Artwork for podcast The Operations Room: A Podcast for COO’s
92. How to Really Use AI in Your GTM Team
Episode 9119th February 2026 • The Operations Room: A Podcast for COO’s • Bethany Ayers & Brandon Mensinga
00:00:00 00:47:10

Share Episode

Shownotes

In this episode we discuss: How to really use AI in your GTM team. We are joined by Donna McCurley, Creator of the AI Sales Operating System™ (AiSOS).

Love The Operations Room? Please support us by rating and reviewing it here.

We chat about the following with Donna McCurley:

  1. Who should actually “own” AI inside your GTM function — and what happens if nobody does?
  2. What does real AI governance look like in practice — beyond policies and buzzwords?
  3. Are AI agents creating hidden shadow systems inside your organisation?
  4. Why are most AI rollouts in sales failing to drive measurable revenue impact?
  5. How do you move from AI experimentation to a true AI sales operating system?

References

  1. https://www.linkedin.com/in/donnamccurley/

Biography

Donna McCurley is the creator of the **AI Sales Operating System™ (AiSOS)** and a go-to advisor for SaaS revenue teams who want to turn AI from “extra noise” into a real growth engine. She leads Global Sales Enablement teams and has helped 100s of sellers cut out busywork, triple their pipeline coverage, and adopt AI workflows that actually stick.

With a background that spans classroom teaching, Fortune 500 enablement leadership, and consulting with companies like McAfee and HPe, Donna has built a reputation for stripping away complexity and focusing on what truly drives revenue. Her work blends proven sales methodologies with practical AI deployment—think Microsoft Copilot, ChatGPT, Gong, and SalesLoft stitched together into one coherent system sellers can actually use.

At her core, Donna is obsessed with solving a simple but critical problem: too many sales teams are drowning in admin work, bloated tech stacks, and outdated playbooks. Through AiSOS, she’s building a future where sellers spend less time clicking around and more time closing deals.

To learn more about Beth and Brandon or to find out about sponsorship opportunities click here.

Summary

07:05 – Turning Copilot into a Revenue Engine

07:15 – AI Governance in Practice

07:18 – Avoiding “Shadow AI” Systems

09:27 – Change Management & Behaviour Shift

13:19 – The Biggest Mistakes in AI Rollout

14:16 – Enablement Over Experimentation

22:00 – The Role of RevOps as AI Overseer

38:00 – Moving from AI Tools to an AI Operating System



This podcast uses the following third-party services for analysis:

Podcorn - https://podcorn.com/privacy

Transcripts

Speaker:

Hello and welcome to another episode

Speaker:

of the Operations Room, a podcast

Speaker:

for COOs.

Speaker:

I am Brandon Mencinga joined by

Speaker:

Bethany Ayers.

Speaker:

How are things going this morning?

Speaker:

Back in the first week of January,

Speaker:

back at work.

Speaker:

It's so hard.

Speaker:

I wish I could be like, oh, I'm

Speaker:

cheery and I'm great and everything

Speaker:

is awesome, but that

Speaker:

week back, I don't know, hopefully

Speaker:

it's not going to be all of January

Speaker:

and it's just the first week, but

Speaker:

I'm tired.

Speaker:

It doesn't help that I went to a

Speaker:

dinner last night.

Speaker:

I clearly have no stamina anymore.

Speaker:

So I'm physically tired.

Speaker:

I'm emotionally exhausted and

Speaker:

not my normal bubbly self.

Speaker:

How about you?

Speaker:

You know, I just realized, I was

Speaker:

thinking about this this morning,

Speaker:

like the actual new year is in

Speaker:

spring, to be honest, because spring

Speaker:

is when, you know, you get more

Speaker:

daylight, it feels better.

Speaker:

There's like a spring in your step.

Speaker:

You're thinking about the future,

Speaker:

you're optimistic.

Speaker:

The new year is not January 1st,

Speaker:

because it turns January 1, we're

Speaker:

forced by the calendar and suddenly

Speaker:

we're supposed to feel like, yeah,

Speaker:

it's a new year. It's a New Me.

Speaker:

It's resolutions and it is deadly

Speaker:

cold outside.

Speaker:

It is dark.

Speaker:

You're being forced back

Speaker:

into work off your comfy couch

Speaker:

and it doesn't feel like there's a

Speaker:

spring in your step.

Speaker:

Yeah, I 100% agree.

Speaker:

And so the one good thing is that

Speaker:

we've had the solstice and the sun

Speaker:

is, the days are lengthening from

Speaker:

here on out, but you don't

Speaker:

notice the difference until

Speaker:

February.

Speaker:

January is just dark and

Speaker:

grim, and

Speaker:

we've been pushing hard since

Speaker:

the sun disappeared sometime in

Speaker:

November.

Speaker:

So what's happening on the work

Speaker:

front? Anything noteworthy?

Speaker:

The really big news is I have a new

Speaker:

sales leader and new marketing

Speaker:

leader starting next week.

Speaker:

Okay, that's big news.

Speaker:

Cannot wait.

Speaker:

Sales and marketing, both starting

Speaker:

in the same week.

Speaker:

Okay, well that's actually pretty

Speaker:

ideal.

Speaker:

Well, they were supposed to start on

Speaker:

the same day, but because of

Speaker:

calendaring issues, we have one

Speaker:

starting Monday, one starting

Speaker:

Tuesday.

Speaker:

So yeah, that must be thrilling.

Speaker:

You're like, oh my God.

Speaker:

It really is.

Speaker:

And also, there were long notice

Speaker:

periods and so

Speaker:

there's been lots

Speaker:

of pre-onboarding so it's not

Speaker:

like they're going to have to spend

Speaker:

months learning.

Speaker:

They probably are not thanking me

Speaker:

for the Google Drive

Speaker:

that I created for them as

Speaker:

their Christmas reading.

Speaker:

But do you have Gemini?

Speaker:

So Gemini, I find,

Speaker:

does research that's like dealing

Speaker:

with a partner analyst,

Speaker:

everything Gemini produces is

Speaker:

super dense.

Speaker:

Has no personality, but I

Speaker:

kind of have like faith in the

Speaker:

amount of research.

Speaker:

But Nano Banana now does

Speaker:

infographics.

Speaker:

And so I'll just take the

Speaker:

really dense information and then

Speaker:

ask it to produce an infographic and

Speaker:

they are fantastic.

Speaker:

And so for

Speaker:

my go-to-market leaders, they have

Speaker:

all of the like super dense

Speaker:

information and then they have

Speaker:

five infographics and I

Speaker:

suspect all they need to do is look

Speaker:

at the infographics, and that sorts

Speaker:

everything out.

Speaker:

But Nano banana was having some sort

Speaker:

of massive issues.

Speaker:

So each infographic took about three

Speaker:

days to produce.

Speaker:

And it just left it and it looked

Speaker:

like it was still producing.

Speaker:

It's like, well, let's see.

Speaker:

And I would check in every day and

Speaker:

then on the third day that arrived.

Speaker:

That's crazy.

Speaker:

So the compute time, sorry, I'm just

Speaker:

like trying to think this route.

Speaker:

So the computer time to create your

Speaker:

infographic took three days.

Speaker:

Is that what you're telling me?

Speaker:

Yeah, but luckily I could run all

Speaker:

five at the same time.

Speaker:

Wow, I'm just imagining the compute

Speaker:

center and Albuquerque

Speaker:

cranking away, consuming

Speaker:

vast amounts of power

Speaker:

to power your infographic.

Speaker:

Well, but I mean, clearly it's not

Speaker:

my infographic that's caught, you

Speaker:

know, I'm in a big queue and I'm

Speaker:

getting one, one billionth

Speaker:

of the compute power, but it's

Speaker:

taking three days to do it,

Speaker:

but highly recommend infographics.

