounds like it needs access to Slack, emails, call
recordings, and support tickets.
Is that accurate?
That is accurate.
And it's all server-side integration.
Gain-side currently has user-level integration,
staircase-level server-side integration.
So you can connect your emails.
And it only looks at customer emails and meetings,
not internal conversations.
And everything else is automated.
But yes, it does need access to that info.
Next question I think a little bit deeper around how AI works.
Question is, to my knowledge, AI doesn't understand language.
It simply applies a statistical model.
And the output's the most probable next word in a sentence.
Is that accurate?
Do you agree with that statement, Chantan?
Yes and no, in the sense that how do you understand language?
Right?
You have a lot more neurons.
But it is not very different if you think about it.
This technology from-- it gets down very philosophical,
what you define as intelligence, and does it truly understand?
Because yes, it's not built to reason, for example.
But you can test its reasoning capabilities.
And it does a very good job of reasoning.
And I would say AI models understand language quite well.
You could ask it to write a sawne in the style of--
I don't know, some person, all starting with P. It can do it for you.
And without understanding language, how is that system doing?
What does understanding a language mean?
Yes, it's not built to understand language,
but it does understand language.
By learning to predict the next token, it understood language.
It somehow got rudimentary reasoning capabilities.
And this technology is more like getting to know a person.
It's a little bit like that.
This was a question from an earlier session, Chantan.
But I'm going to go off script and ask it to you anyway.
Someone asked, does it actually matter if you say please
and thank you when you're talking to chat, TBT,
or any other AI?
Does it actually matter?
Does it impact the output?
Curious to hear your thoughts.
Um, I should not.
Should not.
But again, it just gives it the information that--
it gives it feedback on how useful you found the answer.
But it shouldn't impact the answer itself.
OK.
Thought that was a fun question.
So I want to throw that one in.
The next real question from the session,
do you have a time frame on Google Meet integration,
I think, for AI follow-up perhaps?
Yes.
So Q1 is when we'll be launching AI follow-up for all vendors.
FYI, so we are going to be using the staircase backend
to deliver this feature.
Staircase already supports Google Meet, Teams, Zoom,
Clary, and Chorus.
So you'll get all of them at once in Q1.
Because you don't need to buy staircase for it,
but we're going to use the technology in the backend
to deliver this feature for you.
Awesome.
I think we have time for maybe two more questions.
There is a question, the second one on the screen here,
for a team new to using Gainsight AI tools.
How long does it typically take before you
start to see valuable insights?
Is there immediate ROI?
Does it take time for that data to accumulate
before you get actual insights emerging?
So it depends on the feature.
But AI follow-up, for example, it's immediate value.
As soon as the feature is turned on,
you no longer need to take notes.
The AI is doing them for you.
And cheat sheet, for example, it requires
a rich timeline.
A lot of customers already have a rich timeline.
But if you do not have a rich timeline,
either you start to add the right entries,
or you wait for AI follow-up to populate
a certain amount of data to automatically start
having a rich timeline.
Co-pilot also, it depends on how rich your C360 is.
But then it also comes down to how good you are
at interacting with AI.
And in our beta customers, for example,
some users get a lot of value from it.
Some users not so much.
Because we express certain experience
and skill of interacting with AI.
And we want to reduce that barrier in the coming--
that's why milestone two comes in.
But these are the factors that impact how much value
you get from AI features.
Awesome.
And then I think we have time for one more question.
So I'm going to pick.
OK, final question here.
Can staircase be leveraged if you're not a current gain site
CS user as a standalone?
If so, what other CRM and systems can it be integrated with?
Yes.
Odie, you want to take this?
Thanks.
Absolutely.
Staircase is a standalone product.
You can use it with any customer success platform.
Obviously, it has enhancements with the gain site.
But there are several other customer success platforms
out there.
And we work together with them, usually through the CRM.
From a CRM perspective, Salesforce, HubSpot, Pipe Drive,
we're working right now on Dynamics and SAP.
So if you are any of those, the first three,
it's already operating today.
Awesome.
Well, thank you, Shonton and Ori for answering such great
questions.
I think that's all we have time for today.
So just want to give Shonton a huge other round of applause
for a great session.
[APPLAUSE]
Awesome.
Thank you.
Thanks. Thanks everyone.
Yep.
And then just two quick housekeeping things.
Don't forget to fill in those after session surveys
so we can continue to create amazing content
in future pulses.
And then I think we have one more session for the day.
I think we have a 15-minute break.
I should know that.
We have a 30-minute break.
Sorry.
Extra time.
So 30-minute break before the next session
that'll close out today for a pulse.
