Nick Mehta & Shantan Reddy & Godard Abel 19 min

AI Meets SaaS with Godard Abel, CEO of G2


Godard Abel, CEO of G2, joins Gainsight AI to address key questions about AI pricing models and opportunities in customer success.



0:00

Hi, I'm Nick Neda, CEO of Gatesight, and I want to welcome you to a new virtual

0:07

series

0:08

we've created called AI Beats SaaS.

0:10

What we're going to talk about, how artificial intelligence is particularly

0:14

gender to the

0:15

AI is having a huge impact on software and service, and how all of us are

0:19

figuring it

0:20

out as we go.

0:21

So what we want to do each time is bring out along a thought leader in SaaS to

0:25

talk about

0:26

how they're thinking about AI and their business.

0:28

What we're going to talk about today is a long time friend, Goderd Abel, the

0:32

CEO of

0:33

G2.

0:34

How are you doing, Goderd?

0:36

Great.

0:37

And Chad and me, Nick.

0:38

Awesome.

0:39

And this is an actual Goderd, or is this like an AI version of you, because you

0:42

're probably

0:42

another meeting right now.

0:43

Is it an actual one?

0:44

Let me pinch myself.

0:45

I don't know if I can feel a little pain.

0:48

So I don't think AI can...

0:50

I don't know.

0:51

Yeah, can AI feel pain?

0:52

We'll learn that eventually.

0:54

So Goderd doesn't really need an introduction.

0:58

Many of you, I'm sure all of you, visited G2, used G2, partnered with that.

1:02

We do all of the above-it-game site.

1:04

And he's the CEO of G2, the incredible multi-time entrepreneur, ran a steel

1:09

break before this.

1:10

So it's a sales force, ran big machines before that.

1:13

Amazing background.

1:14

But one of the things I think that would be really interesting is Goderd has an

1:18

awesome

1:18

purview of what's happening not just in his own company, but across all

1:22

software, because

1:23

pretty much every software product on material size is rated and reviewed on G2

1:28

I know I refresh my G2 listing pretty much every day to learn more about what's

1:31

happening

1:32

in our market.

1:33

So we're going to dive in with a fun icebreaker because we always do that game

1:37

site.

1:37

So Goderd, what's one way you've used chat GPT recently, or anytime, that blew

1:44

you away?

1:45

I would say translation, because with my...

1:48

Oh, interesting.

1:50

My dad is also an entrepreneur, and he actually got involved with a window and

1:54

door company

1:54

in Mexico called Aventa.

1:57

And he wanted some help with the business.

1:59

I'm like, "Oh, we need to hire a CFO."

2:00

And I'm like, "Oh, I don't know how to speak Spanish.

2:02

I don't know anything about this."

2:03

But I'm like, "Yeah, yeah, you know, how to actually be T, help me write a CFO

2:06

job description

2:07

for manufacturing company in Mexico."

2:09

And first is in English, and I said, "Copy to the Mexican Spanish."

2:12

And I sent it to my dad plus the president of the company who's Native Mexican,

2:16

and they

2:16

looked at it, and they're like, "This is amazing.

2:18

It's great."

2:19

And I just...

2:20

Also, not to offer...

2:21

Like, zero edits, and we were out sending it to the recruiter, but I was just

2:25

amazed

2:26

at both helping create the English version.

2:28

But they didn't even amazed how it just created perfect Mexican professional

2:33

language, and

2:34

it worked out of the box.

2:36

And so, frankly, I'm also like, "Oh, wow, we got to accelerate our localization

2:39

for G2,

2:40

because we're still in English language only."

2:41

And honestly, I think...

2:42

And I think also then reading about transformers when they're developed with

2:44

Google, I think

2:45

it was around that translation use case.

2:47

So I do think it should really accelerate multi-language, but also global

2:51

collaboration,

2:52

because I think it does seem to just 100% cross-translation.

2:55

Yeah, I totally agree, and my ninth grader is taking introductory Spanish, so

3:01

hopefully

3:02

she's not using CHATCP to do her essays.

3:04

I don't think she is, but...