Speaker:

So I am super excited for the

Speaker:

offsite that is coming up next week

Speaker:

for us. So we have our plan of

Speaker:

action. We've got day two as

Speaker:

you kind of do these things.

Speaker:

So day one is all around the company

Speaker:

and our strategy and alignment and

Speaker:

buy-in and all that jazz.

Speaker:

Excitement energized for the new

Speaker:

year. Day two is all

Speaker:

about the AI upscaling.

Speaker:

So we've got three tracks in the

Speaker:

morning. We've Got our friend

Speaker:

Charlie Cowan coming in to

Speaker:

facilitate the entire day, but also

Speaker:

doing his bit on chat GPT in the

Speaker:

morning for his track.

Speaker:

We've got Johnny Ball coming in

Speaker:

for lovable, for non-developers,

Speaker:

which I think is going to be super

Speaker:

awesome for the company.

Speaker:

And then this other fellow named

Speaker:

David, who's much more of

Speaker:

like a real developer of developers.

Speaker:

He's actually CEO of his company,

Speaker:

but he is a developer at heart and

Speaker:

he's very ingrained in getting the

Speaker:

best out of AI for the development

Speaker:

side in his company.

Speaker:

So he's coming in too.

Speaker:

Do the cursor training in this case,

Speaker:

do that three tracks, and

Speaker:

then in the afternoon we have six

Speaker:

challenges.

Speaker:

And we're going to have people self-select

Speaker:

into the tracks in the morning,

Speaker:

self- select into the challenges,

Speaker:

kind of normalize the group sizes

Speaker:

for the challenges in the afternoon,

Speaker:

and go at it for three hours,

Speaker:

produce a couple of demos of

Speaker:

interest that we can present back to

Speaker:

the company, and you know,

Speaker:

fingers crossed it works well.

Speaker:

I have a lot of confidence in

Speaker:

Charlie to kind of facilitate the

Speaker:

arc of the day to make it all

Speaker:

sensible and kind of drive

Speaker:

it forward.

Speaker:

And he has just so much energy and

Speaker:

enthusiasm that he gets everybody

Speaker:

interested.

Speaker:

It's amazing if you just think about

Speaker:

the energy givers versus the energy

Speaker:

takers, and as leaders, we need to

Speaker:

be energy giver a lot.

Speaker:

And sometimes I naturally have that

Speaker:

energy giving and sometimes I don't

Speaker:

like this week, whereas Charlie

Speaker:

seems to just always be

Speaker:

able to enthuse

Speaker:

others and give energy.

Speaker:

That's really authentic and also

Speaker:

a great communicator can really

Speaker:

simplify the messages in a way

Speaker:

that's really I think for an

Speaker:

audience of varying degrees

Speaker:

of skill I guess when it comes to

Speaker:

AI you can really kind of hit.

Speaker:

Yeah, so our

Speaker:

training we're actually moving away

Speaker:

from chat GPT and focusing

Speaker:

on Claude skills and

Speaker:

so learning how to build

Speaker:

automations and better

Speaker:

outcomes via Claude.

Speaker:

So that's what our training is

Speaker:

about.

Speaker:

Yeah, I feel like we're one step

Speaker:

behind you in this respect.

Speaker:

I suspect in our next go round, it

Speaker:

might be something similar.

Speaker:

I think the thing I'm most excited

Speaker:

about is the lovable

Speaker:

non-developer track because I think

Speaker:

for 2026, I think

Speaker:

in earnest with our CTO behind it

Speaker:

as well, we want to put rest of the

Speaker:

company in a position to ultimately

Speaker:

develop. Actual code that

Speaker:

gets injected into the product,

Speaker:

obviously, with quality oversight

Speaker:

from actual real developers before

Speaker:

anything goes into production code,

Speaker:

obviously. But I think that ability

Speaker:

is going to be phenomenal.

Speaker:

In particular, we've got one person

Speaker:

that's setting up our new

Speaker:

propositions team where she's kind

Speaker:

of going early into markets that

Speaker:

we're not currently in, and her

Speaker:

ability to work with those

Speaker:

individuals, those prospects and

Speaker:

customers, and develop some

Speaker:

prototypes and try to flex the

Speaker:

muscle of what actually is

Speaker:

compelling to these people and do

Speaker:

that using Lovewell as an example, I

Speaker:

think is going be a super obvious

Speaker:

entry point for an individual that

Speaker:

is not a developer to create real

Speaker:

value in the company.

Speaker:

So she's super excited, I'm super

Speaker:

excited. I think it's going to be

Speaker:

fabulous. We'll see how it plays

Speaker:

out.

Speaker:

We've got a great topic, which is

Speaker:

how to really use AI in your GTM

Speaker:

team. We have an amazing guest for

Speaker:

this in Donna McCurley.

Speaker:

She is the creator of the AI sales

Speaker:

operating system and a go-to-market

Speaker:

advisor for mid-market SaaS teams,

Speaker:

turning Microsoft Copilot 365 into

Speaker:

revenue engines.

Speaker:

So Donna had talked about someone

Speaker:

needing to be the AI overseer

Speaker:

and she had pointed to rev

Speaker:

ops as part of that.

Speaker:

In your view, what does governance

Speaker:

actually look like in practice?

Speaker:

So agents don't become this shadow

Speaker:

system.

Speaker:

So there's a lot in

Speaker:

governance that we should be looking

Speaker:

at. So one level is, and

Speaker:

this is where I'm going to talk

Speaker:

about Matomic, is what

Speaker:

data do you have and who has access

Speaker:

to that data, and who

Speaker:

slash what has access to that date,

Speaker:

so humans and

Speaker:

agents.

Speaker:

And so you need to do a certain

Speaker:

amount of data

Speaker:

categorization and

Speaker:

identify what is sensitive

Speaker:

information that should not be fed

Speaker:

into your models and what is

Speaker:

information that's safe to go into

Speaker:

them. Or be used by your

Speaker:

agents.

Speaker:

And somebody like us can help you

Speaker:

with that. Then you also need to

Speaker:

identify and decide what your

Speaker:

policy is on shadow

Speaker:

AI or not.

Speaker:

And depending on your risk appetite,

Speaker:

you can completely

Speaker:

lock it down, have whitelists and

Speaker:

blacklists of what URLs people are

Speaker:

able to access and get quite

Speaker:

draconian with it.

Speaker:

You can decide.

Speaker:

The reason why people are using

Speaker:

shadow AI is the AI tools.

Speaker:

You're either not providing tools,

Speaker:

which at this point, if you're not

Speaker:

providing them, I think you're in

Speaker:

trouble, or the tools that you are

Speaker:

providing are just subpar.

Speaker:

That would mostly be co-pilot,

Speaker:

but it just doesn't work as well

Speaker:

because it's so locked down compared

Speaker:

to Gemini, Claude, and

Speaker:

ChatGPT.

Speaker:

So again, your risk appetite,

Speaker:

maybe need to loosen it up a bit,

Speaker:

but make sure if your data is in

Speaker:

a good place.

Speaker:

You can loosen up your controls,

Speaker:

bring in something that's not

Speaker:

Microsoft to help you.

Speaker:

And then you can

Speaker:

start to think about agents.

Speaker:

And by agents on the desktop, I mean

Speaker:

like clients, not AI agents.

Speaker:

So agents that recognize what's

Speaker:

going on and what your employees are

Speaker:

doing. Employee spyware, which

Speaker:

is a very popular in bigger

Speaker:

companies in startups

Speaker:

and scalps. We don't have to tend to

Speaker:

deal with it.

Speaker:

Or you can do something more

Speaker:

lightweight like browser

Speaker:

plugins that just scan.

