Evaluate B2B Software Easier with G2 AI

0:00

[MUSIC]

0:05

Hello everyone and welcome to Go to Market AI,

0:07

the future of your Go to Market tech stack.

0:09

I'm your host, Sarah McConnell.

0:11

These days it seems like every company has AI,

0:13

but on this show, we want to go a level deeper so you can see firsthand how

0:17

businesses are actually applying AI to solve your business challenges.

0:21

We're going deep into the use cases and showing you live demos of

0:24

the latest and greatest in AI technology.

0:27

Today, I'm joined by Tim Silber, Product Manager on G2 Labs.

0:31

Tim, welcome to the show. I'm so excited you're here.

0:34

>> Hey Sarah, I have to be here,

0:35

equally excited and thanks for having me on.

0:37

>> Absolutely. So let's start with,

0:39

can you tell me a little bit obviously people know who G2 is,

0:41

but I'd love to hear in your words.

0:42

Who is G2? What are you guys doing specific to AI and who are you guys helping

0:46

in the market?

0:47

>> Yeah. So G2 was really founded on the question,

0:50

why is it easier to read reviews for a $100 hotel room and it is a $100,000

0:55

piece of business software?

0:56

Over the last 10 years, we've become the world's largest and most trusted B2B

1:00

software marketplace,

1:01

with around 2.5 million authentic reviews,

1:04

over 2,000 software and services categories and around 80 million buyers

1:08

annually.

1:09

So I'm a Product Manager on G2 Labs team,

1:12

which is a small hyper R&D focus team that's always looking for G2's next

1:16

product line.

1:17

So our goal is to drive innovation by conducting user research,

1:21

running experiments and then quickly prototyping, testing, and launching new

1:24

products.

1:25

>> So one of those more recent products that we'll be talking about today is

1:27

Monty,

1:28

our friendly Mongoose mascot who has really become the face of a bunch of our

1:32

recent AI initiatives.

1:33

>> That's amazing. I will say out of all of season one of Go to Market AI,

1:37

you guys five gold stars for the cutest AI mascot in Monty.

1:42

It's amazing to see you have a personality.

1:44

With that being said, Tim, let's jump into it.

1:45

I'd love to see a demo of what Monty does and how this AI is working into G2,

1:50

since people are already so familiar with G2 as a general product.

1:53

>> All right, let's go forward. So to state the problem,

1:56

the problem while G2 has always been great,

1:58

if you already have some idea of what you're looking for,

2:01

whether it's specific category of software or a short list of products,

2:05

what if you're a buyer and you don't know exactly where to start your software

2:08

buying journey?

2:10

How can we leverage conversation AI to streamline that experience?

2:14

So in April of this year, we launched Monty,

2:16

the world's first AI powered BUB product recommendation assistant to address

2:21

exactly that.

2:22

So if you're ready, I'd love to introduce an Monty.

2:24

>> Yeah, that's awesome.

2:26

>> So Monty currently lives in a few places in G2,

2:29

but simulate a true software buy experience.

2:31

Let's just start from the G2 homepage and chat with Monty,

2:35

so it'll get started.

2:36

We'll ask Monty for help of choosing the best software for my business.

2:40

Let's say I want to learn more about conversational marketing software.

2:48

So Monty is now searching across the thousands of categories and

2:52

hundreds of thousands of products on G2 based on my question to find the right

2:56

category and

2:57

a sure list of products for my needs.

2:59

So Monty starts off by generating a brief summary of the category and

3:03

found two relevant categories,

3:05

conversational marketing and conversational support,

3:07

both of which are relevant to the question if you're not exactly sure what you

3:10

might be looking for.

3:12

What's cool about Monty here is he's actually injected

3:16

the G2 grid, which is an interactive widget.

3:20

So let's take a look while we're in filter grid.

3:22

We can either choose a live grid or a trending grid or even filter by company

3:27

size.

3:27

So let's choose the enterprise grid.

3:30

And I can see that qualified as a leader in this space.

3:33

So let's ask Monty to tell me more about qualified.

3:38

>> That's really cool that you can filter right there in the chat function.

3:41

That you're not bouncing around a bunch of different places or

3:43

prompting it a whole bunch more that then you kind of lose that grid.

3:46

That's really awesome that you can pull that filter.

3:48

That's a great UX feature.

3:50

>> Exactly.

3:51

Yeah, so Monty's now searched through our product catalog.

3:53

We found qualified and is providing some high level information like the 4.5

3:58

star,

3:58

which congratulations by the way, that's awesome from Nanat Reviews.