3:06

Want to make sure she's learning properly.

3:08

I thought I loved that.

3:09

So, Shantan, I think you got the first question.

3:11

I didn't introduce Shantan.

3:12

He's our head of AI, again, he's quickly become the most popular person.

3:17

He's the first day of G8, because we put a huge percent of our roadmap on the

3:20

generative

3:21

AI.

3:22

So, Shantan, why don't you do the first question?

3:24

Hi, hi, got it.

3:25

So my first question to you is, how is T2 using AI today?

3:31

And what are your plans with this big new technology?

3:34

Yeah, and we're using lots of ways.

3:36

What I'm most excited about is our G2 AI Monty.

3:40

And we launched AI Monty in collaboration with CHATCPT, OpenAI.

3:45

So it's built on top of their chat.

3:47

But really, our vision is it's kind of...

3:50

Monty was our mascot, so we said one, we met Monty's sentiment.

3:53

But we think it can be like your best software analyst or your best eccentric

3:57

consultant

3:57

in AI.

3:58

And I think it has over any analyst or any consultant as it can digest all our

4:02

data in

4:03

real time.

4:04

So we've created a plugin where we inject all the G2 reviews, all the G2 data.

4:07

And it's amazing.

4:08

And anyone can try Monty for free at our homepage, please do.

4:11

But the idea is we can much better guide any business professional to find the

4:15

perfect

4:15

app to solve any business problem they want to solve.

4:18

And I think it's a big breakthrough because today G2, we rely on categorization

4:21

, kind of

4:22

a fixed human created taxonomy, over 2,000 categories.

4:25

But most buyers, they just have a business problem.

4:27

They're like, "Hey, I want to prove my customer retention.

4:30

How do I do that?"

4:32

And traditionally, we've said, "Hey, you got to search for customer success

4:35

management

4:35

software."

4:36

Obviously, those are the industry.

4:37

We know that, but most buyers don't.

4:39

And I think this linkage using Monty AI to help people discover the best apps

4:45

is a very

4:45

powerful use case.

4:46

We're very excited about it, G2.

4:47

I love that.

4:48

And you also have to answer one of the very popular questions I think for all

4:52

of us, which

4:52

is, "What do you name your general AI bot?"

4:55

So you went with Monty.

4:57

Obviously, there's Alexa, this theory, lots of choices.

5:00

But let me switch to something a little more business-oriented.

5:03

I think about pricing.

5:05

Not just in your own business, maybe you have some comments there, but pricing

5:09

in your clients,

5:10

how they're pricing.

5:11

A specific problem I think we all have is, "General AI has a real cost to it."

5:15

Many of you know that.

5:16

It's not just free.

5:18

And then it's a little unclear what customers should pay or how they should pay

5:22

for it.

5:23

I didn't think about that.

5:24

Now, and maybe I will start with how we're doing it, G2.

5:26

And then I'll talk about we're doing it with clients.

5:28

And for now, we've made our buyer, Monty, for free.

5:32

And that's probably been our philosophy of G2 anyway.

5:33

We want to educate the software buyer for free with the best insights.

5:38

But it is real-class.

5:39

Now, we're also launching an AI Monty for sellers, which is, and the idea, you

5:44

could

5:44

put this on your game site listing on G2.com.

5:46

But the idea is almost like, "Yeah, no, people are doing this SDR, BDR use case

5:51

."

5:51

But we can train our Monty also to answer FAQs for game site, let's say.

5:56

Yeah, because one is cool.

5:57

The user of buyers digging in and they're like, "Oh, game site, that sounds

6:01

pretty good."

6:01

But now I have some real questions.

6:03

And obviously, traditionally, we try to get them full of lead form handed off

6:05

to you.

6:06

But now, like, hey, why not let Monty answer those basic questions for their

6:08

qualified,

6:09

some point handed off maybe to your bot.

6:11

And but we're also figuring out that does consume potentially a lot of

6:15

resources with

6:15

chat, GPT, and it's expensive.

6:17

And so I think for now, we're going to do to launch it.

