I'm excited to be joined with Lucy

Alexander. She's the director of agentic

prospecting at HubSpot. Welcome, Lucy.

>> I'm so happy to be here. Thanks for

having me.

>> So, you know, I want to jump in. You've

been at HubSpot for a while,

>> but you just took this new role,

director of aic prospecting. This role

didn't exist

at the beginning of 2026 even, right?

>> No. Yeah, definitely new role and uh

definitely a very AI era job title. So

help me understand why uh I mean how

HubSpot's thinking about this role.

>> Yeah. So I view my mission as using

agents to make our sales reps more

productive specifically in terms of

creating pipeline for them to close. So

whether that looks like creating tools

for them to use to accelerate their

pipeline generation or creating agents

that also create pipeline for them

without their intervention. Our

northstar is pipeline per rep. So we are

tied directly to the business outcome,

but my role is more of a technical role

on figuring out how do we build the

tools to get us to that target.

>> So the pipeline per rep probably was a

general goal already. It's just the new

titles helping you think differently

about how you get there. Is that right?

>> Yeah. I mean, we're looking at

additional metrics like do we get there

efficiently? Do we reduce the number of

touches from a rep that it takes to

generate pipeline and can we point them

toward higher value deals? Things like

that as well.

>> This is an awesome segue. So like now

with all these things in mind, like walk

me through how you're optimizing for

pipeline or like like give me an example

of like a lever that you're pulling that

you've seen like a tremendous impact

from or planning on seeing it.

>> Yeah, the first place that we started

was in a place where we didn't need rep

intervention at all. So my background is

originally in kind of the customer space

at HubSpot, but two years ago I took

over a prospect email team and we saw a

huge opportunity to use generative AI to

create completely personalized emails to

work our inbound demand. And so we

already had a program that would send

out an email, for example, on behalf of

a sales rep if they viewed a pricing

page, for instance, and offered the

chance to book a meeting with that rep.

We had strong conversion rates and were

generating a lot of pipeline and

meetings from those emails already. But

by bringing AI in and iterating a lot,

often 5 to 10 iterations before we

reached impact, we were able to get it

to anywhere from 50 to 200% improvements

in the meeting conversion rate of those

emails. So at this point, the team is

generating hundreds of thousands of

dollars in pipeline from those emails

that all are onetoone personalized for

the prospect. So better experience than

getting a kind of a random template and

better business outcomes as well. So we

started there, no rep intervention

required. We're doing it on their

behalf. And now we're thinking more

about, okay, well that doesn't cover

every single use case that a rep has for

writing AI emails. We want to now give

them the tools to be able to write

better emails um and place better calls

and conduct more relevant outreach. So,

we're thinking about things like

outreach agents that help them draft

wellressearched outreach that they can

then add their personal touch to. We're

thinking about research agents that go

and find relevant pieces of company

news. We're thinking about cold call

agents that will prepare a list of folks

for them to log in in the morning and

call rather than having to put that list

together themselves and agents

eventually that they can work with to

upskill on our product, upskill on cold

calling skills, um create really

personalized pieces of outreach much

faster. So, for example, like if they

want to go send a video to a prospect

today that's totally personalized,

they've got to take the time to go make

a slide deck or a couple slides and then

go figure out what am I even going to

say to this prospect? How am I going to

um you know, show them that I've done my

research? And we think we can expedite

that kind of process. So, trying to

think about first what are the basic

jobs to be done that we need to

accelerate with AI and then how can we

layer on creative approaches on top of

that. And it sounds like uh so first on

the inbound side, you went from just

sending them template email if someone

was inbound on side to like optimizing

what that email said. So like you knew

that that there was a ton of signal. Now

it's just saying the right thing and and

putting AI there. Now this is you're in

a more outbound role, right? Is that

accurate with with this?

>> It's it's sometimes outbound. Sometimes

it's also it looks like multi-threading

with additional contacts at the account

that that rep is working. So we'll email

the person who was checking out our

website automatically, but maybe they

want to pull in for HubSpot, you know,

the director of marketing or the

director of sales as well who hasn't

been on our website and they need to put

together a um really strong compelling

message to earn that person's time as

well. So if I understand this correctly,

your old responsibilities you still have

um from like an inbound optimization uh

but now you're layering in other things

so that you could see those signals

as well as more see how that impacts go

for a higher even higher conversion

rate. The thought process is

can we convert even more of those

inbounds by multi-threading all this

other things and that's pretty logical

and then layer in when you're doing

outbound at HubSpot is there a cold

element to it or is it mostly warm and

are you getting other signals that might

not be like what you were previously

working on like expanding those to to

make it more impactful.