4:02

And then also some basic information about the product and functionality.

4:05

And then also including some pricing tiers, which is one of the most frequent

4:09

queries that buyers are asking to Monty.

4:12

And then Monty loves to keep the conversation going.

4:15

But also the opportunity to offer the opportunity to say, hey,

4:18

I want to now contact qualified.

4:20

So we can do exactly that.

4:22

And we inject the GTI to lead forward around the chat,

4:24

pre-populated with my user information to streamline that process and make it

4:29

pretty easy.

4:30

So Monty's helping thousands of buyers every month streamline their software

4:34

purchasing experience.

4:35

>> That's really cool.

4:36

I've obviously used GTI quite a bit in my time as a B2B buyer.

4:40

And if one, I've always found the GTU interface really easy to use in general,

4:43

as far as finding what I'm looking for and then comparing it to other products.

4:47

But it does take time like scrolling through and looking for different,

4:50

you know, getting to the grid and then going back to the reviews.

4:52

So I love that Monty just consolidates everything to your point of using AI to

4:56

make this a really easy buying experience.

4:58

And a lot of times people don't really know where to start.

5:00

So I love that it is able to just, you can prompt it with questions.

5:03

It's not just buttons that you have to choose from.

5:05

It's actually retaining that answering the question.

5:08

So this is really incredible.

5:09

I have to imagine people are quite enjoying this feature.

5:12

>> Yeah, sure.

5:13

And then we certainly had a lot of fun building it.

5:15

And, you know, we had so much fun building it that we started asking,

5:18

what else can we do with conversation with AI?

5:21

And so within a few weeks of watching the recommender,

5:24

we connected Monty to our documentation knowledge base.

5:27

And that's where support Monty was born.

5:29

So support Monty lives in our Modg2 admin dashboard and can answer a wider age

5:34

of questions,

5:34

like how to update your GTU profile, set up integrations,

5:37

starting your GTU review campaign.

5:39

So we can take a look there as well.

5:41

And here's our cute little Monty Gu's cartoon.

5:44

So here's Monty, the friendly AI assistant for G2.

5:46

And let's just say we want some help updating my profile banner.

5:52

Maybe I went through recently brand and I don't remember exactly where I'm my G

5:55

to find that information.

5:56

So we'll go ahead and search through our entire documentation knowledge base.

6:00

And you can see that Monty will start off with information where to go in my G2

6:04

what UI elements to interact with.

6:07

And then also include a link to our documentation.

6:10

So users can learn more.

6:12

Oh, that's awesome.

6:13

I was our admin for G2 for quite a while when I first joined Qualified.

6:17

And so I spent a lot of time in the back end of G2,

6:19

like updating things to your point, updating product descriptions and banners

6:22

and review campaigns.

6:24

And I know I had a fantastic customer success manager and I used to bug her all

6:28

the time

6:29

with these questions.

6:30

So I'm sure, like, I see this, I'm like, oh, I would have used this all the

6:33

time.

6:33

Like, this is such a good user experience because there were questions like

6:36

that.

6:36

We're like, oh, I'm not in here every day.

6:38

You know, it's every couple of months.

6:39

I forgot where the banner is or, you know, how to update my CTA button on my G2

6:44

profile.

6:45

So this would have been really helpful.

6:48

And I'm awesome that customers have the opportunity to use this.

6:50

Yeah.

6:51

Yeah.

6:52

We're really happy with it.

6:53

And, you know, one of the reasons we weren't able to move so quickly after

6:58

launching the

6:58

recommender to developing support Monty is that we were at the same time as we

7:03

were developing

7:03

these two bots working on this internal no-code bot building platform that

7:07

would allow us

7:08

to easily create and refine our AI agents for whatever use case we wanted to

7:12

tackle

7:13

next.

7:14

And it made it really easy to start tweaking that functionality and kind of

7:16

hone in on

7:16

what we want to do and iterate really rapidly.

7:19

And so it wasn't long before we started asking, you know, how can we leverage

7:24

this platform

7:25

to bring buyers and sellers even close together?

7:29

And, you know, there's only so much information, for example, that can fit on a

7:33

single G2 profile

7:34

page.

7:35

And as a buyer, how valuable would it be if you could have your detailed

7:38

product questions

7:39

answered as you're researching products and comparing them to other products on

7:44

G2.

7:44

And as a seller, how great would it be if you can have a sales representative

7:47

always

7:47

ready to answer those questions 24/7, especially while this buyers are

7:52

researching your product

7:53

on G2.