Speaker:

All of the text boxes that you

Speaker:

have that your employees are putting

Speaker:

text into before they

Speaker:

do it and either block it or

Speaker:

just alert that there are employees

Speaker:

that are behaving against

Speaker:

policy and then you can address it

Speaker:

with behavior change.

Speaker:

That's with the CSO hat on of like

Speaker:

super locking it down.

Speaker:

I'm guessing our audience isn't

Speaker:

worrying as much about that.

Speaker:

They're fighting their so to loosen

Speaker:

things up, which is then around

Speaker:

policy and trust and

Speaker:

balancing.

Speaker:

Data security risk with

Speaker:

falling behind and not using

Speaker:

AI at all, an existential risk,

Speaker:

and trying to decide which way you

Speaker:

need to go.

Speaker:

Yeah, that makes sense.

Speaker:

And then do you have an opinion on

Speaker:

this revops ownership thing?

Speaker:

So there's a fleet of agents within

Speaker:

the company.

Speaker:

I think the rev-ops should be

Speaker:

responsible for rev-op.

Speaker:

It's agents but not others.

Speaker:

One of the challenges with

Speaker:

agents going forward is that there's

Speaker:

a lot of cross-functional

Speaker:

work and so I suspect over time

Speaker:

there's going to be some sort of AI

Speaker:

agent office that's

Speaker:

looking at it across the board

Speaker:

rather than specifically within

Speaker:

rev-ups. A lot what Donna was

Speaker:

talking about is.

Speaker:

Specifically within the revenue

Speaker:

function. And so then it makes sense

Speaker:

for rev ops to look at it, but

Speaker:

there's gonna be a point where it

Speaker:

starts to touch finance quite

Speaker:

a bit.

Speaker:

And then it's beyond rev ops

Speaker:

control. And anytime you're actually

Speaker:

talking about cash collection and

Speaker:

cash reporting, then I think

Speaker:

somebody in finance needs to own it.

Speaker:

So the other bit that she had spoken

Speaker:

about was this opinion that on

Speaker:

the AI-SDR front that

Speaker:

the split that she

Speaker:

sees being the optimal split between

Speaker:

AI doing a lot of work on behalf

Speaker:

of the SDR is a 70-30

Speaker:

split whereby 70% is

Speaker:

agent automation and 30% is the

Speaker:

SGR doing the human in

Speaker:

the loop kind of like.

Speaker:

Points of qualification of the

Speaker:

AI agent outputs, I suppose,

Speaker:

in this case, and also to some

Speaker:

extent still doing the live calls

Speaker:

with Outbound,

Speaker:

what's your take on that 70-30

Speaker:

split?

Speaker:

Yes, we ended up talking a bit

Speaker:

around what is the role of an SDR

Speaker:

and what are the new skills that

Speaker:

you're looking for when hiring SDRs.

Speaker:

And I do think this is one of the

Speaker:

roles that's changing most rapidly,

Speaker:

where the people that you are

Speaker:

looking for are way more systems

Speaker:

thinkers.

Speaker:

They're systems thinkers who are

Speaker:

motivated by money is

Speaker:

I'd say the profile that youre

Speaker:

looking for.

Speaker:

It's almost like an ops person that

Speaker:

you're Looking for for SDR

Speaker:

is now because it's like somebody

Speaker:

who plus being motivated

Speaker:

money.

Speaker:

Likes to problem solve, and this

Speaker:

problem you're solving is how to

Speaker:

bring in good prospects, cares about

Speaker:

the problem because they like to

Speaker:

earn money, and are

Speaker:

systems thinking and curious around

Speaker:

technology.

Speaker:

It's not about the resilience

Speaker:

of hitting the phones and cold

Speaker:

calling anymore, although motivated

Speaker:

by money means that you're probably

Speaker:

willing to pick up the phone.

Speaker:

I completely agree, but there's

Speaker:

still, I think, isn't there like a

Speaker:

live outbound call by

Speaker:

a human to another human that still

Speaker:

needs to occur?

Speaker:

It always swings back and forth,

Speaker:

doesn't it? Because you'll end up

Speaker:

with so many cold calls.

Speaker:

Everybody screens it.

Speaker:

You can't get through to anybody.

Speaker:

Nobody answers.

Speaker:

And then everybody gives up on cold

Speaker:

calls, so then the numbers go down.

Speaker:

So then when you start doing it

Speaker:

again, people answer.

Speaker:

So it kind of depends where are you

Speaker:

in that pendulum as to whether or

Speaker:

not cold calls are effective.

Speaker:

What might be more effective for

Speaker:

right now, I don't know,

Speaker:

is, I mean, how many thousands of

Speaker:

emails do you get a day now,

Speaker:

Brandon? There are just so

Speaker:

many. And so many of them are.

Speaker:

Bad, like wrong company.

Speaker:

I got one the other day that was

Speaker:

like, we're so impressed by what

Speaker:

Matomic's doing in the

Speaker:

healthcare monitoring

Speaker:

space or something totally random

Speaker:

and went on and on and on.

Speaker:

And this is like, and then wrong

Speaker:

names, wrong companies.

Speaker:

But then through all of that mess, I

Speaker:

got a really good one that

Speaker:

I haven't talked to a human yet,

Speaker:

but are following through in a

Speaker:

process.

Speaker:

Yeah, and this may be where this

Speaker:

operator precision that you're

Speaker:

talking about comes in mostly.

Speaker:

Yeah. And also like,

Speaker:

what are you offering in your

Speaker:

outbound now?

Speaker:

So this one was LinkedIn

Speaker:

and how to use LinkedIn more

Speaker:

effectively for sales processes.

Speaker:

Had noticed a couple of mistakes

Speaker:

that we were making.

Speaker:

Can you send a Loom video explaining

Speaker:

the mistakes?

Speaker:

Okay. Yeah.

Speaker:

We invest quite a bit on LinkedIn.

Speaker:

We're not, you know, we can always

Speaker:

get better results.

Speaker:

And then I forwarded it around to

Speaker:

the company of like, this is a good

Speaker:

email. But I'm afraid to

Speaker:

ask for the Loom video because I

Speaker:

don't know what's going to happen

Speaker:

and what kind of like aggressive

Speaker:

sales I'm opening myself up to.

Speaker:

And then I was like, well, you know

Speaker:

what? Why should I be afraid?

Speaker:

I'll just ask for a Loom Video.

Speaker:

So now he sent it to me.

Speaker:

I have yet to watch it because I

Speaker:

have no time, but it was like

Speaker:

watch the video and then let me know

Speaker:

if you're interested in chatting

Speaker:

more.

Speaker:

So super lightweight,

Speaker:

not an aggressive push,

Speaker:

but giving me value from the very

Speaker:

first interaction.

Speaker:

Smart prospecting is going to make

Speaker:

more of a difference, whether it's

Speaker:

on the phone or not.

Speaker:

But also you might need to do it on

Speaker:

the phone because thousands of

Speaker:

emails every day.

Speaker:

I've been in sales enablement for

Speaker:

a little over 20 years,

Speaker:

and as a lot of us know, sales

Speaker:

enablement organizations typically

Speaker:

have to run pretty lean.

Speaker:

So being resourceful and figuring

Speaker:

out how to support an entire sales

Speaker:

organization is pretty critical.

Speaker:

You've got account managers,

Speaker:

you've got, from an AE perspective,

Speaker:

you got enterprise, commercial,

Speaker:

S&B, and then you've BDR.

Speaker:

So everybody has a whole lot

Speaker:

of different things that they need

Speaker:

to do. And so my

Speaker:

initial thoughts were,

Speaker:

how do I make sure that the

Speaker:

sales enablement organization can

Speaker:

support everyone?

Speaker:

And when AI came

Speaker:

along, it was like a dream come true

Speaker:

for me.