6:20

We're going to launch it as a beta in November.

6:21

We're just going to bundle it with what we call our new core or essential

6:25

profile so

6:26

you all can try it.

6:27

But we're going to put a limit on it.

6:28

We're still debating what that is.

6:30

But it'll probably be like, hey, you get kind of up to, I don't know, like, $

6:33

100 a year

6:35

for free.

6:36

And then beyond that, we'll upcharge you.

6:38

But we want to make it kind of fictional as to try.

6:41

And by then at some point, yeah, we do have to upcharge if nothing else to

6:44

cover the cost.

6:45

And of course, the other thing, there's a lot of value.

6:47

And also, I saw this as a dream force.

6:49

I was out at Dreamforce and made it with a lot of their executives, like Brian

6:53

Millum,

6:53

their president and a lot of old friends at Salesforce.

6:55

But they also launched their Einstein 1 platform.

6:58

And we're also a customer.

7:00

And I asked us their Salesforce rep, do we just get this for free?

7:04

And I think the answer is no.

7:06

They are trying to make it an upsell add on to access to Salesforce Einstein 1

7:12

platform.

7:13

And because I had the same issues, right?

7:15

A, on the cost side, it cost real money.

7:18

And then B, obviously, if you're offering more value, you're investing a lot of

7:20

innovation

7:21

R&D, you've got to monetize it somehow.

7:23

So I think most people, and I don't know, I think, and hopefully our G2

7:26

strategy is

7:27

a good one where we're going as they give people a taste of the AI features for

7:31

free.

7:31

But then put a cap on it because the compute costs can run out of, can be

7:36

astronomical.

7:37

And there's a lot of value, right?

7:39

So at some point, it's also fair.

7:40

Yeah.

7:41

It says, "Menagers, innovators to charge our customers for it."

7:45

I think it's a great point.

7:47

We're actually on the whole bunch of AI features.

7:49

We've already come out, so I'm coming out.

7:51

Like for example, something that can help you prep the meetings automatically.

7:54

And with the similar strategy, we're worth being basically built into the core

7:59

cost up

8:00

to a certain number of credit.

8:02

And then when you want to go above those credits, you buy those.

8:05

Because I think unlimited consumption with significant cost is not a great

8:09

business model.

8:10

So we all have to figure that out.

8:12

What the others value, right?

8:13

It's probably broader industry trend and maybe it's more consumption-based

8:16

pricing.

8:16

I think AI, a lot of people are saying it may do that anyway because the other

8:20

thing,

8:21

we all rely on, most of us rely on seats and users.

8:24

It's licensed.

8:25

Exactly.

8:26

And that may become less relevant, right?

8:28

Because if bots really do more of the work, it also means our customers, and I

8:32

'm not saying

8:33

this is going to happen, but some of your customers might actually have less CS

8:35

Ms in

8:35

the future.

8:36

Yeah.

8:37

And obviously, they really grow, so they need more humans as well.

8:40

But there could be cases where like, "Hey, maybe they only need half the seats

8:43

because

8:44

a lot of the basic CSM support questions the bots not answering."

8:48

And so both for your business model and for alignment with the customer, right?

8:52

Probably some more consumption-based models like, "Hey, how many questions are

8:56

we answering

8:57

for you or how many customer issues are we resolving for you with our AI?"

9:01

And so I think there'll be some interesting new consumption models that we

9:04

probably will

9:05

all try and probably change how SaaS is priced in a lot of cases.

9:10

So while I'm saying, I think we're all figuring it out together.

9:13

So good to learn from each other.

9:15

I shout it back to you.

9:16

Definitely a lot of these concerns are something we are right now actively

9:20

trying to come to

9:21

a consensus on.

9:23

And we've touched upon each and every one of the points I just mentioned I'm

9:26

pricing.

9:26

Yeah.

9:27

And I should, Shantan, I should connect you to Tim Yandor, if my co-founder, he

9:30

's building

9:30

our AI and I'll connect you guys after it because it's an area we're all trying

9:34

to learn

9:34

in right now.

9:35

Moving on to the next question.