7:54

So another version of Monte I'd love to show you is one that we launched a beta

7:59

program

7:59

for in August, what we're currently calling Monte for Sales, which enables

8:03

customers to

8:04

use our platform to create their own version of Monte trained on their

8:08

organization's sales

8:10

and marketing content to educate those high-end G2 buyers, push them down the

8:14

funnel and ultimately

8:16

sell their software autonomously.

8:18

Yep.

8:19

Well, kind of in the spirit, tricking around champagne.

8:21

Let's take a look at the Monte for Sales, but we have on our own G2 marketing

8:24

solutions

8:25

profile page.

8:26

So here in the bottom right corner, we have a chat with an expert bubble.

8:29

We'll go ahead and interact with Monte.

8:32

And Monte introduces himself as a friendly AI sales assistant for G2 marketing

8:37

solutions.

8:38

So let's ask Monte to tell me about customers who have seen success with G2.

8:49

And while the general product recommender will exclusively use G2 data to

8:53

interpret a user's

8:54

needs and direct them to the right solution, Monte for Sales lets you upload

8:58

your organization's

8:59

own internal and external content and allow Monte to use that content much like

9:04

a member

9:04

of your own sales team.

9:06

So we uploaded transcripts from our best sales calls and webinars, battle cards

9:10

, and as you're

9:11

seeing here, case studies.

9:14

And I really love how Monte approaches this question in particular, because I

9:18

asked for

9:19

some examples and Monte says, you want examples?

9:21

I got multiple stuff around the stores and even including those publicly

9:25

accessible URLs

9:27

to our case studies, so buyers can learn more.

9:30

All right.

9:31

I love that once he has the thumbs up and thumbs down, I think those are like

9:34

such

9:35

small features that make a huge difference.

9:37

I think in like training and making sure his answers are right is that giving

9:40

users the

9:40

option to say like, yeah, this was helpful or wasn't.

9:43

And I'm assuming all of that is ingested to help train Monte on the back end.

9:46

It sure is.

9:47

Yep.

9:48

Yep.

9:49

And so, you know, as a buyer, I might want to now ask Monte more detailed

9:54

product question.

9:55

So something like does G2 provide information about buyers researching my

10:03

product and competitors?

10:06

All right.

10:09

And Monte is again searching through the content that we uploaded to generate a

10:13

relevant response.

10:13

And his knowledge is so encyclopedic that some of our own internal reps will

10:17

actually

10:18

chat with Monte as they're preparing for a sales call to help them find the

10:21

relevant

10:22

case studies and success stories and create that tailored value proposition for

10:26

the persona

10:26

they're speaking to for them.

10:28

And so that totally makes sense.

10:31

Like, if you think about buyers, like our sellers, if they're, you know, we've

10:35

got all

10:35

of our data in some warehouse like a guru where they have to go search through,

10:40

if they

10:40

have to go through and like dig through all these different folders, it's so

10:43

much easier

10:43

just to ask the conversational AI and say, Hey, like, can you go, you're going

10:46

to get

10:47

through the law faster than I am, like you're built to read it, go grab that

10:50

information

10:50

and bring it in so I can see where absolutely sellers would use this for their

10:55

own benefit

10:56

as well as buyers.

10:57

Yeah.

10:58

Yeah.

10:59

Yeah.

11:00

It is funny as we've been exploring a beta program, we do see a lot of sellers

11:04

internal

11:04

usage for that reason.

11:06

They give the persona and then Monte finds the case studies and has that

11:10

tailored value

11:10

pitch kind of instantly.

11:13

And so in this case for this question, Monte correctly identified the feature

11:16

we're talking

11:17

about is buyer intent data.

11:19

Tell us a little bit more about the successes that some customers are seeing

11:21

with buyer intent

11:22

data and then what more information to provide.

11:26

And so Monte loves keeping the conversation going, but I'm kind of sold here.

11:30

So let's say that's, that's great.

11:33

Can I book a demo?

11:37

And very similarly to the recommender example we saw earlier, Monte's in a

11:41

second, you can

11:42

go ahead and inject the GTL form into JET.

11:50

And once again, it's pretty popular with my information to make that process

11:54

even easier.

11:55

So we're pretty excited about it.

11:57

Very cool.

11:58

I think to your point again, like we have a GT profile page that we maintain a

12:01

qualified

12:02

that we can't have all of our information on there.

12:04

Like there's so much product information if people have really specific

12:07

questions that

12:07

we don't necessarily have listed in like the descriptions or in the case

12:11

studies or what

12:12

is being able to feed that information to Monte for sales in the back end, make

12:16

sure

12:16

it has all those answers.