Speaker:

Because when I think about

Speaker:

AI, I think of building

Speaker:

agents as if I were hiring

Speaker:

another person in my department,

Speaker:

who will show up every single

Speaker:

day, they're never sick,

Speaker:

they do exactly what they're

Speaker:

told, and they worked 24-7.

Speaker:

As long as I teach

Speaker:

them how to do this.

Speaker:

So I built what I refer

Speaker:

to as AISOS,

Speaker:

which is simply an AI sales

Speaker:

operating system.

Speaker:

And at a high level, what it does

Speaker:

is when you look across sales

Speaker:

activities, or

Speaker:

your sales process, what is does

Speaker:

is it does exactly what

Speaker:

task a seller needs to

Speaker:

do. So when you,

Speaker:

hey, I'm in, you know, stage

Speaker:

one, and I'm doing, let's say

Speaker:

I need to do discovery.

Speaker:

What are all the things that a

Speaker:

seller needs to do during that?

Speaker:

You know, they may have to research

Speaker:

the account.

Speaker:

They may need to research The Buying

Speaker:

Committee.

Speaker:

They may to scrape for buying

Speaker:

signals.

Speaker:

All of this now is something

Speaker:

that an AI agent can do.

Speaker:

Whereas in the past, this could

Speaker:

take a lot of hours for sellers

Speaker:

to do. We know that sellers

Speaker:

like to skip steps and

Speaker:

so, you know, it's like, oh I can

Speaker:

wing it Well now a

Speaker:

seller can show up for that

Speaker:

discovery call completely prepared

Speaker:

with all that account information

Speaker:

buying committee buying Signals

Speaker:

and even outreach messages that

Speaker:

got them to that place and that's

Speaker:

why I developed it because I thought

Speaker:

how do I make Sure, I'm supporting

Speaker:

the sales organization so that

Speaker:

they're able to convert and

Speaker:

feel confident and comfortable

Speaker:

showing up The second

Speaker:

thing is we can look across

Speaker:

a lot of sales organization and a

Speaker:

lot of sales leaders have been

Speaker:

promoted into a sales manager

Speaker:

because they were good as a

Speaker:

seller themselves, but it doesn't

Speaker:

necessarily mean that they've been

Speaker:

trained on what it

Speaker:

means to be a real sales leader.

Speaker:

And so you see these sellers

Speaker:

who want to perform at an A

Speaker:

level, but they may not have a

Speaker:

sales leader who has the

Speaker:

competencies to help

Speaker:

them with maybe a skill that they're

Speaker:

struggling with or product

Speaker:

knowledge.

Speaker:

And so they're at a disadvantage

Speaker:

when that happens.

Speaker:

And then also sales leaders are

Speaker:

spread really thin.

Speaker:

They may have eight sellers that

Speaker:

they are trying to support.

Speaker:

And it's more like bring your deal

Speaker:

to me, yes, yes yes, go or

Speaker:

not really having that opportunity

Speaker:

to coach.

Speaker:

So AISOS does that.

Speaker:

It is your 24-7,

Speaker:

very highly competent, PhD

Speaker:

level sales coach.

Speaker:

So a seller can go in and

Speaker:

say like I'm struggling with

Speaker:

being able to you know

Speaker:

This client has went dark You know I

Speaker:

don't know what to do now and

Speaker:

they can actually type in those

Speaker:

questions and the AIS OS

Speaker:

will provide them with hey Here's

Speaker:

five ideas like which one resonates

Speaker:

with you and actually build out

Speaker:

like let's say I had a seller the

Speaker:

other day This was their exact

Speaker:

situation and you know

Speaker:

it provided them with the reverse

Speaker:

timeline Hey, in the discovery

Speaker:

call, your client,

Speaker:

Joe, said that this was their

Speaker:

timeline they were trying to make.

Speaker:

Let's reverse engineer that and

Speaker:

provide them with the timeline and,

Speaker:

hey, Joe.

Speaker:

If we're still going to try to make

Speaker:

that November timeline,

Speaker:

here's how this is going to look.

Speaker:

We probably need to move forward.

Speaker:

He was able to get Joe back

Speaker:

on that joke, back on a call.

Speaker:

So it's things like that that his

Speaker:

sales manager probably wouldn't have

Speaker:

been able to have done.

Speaker:

We tend to be very practical in

Speaker:

the podcast.

Speaker:

And what I love about your content

Speaker:

is you're

Speaker:

building all of this.

Speaker:

You're not buying a bunch of tools.

Speaker:

Like, can we talk a bit around how

Speaker:

are you actually doing it?

Speaker:

The post that I got really

Speaker:

interested in was when you took all

Speaker:

of your top sellers recordings,

Speaker:

analyze them and discovered what

Speaker:

was actually why you were actually

Speaker:

winning.

Speaker:

So rather than answer, maybe that

Speaker:

one is like.

Speaker:

How to get started for

Speaker:

somebody who doesn't have an A.I.

Speaker:

S.O.S. And doesn't have a Donna but

Speaker:

wants to have that in their

Speaker:

organization, what would you

Speaker:

suggest?

Speaker:

This is how I got started.

Speaker:

Is I went into

Speaker:

CoPilot, and when you go into

Speaker:

Copilot, there is a

Speaker:

AI agent, like it's really, really

Speaker:

intuitive. When you go in to the AI

Speaker:

agent I believe that there are

Speaker:

three things that you need to do.

Speaker:

The first thing that you to do is

Speaker:

you have to decide what

Speaker:

output you want that agent to

Speaker:

do, it's just as simple as, you

Speaker:

know, I want it to write an email,

Speaker:

like you are my Ellie

Speaker:

email. And so understanding what

Speaker:

you're trying to achieve with that

Speaker:

agent is first and foremost.

Speaker:

The second thing you do is you have

Speaker:

to give it instructions.

Speaker:

And we all do this when we

Speaker:

hire people. We give them

Speaker:

instructions.

Speaker:

Hey, here's how you write an email.

Speaker:

Now. I actually use

Speaker:

a framework to

Speaker:

help me rinse and repeat my

Speaker:

instructions over and over again,

Speaker:

which I like to teach.

Speaker:

And so being able to write

Speaker:

instructions is the second thing,

Speaker:

and it is simply, hey, how

Speaker:

do you do the task that you assign?

Speaker:

And then the third part of that

Speaker:

is the knowledge base.

Speaker:

And the knowledge base is you think

Speaker:

of it as best practices.

Speaker:

So what are some really good emails

Speaker:

or campaigns or writing

Speaker:

instructions that you have

Speaker:

and you take that and you put it

Speaker:

in the knowledge base.

Speaker:

That's it.

Speaker:

Like you're done.

Speaker:

You've given it an output, you've

Speaker:

given an instructions and you've

Speaker:

given it a knowledge base, now go

Speaker:

play with it.

Speaker:

Now go iterate and see what

Speaker:

did that produce?

Speaker:

What was the outcome?

Speaker:

Ooh, how can I go back and better

Speaker:

improve my instructions because

Speaker:

I want it to remove maybe the

Speaker:

hope you're doing well kind of stuff

Speaker:

or I want to use

Speaker:

a specific email framework,

Speaker:

things like that, and that's

Speaker:

the best way to get started.

Speaker:

So, I have a couple of questions.

Speaker:

One is the first way

Speaker:

of getting started is basically

Speaker:

thinking about how

Speaker:

your AI employee

Speaker:

can do stuff. So write emails for

Speaker:

you, help you unlock a deal,

Speaker:

the person that you go to and say,

Speaker:

do this specific task.

Speaker:

But what I always feel like is a

Speaker:

really big unlock is when it gets

Speaker:

starts to be tied in with

Speaker:

automation.

Speaker:

So when you're talking about the

Speaker:

beginning, the employee that works

Speaker:

24 hours a day, so your SDR

Speaker:

who's prospecting all day.