9:38

What have you given that you're a premier company, that evaluates software

9:43

sellers?

9:44

What are some interesting things that you have seen among other companies?

9:48

Anything innovative around AI that things stood out?

9:51

Yeah.

9:52

And I think, I mean right now, I see everyone's experimenting.

9:54

I'm probably a bit still dream force brainwash.

9:57

So I think all the things we're trying to do.

10:00

But it's obviously not just Salesforce.

10:01

You know, how the spot's also partner of ours.

10:04

And I think they're co-founder of Darmesh.

10:06

I think it's really cool what he's doing with chat spot.

10:09

And he's also focused on it all time.

10:11

So frankly, I see every vendor now really innovating.

10:14

I think some of the cooler ones are like writer, where they're just like one

10:18

specific use

10:19

case around content writing, marketing writing.

10:23

And I think they're, so we're starting to use tools like that to generate our

10:26

own content.

10:27

And so I think, obviously we all hear about BARD and OpenAI, but there's also

10:31

all these

10:32

startups just building really cool purposeful tools.

10:34

But I think it's still early days.

10:35

I think also the reality is my sense is like in the AI hype cycle, we may be,

10:40

you know,

10:41

if I could pass the peak and now I think a lot of people are saying, where's

10:45

the revenue,

10:45

right?

10:46

Where's the business impact?

10:48

But I'm a believer, you know, but I think we're also in that point of hype

10:50

cycle.

10:51

Now a lot of people have launched things, but most of them are still.

10:54

And I saw a BCD presentation.

10:55

Most things are still in beta head alpha.

10:57

You know, most things are not getting commercial use.

10:59

So I think overall, a lot of interest experimenting, but I don't know, besides

11:03

opening, I needed

11:04

to grow into a billion.

11:05

But I don't know if anyone's had kind of revenue, you know, besides them had

11:08

run away success

11:09

in generating revenue through it.

11:11

That's totally right.

11:12

Yeah.

11:13

One thing that you spoke about recently, you were a SaaSter and you spoke about

11:17

something

11:17

I've been thinking a ton about, which is AI can actually also disrupt the way

11:21

people

11:22

build user experience and user interfaces.

11:24

Can you talk a little bit more about what you think of that?

11:27

Yeah.

11:28

And I think, you know, I think the most software, to be honest, like we all

11:32

have bad interfaces.

11:33

And I built CPQ software for years, Steel Breaknow, Salesforce, CPQ and big

11:37

machines.

11:38

So we always had like forms based systems where like, yeah, you have 10

11:43

requirements

11:44

and the features you want, right?

11:45

But you're always filling in forms and fields as a user.

11:48

Or like if you're using a CRM system, it's like a put in the 10 required fields

11:51

on your

11:51

opportunity.

11:52

But frankly, no user ever likes it.

11:54

Frankly, most users hate it, right?

11:56

Which is why those systems, traditional CRM, even CPQ has a hard time off in

12:00

getting adoption.

12:01

You sort of have to force your users to use it and fill in all the forms and

12:03

there's

12:04

rules that kick off and tell you what you've entered wrong and forces you into

12:07

more workflows

12:08

and approvals, right?

12:09

But it's all I think to most salespeople, customers, successpeople, all these

12:11

traditional

12:12

systems feel more dangerous than help.

12:13

And obviously their managers needed, the company needs it to forecast.

12:16

We all know why the business needs it.

12:18

But I think AI could totally change that, right?

12:21

And I think the vision that most people are seeing, like if you're a sales rep,

12:23

customer

12:24

success rep, you can just do your work, right?

12:26

You just work in Gmail, you work in Slack, you work in Google Docs and behind

12:30

the scenes

12:30

and see AI is digesting all your work and then can update your opportunity for

12:35

you.

12:35

And we'll probably do it better.

12:36

Because we also have that challenge with sales reps, right?

12:39

Either they're too optimistic, too pessimistic, but there's that human element,

12:42

whereas testing

12:43

all your emails, it's listening to your zooms with the customer and analyzing

12:47

all that data

12:48

in real time, it'll do a better job than the human would anyway.