12:18

I will also say we love anyone that comes to our website from G2.

12:22

Like as far as leads go or referring sources like our reps know what it's from

12:26

G2 like

12:26

that's our most high intent potential buyers.

12:29

And I feel like this is only going to make it even better because they've

12:31

already asked

12:32

those questions.

12:33

It takes out a lot of the work for a sales reps like it's doing a lot of that

12:35

work they

12:36

were going to have to do when they came to our website.

12:38

By the time they're there, they're ready to buy.

12:40

So this is really cool.

12:42

Yeah, there's that pre-qualification, education aspect that our beta

12:46

participants are really

12:47

loving.

12:48

Yeah.

12:49

And so if we have some time, I'd love to give you a quick demo of our self-

12:52

service platform

12:53

that customers are using to add content and access performance metrics.

12:57

So we'll go, yeah, Monte for sales privilege and my G2, where we currently have

13:02

two tabs,

13:03

configuration and performance.

13:04

So on the G2 Labs team, we believe in shipping early and shipping often.

13:08

So we have an overall these pages coming within the next couple of weeks that's

13:11

going to improve

13:11

the functionality look and feel these pages.

13:13

But let's just start with a look at the content upload process.

13:17

So I'll grab one of these links, just so I have something for demo purposes.

13:21

And then we'll hit add content where I can drag and drop local files like PDFs,

13:26

Wordocs

13:26

and a bunch of other support, bio types, or I can just paste in a list of URLs.

13:31

So let's just add that URL to Scrabbed and hit Create Content.

13:36

You can see Monte correctly identifies the content that we uploaded as a case

13:39

study and

13:40

begins processing that content, which in a few moments it'll finish, and

13:44

immediately

13:44

be available for the chat bot to use in conversations.

13:48

And we can even evaluate the impact of that content on conversations directly

13:51

from the

13:52

interface by hitting Preview Chat.

13:57

And on the reporting side of things, let's click on the performance tab.

14:00

So let's go ahead and do a second to load.

14:02

And we start off with some high level performance metrics for Monte, including

14:05

the total number

14:06

of conversations and the total number of links clicked throughout the chat

14:10

history.

14:11

And we're also offering the ability to read anonymized conversation with buyers

14:17

So here's my chat that we saw earlier, which not only helps our customers

14:21

buying tune Monte's

14:22

performance as you're kind of deciding what content to use, but also provides

14:27

insights

14:27

into the kinds of questions that buyers are asking as they're researching your

14:31

product

14:31

on G2.

14:32

So that's pretty much my demo.

14:35

This is such a cool feature for two use cases immediately come to minus one to

14:38

your point.

14:39

Like you could go in here as the admin of G2 and look and say, Oh, shoot,

14:42

someone asks

14:43

this question.

14:44

I have documentation for that.

14:46

I forgot to give that to Monte because maybe we just changed something or we

14:48

had a new product

14:49

release.

14:50

And it's sort of this like good reminder, but to the what people are asking my

14:53

immediate

14:54

use cases of marketers, I'm like, Oh, I could use this to fuel what content I

14:57

need to create.

14:58

So people are asking a bunch of questions and I realize, Oh, we don't speak to

15:01

that very

15:01

often in any of our product documentation or like I should start creating top

15:06

of funnel

15:06

assets that answer these questions that people are asking.

15:09

So I feel like not only from a product marketing side is like, okay, I need to

15:12

make sure I'm

15:12

like feeding this all the relevant information, but from a like a content

15:16

marketing side,

15:16

I can use this to fuel that content.

15:18

So a lot of really fun use cases for a go to market team here.

15:22

Exactly.

15:23

Tim, this was such a cool demo.

15:25

I love Monty, not only because he's a cute monger, but I think that this is

15:28

just such

15:28

a good like buyer experience to good.

15:31

It's good for both sides of this, like it's making a great experience for those

15:33

that are

15:33

coming to look and learn more, but also for the companies.

15:36

Do you choose always been a great platform to use?

15:38

And this is just making it so much better for users.

15:41

So thank you for taking us through that demo.

15:43

Now if you're ready, I would love to go into our lightning round Q and A and

15:47

have you answer

15:48

a couple of questions if that's okay.

15:49

We're ready.

15:50

Let's do it.

15:51

So the first one is how long have you been building AI into G2?

15:55

Yeah.

15:56

So the concept of AI has gone through a bunch of kind of its own rebrand over

16:00

the years,

16:01

and Zales catch all for what we used to call things like machine learning or

16:05

natural language

16:05

processing.