Speaker:

And I'm really struggling, and I've

Speaker:

been struggling all year, no matter

Speaker:

who I speak to, about this

Speaker:

AI automation true

Speaker:

agent and how to do

Speaker:

that.

Speaker:

Okay, I love that you asked that

Speaker:

because here's where I like to

Speaker:

say is this is my kind of soapbox

Speaker:

is I feel like everybody's

Speaker:

trying to get to

Speaker:

ninth grade way too fast.

Speaker:

So what I like to say, is that

Speaker:

if you have got to build out

Speaker:

your agent to

Speaker:

advance them from first grade to

Speaker:

second grade to third, you need

Speaker:

to get your agent, one agent

Speaker:

performing really well before

Speaker:

you ever start to automate it.

Speaker:

And so once you

Speaker:

have your agents automating

Speaker:

well, then let's say

Speaker:

the next part of this conversation

Speaker:

in a workshop, if we were doing a

Speaker:

workshop together, I would say,

Speaker:

hey, let's get your agents

Speaker:

that can offload non-revenue

Speaker:

producing work, the research,

Speaker:

the signal scraping, all that.

Speaker:

Let's get those built.

Speaker:

Now let's talk about how to

Speaker:

run a trigger.

Speaker:

Let's say we built four agents in

Speaker:

our first workshop.

Speaker:

Now let's take those four agents

Speaker:

and now let's talk about how we

Speaker:

could automate them Like how

Speaker:

do we create that trigger to

Speaker:

put them to just kind of run on

Speaker:

their own and I wake up Monday

Speaker:

morning And they're there.

Speaker:

Okay, so let's a real example

Speaker:

here You're let's

Speaker:

say you're an organization and my

Speaker:

first question in this workshop too

Speaker:

before we build out your you know

Speaker:

Your auditions is thinking about

Speaker:

okay. What are your growth levers?

Speaker:

And so most organizations they

Speaker:

will say like hey I have an

Speaker:

operational efficiency that I

Speaker:

need to pull, and I have a land

Speaker:

and expand for my account managers.

Speaker:

Okay, let's separate those

Speaker:

workshops. Let's only talk about

Speaker:

operational efficiency.

Speaker:

Let's not talk about operational

Speaker:

efficiency because, again, I'm going

Speaker:

to be a little bit selfish today on

Speaker:

what we're looking for in our

Speaker:

organization, and that is

Speaker:

pipeline generation.

Speaker:

So new business pipeline generation,

Speaker:

that's a part of it.

Speaker:

Operational efficiency is going to

Speaker:

be a part. I think you'll like this.

Speaker:

Awesome. Okay.

Speaker:

So I was just trying to figure out.

Speaker:

So yeah, I guess if I just say the

Speaker:

number one unlock for us

Speaker:

is pipeline generation.

Speaker:

If you look at that as the

Speaker:

most important thing.

Speaker:

How should we get started?

Speaker:

Ooh, okay, wait.

Speaker:

Let me pump the brakes.

Speaker:

So you're looking more

Speaker:

at your lead generation

Speaker:

than you are more

Speaker:

about the efficiency of

Speaker:

your pipeline.

Speaker:

Yes, so we don't have enough

Speaker:

pipeline to worry about how

Speaker:

efficient it is.

Speaker:

We just want more in the top and

Speaker:

then we can worry about the

Speaker:

efficiency later on.

Speaker:

Okay, well, because once we

Speaker:

get it in, we win 40%

Speaker:

of it.

Speaker:

And so but I think that's because

Speaker:

customers have to work really hard

Speaker:

to find us. And by the time they

Speaker:

find us, they want to buy us.

Speaker:

So I want to put more

Speaker:

companies that are interested.

Speaker:

So even if our win rate goes down,

Speaker:

we at least know that we're talking

Speaker:

to more and kind of planting seeds

Speaker:

for the future of other companies

Speaker:

who would like to work with us.

Speaker:

So your agents, what

Speaker:

comes top of mind for me is, your

Speaker:

agents would be, hey, if we were to

Speaker:

take a seed list, a list of

Speaker:

companies that have done business

Speaker:

with us in the past and have

Speaker:

had success, and then we wanna find

Speaker:

more of those.

Speaker:

So the first agent I would build

Speaker:

is gonna be that look-alike agent,

Speaker:

where we're taking that seed list

Speaker:

putting it in, saying, hey what's

Speaker:

not in our CRM, and being

Speaker:

able, hey now go out and find more

Speaker:

companies that are like this.

Speaker:

After that, now that you have,

Speaker:

let's just say you're a

Speaker:

targeted.

Speaker:

New opportunity list

Speaker:

so the next thing is because

Speaker:

you guys understand so much

Speaker:

historically about these companies

Speaker:

in the past is taking

Speaker:

those calls and doing kind

Speaker:

of what my post said taking those.

Speaker:

Calls and extracting buyer

Speaker:

intelligence from those is

Speaker:

first and foremost so now we

Speaker:

have an an agent that is

Speaker:

finding lookalikes and we

Speaker:

have looking at our past we

Speaker:

have agent that's going to find,

Speaker:

you know, buyer insights.

Speaker:

And so those buyer insights is I

Speaker:

want to know who are the people

Speaker:

that are on these calls?

Speaker:

What questions are they asking?

Speaker:

What objections are they running

Speaker:

an AI agent to

Speaker:

look for those types of patterns

Speaker:

is a goldmine.

Speaker:

And so when we go back

Speaker:

to writing the instructions, you're

Speaker:

saying, hey, I'm going to load, you

Speaker:

knows, calls from let's just take a

Speaker:

month, you don't from last month.

Speaker:

I'm going to load those in there,

Speaker:

and I want you to tell me who was

Speaker:

on the call, what questions they

Speaker:

asked, what objections they pulled

Speaker:

up, what did they like, what

Speaker:

pains did they, all the things that

Speaker:

we normally do, but

Speaker:

we want to take it from those real

Speaker:

live calls.

Speaker:

And then once we have that,

Speaker:

the third agent is going to go out

Speaker:

there and actually scrape for

Speaker:

those signals.

Speaker:

So that seed list that we

Speaker:

found, that we used to find

Speaker:

new clients, now I'm take

Speaker:

that seed list, and I'm going

Speaker:

to reduce it,

Speaker:

filter it by buying signals.

Speaker:

Because as we know, only 3% of the

Speaker:

market is ready to buy at any given

Speaker:

time.

Speaker:

So I want my sellers focused

Speaker:

on higher probability

Speaker:

deals.

Speaker:

So let's say once a week, I do

Speaker:

that automation for those

Speaker:

scraping for those buying signals,

Speaker:

those intent signals that I found in

Speaker:

my call extraction,

Speaker:

I'm go out and looking for those.

Speaker:

So it could be new businesses open,

Speaker:

it could new sales leadership.

Speaker:

I don't know what your signals are,

Speaker:

but those are the ones that I'm

Speaker:

scraping for and every Monday

Speaker:

morning my sellers wake up,

Speaker:

that call report is something that's

Speaker:

in their Slack channel or email

Speaker:

is in there for them to say,

Speaker:

hey, of my 100 account lists,

Speaker:

here's the ones that are actually

Speaker:

dealing with something.

Speaker:

So I'll pause there from

Speaker:

a lead generation perspective.

Speaker:

That sounds tremendous to me.

Speaker:

So I am now very curious, like,

Speaker:

what are we talking about here in

Speaker:

terms of the tech stack?

Speaker:

I'll be honest with you, I am 100%

Speaker:

using just Copilot for everything

Speaker:

that I just described.

Speaker:

Before I say this, I want to go

Speaker:

back in because everybody tries to

Speaker:

get into ninth or tenth grade way

Speaker:

too quickly, into automation way

Speaker:

too quick.