12:51

But I think a lot of these systems will probably do it behind the scenes, you

12:54

know?

12:55

And maybe the users are just working in a few systems that they actually use to

12:57

interact

12:58

with customers, at least on the front of the business, right?

13:00

And behind the scenes, the AI is capturing all the data, all the insights and

13:05

feeding the forms, if you will, feeding the structured data into the system.

13:08

So I think it could be totally revolutionary.

13:11

And I'm sure you guys are working on that game site, but I think that

13:13

traditionally user

13:14

interface app, enterprise app, and I'd also say for sites like G2, we're more

13:17

of an information

13:18

site and traditionally rely on search.

13:21

But even search, like I said, you have to know the category name, it's a bad

13:23

experience

13:24

site, but this interactive Q&A experience, that's where we're trying to go with

13:27

Monkey.

13:28

So we also think it's going to change websites like ours that people go to to

13:30

learn.

13:31

So I do think in five or 10 years, yeah, I think the UX of all apps, all sites

13:36

will

13:37

probably be totally different.

13:38

That's awesome.

13:39

I love this one.

13:40

This is one I'm super passionate about, because I agree, software today tells

13:44

users to go do

13:45

things, fill up forms and then take action.

13:47

Largely every product like that.

13:49

And we were actually trying to talk last night about, you know, we're building

13:52

something

13:52

where we'll take, you know, meeting transcripts, sort of a notes automatically,

13:56

it's not

13:56

going to be a match and I, and things like that.

13:58

But then the next level is how do you go do some of those action items for the

14:02

CSM, you

14:02

know, send the email, update, you know, update all the fields and salesforce,

14:06

like do everything

14:07

for them.

14:08

So I think it's a big opportunity.

14:09

The other time I was going to make, which I think is really interesting is like

14:12

you go

14:12

back to leapfrogging with AI.

14:15

You know, I feel like all of us constantly work on our user interface, change

14:18

this drop

14:19

down, make this thing look different.

14:21

And ironically, I'm definitely old enough to say, I can't say go to the viewer,

14:24

but

14:24

like, you know, having started by the kid using MS.

14:28

DAW like the command line, it's almost like the command line typing in a prompt

14:33

is actually

14:34

the best user interface.

14:35

And we went through years and years of like drop down to UIs and web apps and

14:40

we might

14:41

go all the way back down to a command line and just typing to answer questions,

14:45

which

14:45

is kind of like monster, you know, a new website, right?

14:47

So just type and figure out how you super human and same thing, right?

14:50

Like what are they differently in Gmail?

14:52

It's like all keyboard shortcuts, right?

14:54

Exactly.

14:55

I think my first computer was an Apple IIE, but that was also like green screen

15:00

commands,

15:00

right?

15:01

You just type it.

15:02

Yeah.

15:03

And ironically, maybe we're back to that, but maybe I think even beyond typing,

15:06

I think

15:06

now with voice recognition, right?

15:08

And like, you know, maybe also we just talk to our systems.

15:11

Yeah, I mean, especially depends on the preference.

15:13

Like if you're an engineer and I, yeah, school or if you're a physicist, like

15:16

you Nick,

15:17

you know, like we probably like typing, but then probably most salespeople that

15:20

are more

15:21

like extroverted, there's like, Hey, let me just talk.

15:24

And probably the AI will be able to see, you know, voice to text is great now.

15:28

And then he had parts of the text and updates their CRM better than they ever

15:31

would.

15:31

And they're just out there talking to people and selling, you know, like same

15:34

thing with

15:35

great CSM.

15:36

They probably want to think with the AI, probably won't do like it back to the

15:38

pain.

15:39

It probably won't be the human emotion and the human connection.

15:43

And even people say the AI test, like it's as smart as us.

15:46

I think that's achievable, but I don't think the AI, I can't see that we're

15:50

like connection,

15:50

feeling emotion.

15:53

And I think in any, you know, human customer relationship, both are important.