>> Totally. Yeah. I think of it as broadly

first and third party signals. So you

have folks who are showing intent to

purchase by coming to your website or

going inbound. You've also got signals

externally either from data vendors or

from company news that's just publicly

available. They got a funding round for

instance. And we try to bring those

things all together so that our reps can

understand the holistic picture of that

account. prioritize their entire book of

business for them so that they don't

have to constantly be working out of

like which sort of campaign or play do I

go after next and then when they are

working that that specific account um

they're able to like generate talking

points or generate emails or generate

creative outreach strategies that pull

in that holistic picture rather than

maybe before they were focused primarily

on a competitor rip out or something

like that. So, it sounds like you're

doing all top of the funnel now and just

like optimizing it with AI, which is

awesome. And then how are you thinking

through build verse buy? Cuz it sounded

like you're building agents, but we know

that there's out ofthe-box solutions,

some of which that are getting pretty

good reviews.

>> Yeah. And we have tried a mix of tools.

We of course test our own tools and our

target audience as HubSpot. Our company

is smaller than HubSpot. So often times

we do need to customize our own products

a bit more. On top of that, um, there

are some amazing tools on the market.

Like I mentioned before, my ultimate

goal is pipeline per rep. And so we're

totally willing to test out different

tools to see what will get us there the

fastest. And sometimes we buy first and

then, you know, decide to build later.

But we have a really multifaceted sales

org as well. So we're not necessarily

building oneizefits-all solutions. We

have BDRs, we have account executives,

we have three different segments, we

have specialist, we [clears throat] sell

certain tools. So not everything is

going to work for every single every

single rep and we sometimes will build

in one area, buy in another, too.

>> So do you also have revenue metrics as

well as pipeline metrics under your

purview?

>> Technically, not under my purview, but

that's absolutely something we look at.

We want to make sure that our pipeline

that we're generating is actually

converting into dollars at the end of

the day.

>> Got it. So, so you're not necessarily

optimizing for reps on calls as much as

you are to get them calls.

>> Correct.

>> And then

>> another team that handles all of the

cool things around call prep and

followup and making sure that we're

nudging reps to do the right activities

on deals to progress them too. I'm most

curious about so understood pipeline per

rep core metric. How has that number

changed in the last year in terms of

just like the whole number?

>> Are you talking about like how much or

the like metric that we're using?

>> Yeah. Your your pipeline per rep goal in

2025 versus 2026.

>> Yeah. I mean as a company we're always

looking to grow. So we are trying that

works out to growth targets in terms of

what our reps bring in and when we work

back up the funnel that means we need to

generate more pipeline or drastically

improve our close rates to generate more

of that revenue. So, um, we have all of

our different sort of demand sources

feed into that revenue target at the end

of the day and we hope that these tools

we hope and we measure that these tools

are actually giving us incremental lift

over reps who are not using the tools or

reps versus their past selves. Does that

answer your question?

>> Yeah. So, I think where I'm trying to

get is like relative to other years

because you have this tool AI. So, like

increasing goals fortunately or

unfortunately seems to be the norm. uh

we we have uh bigger expectations

especially as a public company to

you know do what's right by your

stakeholders in addition to the company

and and and grow that way but AI an

inflection point right like where where

we could do more with less and as a

whole like theoretically have you seen a

greater percentage increase in the

year-over-year goal attainment from like

2025 to 2026 or 24 25 versus like 5

years ago or is it I'm just trying to

see like if you have a forcing function

with like higher goals that you haven't

had in the past just from that you could

cure with better ena better old school

enablement so to speak theoretically

like that pushes you to optimize even

more.

>> Yeah, it's interesting you bring that

up. I think we're having a conversation

about it about what incentivizes people

to like really think 10x rather than

just in the incremental. Right. We're

such a large company. I think we h

inevitably are going to push ourselves,

but there are so many different levers

that we have. We have so many different

demand sources. So, for example, when we

proved that AI emails could convert

folks into booked meeting or like get a

lot more booked meetings than before, we

had aggressive growth targets the

following year for how many meetings

generated from email we needed. that

growth percentage

>> was

depending on the segment between like 30

and 50% year-over-year growth because

that was a proven that was an area where

we were starting to see signal that yes

wow this could drive a lot of

incremental growth now there are other

areas where for example working with our

BDR team we rolled out a tool with them

that was also an AI kind of research and

drafting tool we don't expect that they

are going to become 30 to 50% more

productive in the span of a year. They

are people with, you know, a training

training enablement like they're

learning the skill set of how to reach

out to folks like their early career.