Speaker:

And I have to say, guys, please,

Speaker:

please always build individual

Speaker:

first, all those agents that I went

Speaker:

through. Make sure that they are

Speaker:

working, they're providing the

Speaker:

output individually before

Speaker:

you move to what Brandon and I are

Speaker:

about to talk about, and that's

Speaker:

automation. So these entire

Speaker:

automation can be done in Copilot

Speaker:

Studio.

Speaker:

So there are workflows that

Speaker:

you can build in there to say, to

Speaker:

attach it to your CRM and

Speaker:

extract data.

Speaker:

And a lot of people will push back

Speaker:

on this and is what if our data

Speaker:

isn't clean?

Speaker:

I don't care.

Speaker:

Start pulling what you can.

Speaker:

Your goal is to learn.

Speaker:

We're in a phase where we have to

Speaker:

learn fast.

Speaker:

And I'm like, no, identify three

Speaker:

fields. Like your name,

Speaker:

address, it's like figure out

Speaker:

what fields that you

Speaker:

can start to identify and

Speaker:

just start playing with it.

Speaker:

And what you'll do is you'll start

Speaker:

to realize, hey, this field

Speaker:

in CRM has to be populated.

Speaker:

We're gonna start to mandate that

Speaker:

because I need to pull these

Speaker:

opportunity records or I need to

Speaker:

pull this specific object.

Speaker:

You'll learn that.

Speaker:

So start small.

Speaker:

But you can do all of this in

Speaker:

Microsoft Studio, which are

Speaker:

two different, I'm going to call

Speaker:

them instances.

Speaker:

They might be two different

Speaker:

platforms that Microsoft does,

Speaker:

but they can create those workflows.

Speaker:

They have API connections, so

Speaker:

you can every single.

Speaker:

Thing in there.

Speaker:

So, half of our listeners are going

Speaker:

to be Microsoft houses and be able

Speaker:

to have access to that.

Speaker:

The other half are going be Gmail,

Speaker:

Slack, ChatGPT,

Speaker:

and then maybe

Speaker:

using Klay, we are

Speaker:

looking at a new product called Get

Speaker:

Cargo, which seems to be the

Speaker:

more modern Klay and Klay is already

Speaker:

so modern, it's like I feel sorry

Speaker:

for Klay that it's getting displaced

Speaker:

by Get Cargo.

Speaker:

And then Zapier was already

Speaker:

displaced by Clay.

Speaker:

But I'm guessing that they all have

Speaker:

similar functionality to Microsoft

Speaker:

Studio, so it's some Zapier

Speaker:

automation.

Speaker:

You're right.

Speaker:

I am actually working a lot.

Speaker:

So Zapier and then my favorite

Speaker:

is Make.

Speaker:

There's N8n and then there's a few

Speaker:

more that I've been playing around

Speaker:

with, but I'm not quite ready to

Speaker:

say their names because I'm still

Speaker:

playing around with it, but I like

Speaker:

it when I see.

Speaker:

But there's lot of different, if

Speaker:

you're not a Microsoft user,

Speaker:

then yeah, go to N8 and go

Speaker:

to Make.

Speaker:

Any of these Zapiers that

Speaker:

allows you to have

Speaker:

one task connected to the

Speaker:

next task works just fine.

Speaker:

And here's what's beautiful.

Speaker:

When I say some of these tools,

Speaker:

there are tools now

Speaker:

that, and I wish I had enough

Speaker:

experience with it that I can go,

Speaker:

I highly recommend this one, but not

Speaker:

yet, is there are schools now that

Speaker:

they've written all the automation

Speaker:

for you. You are literally dragging

Speaker:

and dropping things like,

Speaker:

you know, Gmail and

Speaker:

do this and do that.

Speaker:

And it is taking me like

Speaker:

10 minutes. And it does all

Speaker:

of the Zapier make

Speaker:

stuff for you, it's doing it

Speaker:

all for you.

Speaker:

So it's coming quickly even

Speaker:

if it's the one that you're not sure

Speaker:

if you want to share you want to

Speaker:

share go on preview us with it.

Speaker:

It's relay.app

Speaker:

and I think his name

Speaker:

is Justin,

Speaker:

the CEO.

Speaker:

He has, first of all, an incredible

Speaker:

story of how he has taken,

Speaker:

let's just say, a large, let' say,

Speaker:

55 plus marketing organization and

Speaker:

moved it down to one person with

Speaker:

like 80 something apps doing the

Speaker:

work. So a really good story.

Speaker:

And then he does a phenomenal

Speaker:

job on YouTube teaching

Speaker:

how to use his platform

Speaker:

to do all these integrations.

Speaker:

It's a game changer.

Speaker:

Yeah, it's a games changer.

Speaker:

But again, I've only been playing

Speaker:

with it for a week, so don't

Speaker:

judge me.

Speaker:

I think clay is going to be an

Speaker:

incredible tool for a lot of

Speaker:

organizations to get up and running.

Speaker:

I started using it about a year ago,

Speaker:

and I actually hired a

Speaker:

consultant to do the work because I

Speaker:

didn't want to try to figure out how

Speaker:

to work clay the best way, but I'm

Speaker:

not sure it's going to have legs for

Speaker:

the long term.

Speaker:

You know, it's my personal opinion.

Speaker:

Have a look at Get Cargo apparently.

Speaker:

Obviously.

Speaker:

Thank you for sharing.

Speaker:

Because if we go back to your first

Speaker:

two agents, you make it sound really

Speaker:

easy. So we have the seed list.

Speaker:

Now we're going to go find ones that

Speaker:

look like that.

Speaker:

How does it find it?

Speaker:

Do we need a database?

Speaker:

Do we a zoom info or an Apollo

Speaker:

or whatever?

Speaker:

Or does it just discover it on the

Speaker:

internet?

Speaker:

So this is like the

Speaker:

honest truth is I had

Speaker:

an SMB rep.

Speaker:

So when you think about the SMB, it

Speaker:

goes back to like Brandon, you're

Speaker:

like, he got a scrape for it's

Speaker:

like digging for a needle in a

Speaker:

haystack. So I had a

Speaker:

SMB seller who had, I'm

Speaker:

making this number, let's say 5,000

Speaker:

accounts and she would say 3%

Speaker:

of them are good.

Speaker:

Like, you know, I have a list of

Speaker:

accounts, but she's like, I want

Speaker:

more accounts like this one

Speaker:

and this one, and this and we

Speaker:

use this very and I know

Speaker:

I love that you said you make it

Speaker:

sound simple because I

Speaker:

swear it is like it's so simple

Speaker:

so I wrote this prompt

Speaker:

and I think you have to really write

Speaker:

really good prompts and I think you

Speaker:

get good at writing prompts by just

Speaker:

writing a prompt And

Speaker:

so literally recording

Speaker:

her to figure out exactly

Speaker:

what that needle looked like

Speaker:

in that haystack that she wanted and

Speaker:

then taking those accounts that

Speaker:

she had already closed that

Speaker:

were good and then saying go find

Speaker:

more of these.

Speaker:

We found and it took her a year of

Speaker:

doing it. She said that she was like

Speaker:

you know for the past year I haven't

Speaker:

found any you know more accounts.

Speaker:

We ran that.

Speaker:

I had a 30-minute call with her and

Speaker:

I will say in 20 minutes we were

Speaker:

off the because it scraped

Speaker:

and found her six new accounts

Speaker:

to go after.

Speaker:

True, true, true.

Speaker:

Every bit of that is just solid

Speaker:

truth, is she now had six

Speaker:

new accounts that were not in her

Speaker:

CRM that she had never heard

Speaker:

of that were ideal for

Speaker:

her to go after.

Speaker:

So I don't mean for it

Speaker:

to sound simple, but I

Speaker:

can do it.