15:57

And I think that's the part, if we can take away all the, the grind, the

16:00

filling in forms

16:01

of workflows and then our, you know, our teams can just really build a

16:05

connection with

16:06

the emotional connections with customers and people.

16:08

Like, I think that would be a great place to end up.

16:11

I agree.

16:12

Yeah.

16:13

Something.

16:14

I'll be giving you one last question, right?

16:15

Yes.

16:16

Last question from me.

16:17

So let's get to customer success because I am a customer success guy.

16:24

So what are the biggest opportunities from your perspective to using generative

16:29

AI in

16:29

CS and also what are some potential threats in this space?

16:34

Yeah.

16:35

Well, I definitely, I think we talked about it earlier.

16:37

I think the idea of just asking, answering customer questions, yeah, which a

16:41

lot of CSM

16:42

I think spend a lot of time doing it.

16:44

Yeah.

16:45

And I think customers are like, oh, how do I get the most value or how do I

16:47

reconfigure

16:48

my workflow to do this?

16:50

You know, I think those kinds of questions, I think the AI will crush and I bet

16:54

you guys

16:54

are working on that, you know, as is probably everyone in customer success and

16:57

certainly

16:57

support use cases.

16:58

Same thing.

16:59

Right.

17:00

Everyone's saying, hey, that's such a great AI bot use case because probably

17:03

most cases

17:03

can be resolved without the human, you know, by the AI and then the human is

17:07

there for the

17:08

really tricky technical or if the customer needs some emotional love.

17:11

And I think the AI can also sense that, you know, and sense when does a human

17:15

touch needed

17:16

either because of emotion or technical challenge that the AI can't resolve.

17:20

So I do think that frontline customer Q and A customer support, I think that

17:25

will get

17:26

generally the AI is one of the best, it's one of the very best use case for

17:29

generative

17:29

AI.

17:30

So I do think everyone in customer success base will work on that.

17:33

And so I do think that's a massive opportunity because the end of the day the

17:35

customer will

17:36

get quicker, maybe better answers in real time 24/7.

17:40

So I think that a lot of customer success opportunity there.

17:43

I love that.

17:44

Yeah.

17:45

When I think about, you know, whether it's Chats and PT, it's kind of aggreg

17:47

ating and

17:48

using all the knowledge of all humans to answer questions when that's basically

17:51

the way it

17:52

works.

17:53

And I think in your company, you know, this stuff will work and I'm going to

17:56

get all

17:56

the knowledge of all your team to help individual customers.

17:59

So let me close out with some rapid, rapid association.

18:03

So each one, just tell me the first thing that comes to mind.

18:07

I try to get the key.

18:12

But bought.

18:13

Okay.

18:14

All right.

18:15

AI more generally.

18:16

Yes.

18:17

Machines being sentiment.

18:18

Yeah.

18:19

Exactly.

18:20

Or like I said, soon, some days smarter than us, you know, on the intellectual

18:23

plane.

18:24

A hundred percent agree.

18:25

And then a Terminator to good AI movie.

18:28

I think of Arnold and especially I'm half Austrian, seven Austrian mom and I

18:33

just think

18:33

of Arnold, amazing Austrian American accent.

18:37

I'll be back.

18:38

I'll just say, I hope you'll be back for another one of these.

18:40

You think great, great discussion.

18:42

Obviously check out G2 if you haven't, if you're the one person who has it.

18:47

And then on the game side, we actually, you'll see a link in the webinar.

18:51

We have a great page educating our AI for customer success and obviously talk a

18:55

little

18:56

about what we're doing in a game site.

18:57

Thank you so much.

18:58

I really appreciate it.

18:59

Thank you, Sean, Ben.

19:00

And thanks you all for watching.

19:01

Thanks.

19:02

Thanks.

19:03

Thanks, Sean, and look forward to collaborating with you more on our AI.

19:06

Love to.

19:07

Thanks, MeBonzo.

19:08

Bye.

19:09

[ Silence ]

X

This is a test comment.

X

This is a longer test comment to see how this looks if the person decides to ramble a bit. So they're rambling and rambling and then they even lorem ipsum.


author