So, we're not going to set a target of

30 to 50%. And there's kind of this

balance between what is realistic for

somebody to attain. How do we set the

right quotas to set people up to succeed

as people and then also how do we um

like pilot things and incentivize

growth. So in that particular case for

example when we rolled out that tool we

ran a pilot where we incentivized people

to go far beyond their typical quota and

then once we proved that it worked we

scaled up that tool we rolled out to

more folks and then that sets us up to

be able to raise quotas and incremental

ways later. I love that and I think um

it's it's a very pragmatic approach

versus like the Shopify which wasn't

laid with KPIs the now famous email like

just figure it out do this you can't

hire this way like that's uh

another public company's point of view

and and actually they're doing well in

their own right but what KPIs outside of

pipeline maybe are more like leading

indicators right like if you're looking

at pipeline as lagging and and some some

that you're using Like some things I've

heard selling time per rep, like actual

selling time. I'm sure there's a bunch

others, but that that one in particular

has been top of mind. How are you

thinking about those other leading KPIs

that might be easier to hit short term?

>> Yeah. Um, we generally want to use

automation and agents to take admin work

and repeatable things off of rep's

plates so that they can spend more time

doing the uniquely human pieces of this.

We are not looking to rooc call a bunch

of people. We want our wonderful human

beings to go pick up the phone and have

a real genuine conversation with

someone. And so one of the things that

we look at is have they shifted time

from emailing to calling by looking at

email volume sent by reps and call

volume by rep so that we can start to

see okay are we actually seeing that

like automation is taking a bigger and

bigger slice of these emails and that

reps are then able to reallocate that

time to more calling. We also have a

metric that we've worked on internally

that basically quantifies the value of

like a minute of rep time. So if they

were able to reallocate that five

minutes that they would spend

researching and writing an email to an

activity involved in progressing a deal

or closing a deal, then we expect we

could bring in x more revenue, but

that's a little bit more on the lagging

side. So really, it's about like the

inputs. Are we seeing that the number of

companies that they reach out to are up?

Are we seeing that they're specifically

reaching out a lot more and a lot more

quickly to the best companies in their

book? And then are they calling more and

sending manual email less?

>> Uh my I I I I laughed in my head

thinking through the Facebook anecdote

of like every calendar invite showing

what that cost the company.

>> Yes.

>> I don't know if you've seen that of like

>> I've seen that. Yeah. Exactly. like we

do really think about the most valuable

asset that we have

>> is our humans attention and so we want

to be putting that on the right places

where there it's the highest highest

leverage for the company and like also

what our prospects want they want to

have a genuine conversation with

someone.

>> I'm curious now like with with the newer

things and and you're newer in role

understood but like what are some of

like the early successes that you've

seen or the results I should say?

>> Yeah. Um yeah very new two weeks into

the role but was

So, anything now would be completely

luck.

>> Yeah.

>> Yeah. I'll talk I mean I was I had my

feet in this world beforehand. Um and

couple of the successes we've seen,

number one, call volume did go up

significantly when we started to

automate a lot more of our emails um on

behalf of reps and the conversion rates

of those comparable rep sequences versus

the AI driven email sequences were the

same if not better for AI. So that tells

us, okay, we're starting to get

repetention toward the more human

channel of calling. Um, I can't remember

the exact stat off the top of my head,

but I think it was around a 10% increase

in the number of calls that a rep was

making per day. Um, so that's a good

leading indicator. The other thing that

we've seen is the value of

prioritization. So, we're using agents

to go comb through and review accounts

before we tell a rep like this is the

most important one for you to go after.

So, we basically talked with a bunch of

reps and we said, "How do you evaluate

an account?" Um, when you go get an

account in your book, what are what

specifically are you looking for on

their website? What are you looking for

on LinkedIn? All of these elements that

we hadn't really been able to pull at

scale. We replicated that um with our

own like in-house scrapers and third

party tools as well. And then now we're

able to surface a really good list. So,

we were actually able to get it to the

like the top 30% of accounts in a rep's

book generate 95% of their revenue,

which was a really good signal for us

that we had actually found. If you just

focus on these top 30% of accounts

rather than spraying praying with

everyone, then you're going to drive

much better revenue outcomes.

>> That's incredible. Thank Thank you so

much for uh peeling back the kimono a

bit and showing us what uh what's up at

a powerhouse.

>> Yeah, happy to.

>> And thank you for your time, Lucy. This

has been incredible. Of

>> course.