Speaker:

It really is just

Speaker:

fun to go, ooh, what's a good

Speaker:

prompt that will extract this type

Speaker:

of data? And you don't need

Speaker:

Apollo or anything like that

Speaker:

for this one that we just said.

Speaker:

You literally just have to

Speaker:

really know your ICP and get

Speaker:

granular.

Speaker:

And then you have to be super clear

Speaker:

with instructions.

Speaker:

And I love giving examples.

Speaker:

So those example accounts

Speaker:

that she had had success in,

Speaker:

giving those and saying,

Speaker:

hey, the reason it was good is

Speaker:

because they were opening new store

Speaker:

locations, you know, they were

Speaker:

expanding into new

Speaker:

territory, like all those little

Speaker:

details that we take for granted.

Speaker:

Your agent.

Speaker:

It's like food for them.

Speaker:

Like, give it to me.

Speaker:

Don't be afraid to overshare

Speaker:

with your agent." I love that I just

Speaker:

said that. Don't be afraid.

Speaker:

Overshare with your agent.

Speaker:

But it's crawling over the internet.

Speaker:

And that's why I was asking, so it's

Speaker:

just looking at companies in the

Speaker:

internet that look like

Speaker:

the seed list.

Speaker:

Yeah, whatever your instructions

Speaker:

say, you're exactly right.

Speaker:

It is Curl in the Web.

Speaker:

And I will tell it things like, go

Speaker:

to Yelp reviews.

Speaker:

I'm being very specific

Speaker:

about where I want it

Speaker:

to crawl.

Speaker:

You know, go the events page.

Speaker:

Like, here, if I were

Speaker:

telling you, Bethany, how to get to

Speaker:

my house, and I just said, head

Speaker:

north, like,

Speaker:

help me out, Donna.

Speaker:

And so, I do, I tell

Speaker:

a lot of people that agents are like

Speaker:

really good recipes.

Speaker:

If you want a quarter cup

Speaker:

of sugar in there, you better

Speaker:

say a quarter of a sugar level

Speaker:

at all. You have got to get

Speaker:

detailed if you want to have really

Speaker:

good output.

Speaker:

I'm a big pyramid user

Speaker:

when I think about this.

Speaker:

So you guys mentioned your industry,

Speaker:

your law firms, real estate,

Speaker:

you've mentioned that.

Speaker:

So that's that industry top layer.

Speaker:

And then that account layer is what

Speaker:

we just did with our seed list.

Speaker:

So we went and we found lookalike

Speaker:

accounts.

Speaker:

And then what we wanna do is we want

Speaker:

an agent to extract

Speaker:

the right personas for us.

Speaker:

So each of the layers of the

Speaker:

pyramid kind of get things a little

Speaker:

bit more fine tune, fine tune

Speaker:

fine tune.

Speaker:

Now I'm gonna extract my

Speaker:

specific personas that

Speaker:

have that role and responsibility

Speaker:

that is gonna align

Speaker:

with the value that I'm looking at.

Speaker:

And then I want to take

Speaker:

that list, that lead list

Speaker:

as maybe a CSV and

Speaker:

upload it into the CRM

Speaker:

and just call it

Speaker:

my lead list and work from it before

Speaker:

I convert it.

Speaker:

And then we've created an agent

Speaker:

that knows

Speaker:

good prospecting emails

Speaker:

or what it is that the customers

Speaker:

want to hear about.

Speaker:

Yes, yes.

Speaker:

And so I love how you're

Speaker:

starting to do this is that, you

Speaker:

know, I think the reason I

Speaker:

found this pretty easy is

Speaker:

I've always developed SOPs.

Speaker:

And so for me to take those SOPs

Speaker:

and move it into AI was like,

Speaker:

oh, hallelujah, but

Speaker:

I recognize not all organizations

Speaker:

have taken the time to

Speaker:

write these out.

Speaker:

And, so I would say best practice

Speaker:

is grab an Excel sheet,

Speaker:

Write those in there where

Speaker:

you're saying industry,

Speaker:

account, buyers, competition,

Speaker:

and then looking at each one of

Speaker:

those headliners and saying,

Speaker:

okay, for personas, what

Speaker:

do we need to do with personas?

Speaker:

Well, maybe we need a scrape for the

Speaker:

buying committee.

Speaker:

We definitely need to

Speaker:

outreach, okay?

Speaker:

So we've got to outreach to

Speaker:

different personas, different ways.

Speaker:

So, we want to create

Speaker:

an outreach agent

Speaker:

that knows our personas

Speaker:

and we want specific

Speaker:

emails that are written to those

Speaker:

personas. If I'm going after

Speaker:

somebody who has fiscal

Speaker:

responsibility, then I want to make

Speaker:

sure that I'm addressing that

Speaker:

and I'm respecting their time and

Speaker:

really knowing that my message is

Speaker:

going to resonate with them.

Speaker:

So, I think there's different layers

Speaker:

that we have to look at

Speaker:

for when we see our pyramid is

Speaker:

to say, okay, what does really

Speaker:

good look like?

Speaker:

And just get started that way.

Speaker:

And then what's beautiful is that

Speaker:

you go back and you're like, oh,

Speaker:

okay.

Speaker:

We just scraped our

Speaker:

buying conversations, our really

Speaker:

good accounts, and we just found

Speaker:

that we totally missed a persona.

Speaker:

So now let's take that persona,

Speaker:

let's educate our agent on this new

Speaker:

persona and what actually

Speaker:

is resonating.

Speaker:

And I know what's resonating because

Speaker:

I went and scraped all my really

Speaker:

great accounts that you're working

Speaker:

with, right?

Speaker:

And when you talk about scraping to

Speaker:

get the personas, because LinkedIn

Speaker:

really blocks you,

Speaker:

or does LinkedIn not block, how were

Speaker:

you scraping? What are you scraping

Speaker:

to these personas?

Speaker:

I respect LinkedIn,

Speaker:

but from a leadless perspective,

Speaker:

you can absolutely download

Speaker:

your entire first-degree connections

Speaker:

as a CSV.

Speaker:

Anyone who is connected to me, I

Speaker:

can.

Speaker:

Scrape that or download

Speaker:

it, my contacts from LinkedIn

Speaker:

anytime I want.

Speaker:

So that'd be the first thing I'd

Speaker:

say. But then also, these people

Speaker:

are out on the

Speaker:

World Wide Web everywhere.

Speaker:

They're doing podcasts.

Speaker:

They are in news articles.

Speaker:

And so when I write my

Speaker:

instructions for my agent,

Speaker:

I am telling it, I want you

Speaker:

to go like, do not stop

Speaker:

at anything. I want to look

Speaker:

at podcasts. I want you to look at

Speaker:

news articles.

Speaker:

I want you to look at Yelp reviews.

Speaker:

I want you to look at the about

Speaker:

section of the website and

Speaker:

I want you to find people

Speaker:

who are talking about XYZ

Speaker:

or people with this title

Speaker:

or people who

Speaker:

are you know discussing this

Speaker:

challenge.

Speaker:

And that's a scraping.

Speaker:

The instructions go out and it's

Speaker:

like a little agent that says,

Speaker:

okay, let me go find what Donna told

Speaker:

me to go find, I'll scrape that,

Speaker:

wonder if she wants that, take that,

Speaker:

and then it's gonna bring that list

Speaker:

back and I can look at that list and

Speaker:

go, I actually don't want that.

Speaker:

Rerun and here's some new,

Speaker:

better instructions for you.

Speaker:

And then when you're writing the

Speaker:

content for the different personas,

Speaker:

does each one end up being its own

Speaker:

agent?

Speaker:

So you have your finance content

Speaker:

agent versus for me.

Speaker:

I kind of keep things pretty simple.

Speaker:

So when I built my

Speaker:

outreach agent, I have different

Speaker:

variables that I'm asking.