16:06

But G2 has been pretty heavily invested in this space for a few years, both

16:10

internally

16:10

and externally.

16:11

So for example, we have internal models that help us to think things like surf

16:16

acing the

16:17

right software categories when new products are submitted to G2.

16:21

We have buyer facing models that generate pros and cons for products based on G

16:24

2 review

16:24

data.

16:25

And then we also have customer facing models that provide marketing and teleg

16:28

ance data

16:29

based on trends and review responses.

16:32

But like I'm sure you've heard most people who come on the show say having

16:36

access to

16:37

the newest generation of large language models via an API has made it so much

16:40

easier to test

16:41

if any given problem is worth solving with AI.

16:45

And that abstraction really helps us kind of make that value judgment if we

16:50

should just

16:51

use an off the shelf foundation model or build something ourselves.

16:55

Yeah, that's awesome.

16:57

Next question is what you showed today with Monty.

16:59

Is this generally available now for customers?

17:02

Yeah, so the product recommendation system intended to help software buyers is

17:07

in GA

17:08

as is support Monty for customers.

17:10

Our new Monty for sales product is in limited beta right now.

17:13

We have about 75 customers live with Monty for sales on their profile, but we

17:18

plan to move

17:18

to an open beta within the next month or so.

17:21

Oh, that's amazing.

17:22

And speaking of customers, can you speak to some of the success that your

17:25

customers are

17:25

seeing with any of those Monty for support, Monty for sales?

17:29

Anything you want to speak to from a customer standpoint there?

17:31

Yeah.

17:32

So the response to the Monty for sales beta so far has been kind of amazing.

17:38

When we get on calls, customers keep using the word impressive to describe

17:42

performance.

17:43

And that makes us feel all warm and fuzzy inside.

17:46

But quantatively, we're seeing some customers 10x their G2 leads.

17:50

Short in the sales cycle and Monty has actually already closed several deals,

17:54

like completely

17:55

on his own through cell service, which is really amazing.

17:59

And from the inside side, our customers love being able to see what buyers are

18:02

asking as

18:03

they're researching their product on G2.

18:05

Kind of like what you mentioned, which is generally a totally different stage

18:08

of the funnel than

18:08

when our customers usually get to talk to their prospects.

18:12

And they've been using these insights to fine tune their marketing narratives

18:16

as a result.

18:16

Absolutely.

18:17

Okay.

18:18

And the last question is, do you mind sharing a little bit more is what's next

18:21

for G2 and

18:22

AI or Monty?

18:23

Like what's on your roadmap?

18:24

How are you guys thinking about AI within G2 in the future?

18:27

Yeah.

18:28

So first one of the cool things about our labs team is that we don't have a

18:32

roadmap for more

18:33

than a few weeks app, which allows us to run tests and be really responsive to

18:37

the market.

18:38

But I'd be sure a few things we're thinking about both in labs with Monty and

18:42

across G2.

18:44

So for Monty, we're going to be focusing on enhancing reporting to make sure

18:48

customers

18:48

have already fully get insights on how Monty is using their content, as well as

18:52

generating

18:53

themes from conversations.

18:56

And one of the most frequent questions we get from beta participants is, hey,

18:59

can I

18:59

bet Monty on my own website landing pages?

19:01

And so we have a few customers testing this as well.

19:04

But in general across G2, we're also looking at ways we can use AI for quality

19:10

of life

19:11

improvements across our product lines, like making it easier for reviewers to

19:15

turn their

19:15

experience with the product into high quality reviews, making it easier for

19:19

sellers to take

19:20

action and respond to those reviews, and also further improve our search and

19:24

recommendation

19:25

process so buyers can get a truly personalized experience and find exactly the

19:28

right product

19:29

for their needs.

19:30

Yeah.

19:31

That's awesome.

19:32

I heard someone say that they don't have more than a couple of week roadmap,

19:35

which I have

19:35

to imagine.

19:36

You kind of said earlier, your guys are testing and shipping frequently and

19:39

often.

19:39

So that is a really cool concept, especially for something, like you said, AI

19:43

is, it's

19:44

not new, but the access we now have to opening eye and to these models is

19:48

making it move

19:50

so fast.

19:51

I love that you guys are able to be that agile.

19:53

I guess as you guys are deciding like what to do and when to ship, that's, that

19:57

's amazing.

19:58

Tim, thank you so much for joining the show.

20:00

It was great to have you on and to see a demo of G2 and of Monty.

20:05

There are always a pleasure.

20:06

I love the show and thanks for having me.

20:08

(amazing background music)