Speaker:

And I'm asked things like, what

Speaker:

persona are we writing to?

Speaker:

Do you know their pain point?

Speaker:

And often that means just, hey,

Speaker:

upload your discovery call or

Speaker:

a discovery call, upload

Speaker:

that. It can extract

Speaker:

information from uploaded calls

Speaker:

as well. But in this case, it would

Speaker:

be the variables of if

Speaker:

you'll tell me who the persona is

Speaker:

and what the pain point is, and I

Speaker:

think there's another thing that's

Speaker:

in there, then.

Speaker:

I'll go do the work and it

Speaker:

goes and does the work because of

Speaker:

what's in its knowledge base in

Speaker:

its knowledge base are

Speaker:

personas and the

Speaker:

persona cards you know that have all

Speaker:

the psychographics of

Speaker:

that persona and then the other

Speaker:

thing that's in the knowledge base

Speaker:

our email frameworks

Speaker:

to say hey will you use

Speaker:

um you know this framework

Speaker:

for for them things

Speaker:

like that so that i know that it's

Speaker:

going to provide value to

Speaker:

the CFO or COO or

Speaker:

something like that.

Speaker:

So we have the agent that's going

Speaker:

off. It's a ninth grade agent.

Speaker:

It really knows the right personas,

Speaker:

right accounts, right personas.

Speaker:

We have our CSV file.

Speaker:

We're using them as leads.

Speaker:

We have a great email

Speaker:

writer, and now it's time to

Speaker:

automate.

Speaker:

Yes, I love that you let me get

Speaker:

here.

Speaker:

Is this when we get the 24-hour

Speaker:

SDR?

Speaker:

Yes, and I won't give it

Speaker:

the title of an

Speaker:

SDR because I do genuinely

Speaker:

believe in a 70-30 split,

Speaker:

meaning I think that AI can do

Speaker:

70% of the work and

Speaker:

that the SDR can do

Speaker:

that 30%.

Speaker:

And I really think that

Speaker:

organizations need to move into that

Speaker:

mindset of,

Speaker:

in 2026, these department

Speaker:

leaders will be managing both agents

Speaker:

and people, like it's happening.

Speaker:

And so,

Speaker:

that 70-30 split there, so

Speaker:

what non-revenue producing work is

Speaker:

our agent taking care of, and then

Speaker:

what is our next role

Speaker:

taking care, whether or not that's

Speaker:

SDR or AE.

Speaker:

And being able to say, okay, you

Speaker:

are my automation layer.

Speaker:

And what we're going to do is we're

Speaker:

gonna take these three agents and

Speaker:

we're gonna create a workflow that

Speaker:

allows that to say go

Speaker:

out in research, then

Speaker:

go out and find the buying signals,

Speaker:

and then go out and write an

Speaker:

outreach, whatever that process

Speaker:

is, the workflow,

Speaker:

so whether or not you use Make or

Speaker:

Zapier or Copilot Studio,

Speaker:

it doesn't matter, but you're going

Speaker:

to string those agents

Speaker:

together.

Speaker:

And now I say

Speaker:

that I, as a human,

Speaker:

initiate that trigger and let's

Speaker:

just call it a

Speaker:

call agent trigger, whatever you

Speaker:

want to call it, a prospecting

Speaker:

agent. I initiate that,

Speaker:

the workflow goes to each one of

Speaker:

the agents, does what it's supposed

Speaker:

to do, and I, as the human,

Speaker:

then see what that output is

Speaker:

and say, approved,

Speaker:

go do.

Speaker:

So, I think having that human

Speaker:

element as a part of the

Speaker:

workflow is critical

Speaker:

to getting started.

Speaker:

Human in the loop, particularly to

Speaker:

make sure it looks good.

Speaker:

Did I not say that?

Speaker:

Did I say it?

Speaker:

He said, yeah, I'm just using

Speaker:

the trendy term

Speaker:

for it.

Speaker:

I'm a big believer that all

Speaker:

of the AI agent needs to tie to

Speaker:

revenue.

Speaker:

And so I think that somebody in

Speaker:

rev ops should be the

Speaker:

AI overseer to

Speaker:

say, hey, what agents are we

Speaker:

creating?

Speaker:

And I think an agent card needs to

Speaker:

be created so you know what the

Speaker:

input and outputs are going to be.

Speaker:

And then I think there

Speaker:

has to be that governance around

Speaker:

those agents that get created.

Speaker:

I just had another thought as

Speaker:

an idea.

Speaker:

So SDRs often were responsible

Speaker:

for all the outbound, a lot

Speaker:

of like being really smart on who

Speaker:

you're gonna contact, doing a lot

Speaker:

this manually in the past, and

Speaker:

then also a lot cold calls.

Speaker:

I don't know in America, but in the

Speaker:

UK, like cold calling still works.

Speaker:

So there is a lot like dialing

Speaker:

and calling.

Speaker:

Do you think that in the

Speaker:

future, the SDR role might get

Speaker:

split between?

Speaker:

The people who are doing all the

Speaker:

building, automating and

Speaker:

creating this air cover and

Speaker:

making sure it works and

Speaker:

people who just do coke

Speaker:

holes all day.

Speaker:

Yeah, I think the role of

Speaker:

an SDR is dramatically

Speaker:

going to shift.

Speaker:

And I'm seeing a lot

Speaker:

of these SDRs asking

Speaker:

questions around how do

Speaker:

I improve my

Speaker:

skill sets to be able to

Speaker:

do what this AI is coming.

Speaker:

So I think there's going to be a

Speaker:

major shift. And I do think that

Speaker:

we'll still have that cold calling

Speaker:

that's there, but I don't even

Speaker:

know if I would be guessing

Speaker:

it. I am horrible at predicting the

Speaker:

future, but

Speaker:

but I do see it evolving.

Speaker:

Yeah, because I guess I just see

Speaker:

from the SDRs I see in the

Speaker:

world, the ones who are kind of

Speaker:

loving the cold call, love the

Speaker:

interaction, and the ones who are

Speaker:

actually way more technical and

Speaker:

get really into the process.

Speaker:

Because it's hard to find somebody

Speaker:

who wants to do a lot of process

Speaker:

and hit the phones.

Speaker:

They don't tend to be the same.

Speaker:

Yes, very true.

Speaker:

So yeah, it'll be interesting.

Speaker:

You know, there's

Speaker:

AI agents that are now calling and

Speaker:

they are really good.

Speaker:

And like, that was fun.

Speaker:

I don't know.

Speaker:

I don't. And that's why I'm here

Speaker:

just to say everybody, like,

Speaker:

I do know what's currently

Speaker:

available. I do what we can

Speaker:

currently take advantage of, like

Speaker:

get started.

Speaker:

Like that's my mantra, get started.

Speaker:

So I think you've preempted it,

Speaker:

but you might just have to answer it

Speaker:

again.

Speaker:

Which is, if our

Speaker:

listeners can only take one

Speaker:

thing away from the episode today,

Speaker:

what is it?

Speaker:

Two word, get started.

Speaker:

Don't be afraid of it.

Speaker:

Go in there and implement the three

Speaker:

things that I said, your output,

Speaker:

your instructions, and your

Speaker:

knowledge base.

Speaker:

And then see what it gives you.

Speaker:

And if it's not what you wanted, go

Speaker:

tweak the instructions and try

Speaker:

again.

Speaker:

Get started, it's moving fast.

Speaker:

On that note, I will get started on

Speaker:

my target list that I'm going after

Speaker:

for these law firms based on your

Speaker:

guidance here. So thank you very

Speaker:

much, Donna, for joining us on the

Speaker:

operations room. If you like what

Speaker:

you hear, please subscribe or leave

Speaker:

us a comment and we will see you

Speaker:

next week.

Links

Chapters

Video

More from YouTube