Seeing Through The Marketing Data Mirage
#1

Seeing Through The Marketing Data Mirage

The metrics look great, but the pipeline doesn’t. That tension sparked a frank conversation with Bill Hobbib, CMO of Demand Science, about the marketing data mirage—why so many programs appear to win on dashboards yet fail where it counts: qualified opportunities and predictable revenue. We dig into what really signals buying intent, how to stop chasing ghosts, and why AI-only content is quietly eroding brand trust. We start by breaking down the core problem: clicks and topic interest are no...

Seth Marrs: Hello everyone and
welcome back to the Innovative

Revenue Leader Podcast.

Today we have the opportunity to
take a deep dive into marketing

to what what is called the
marketing data mirage.

So that's that occurs when
metrics suggest programs are

performing well, but the
underlying signals are

unreliable, they're inflated,
they're not tied to real buying

behavior.

So so basically what you think
is happening is kind of not

happening, which makes it very
hard to do the things that

you're trying to make decisions
on things.

So it makes weak programs look
strong, mass gaps in execution,

enforces sales team to chase
opportunities that were never

real to begin with.

So I think many people are
dealing with this, so I'm

looking forward to digging in.

Demand Science just released a
report around it, and they they

talked to 750 people, senior
marketing leaders, focusing on

this specifically.

So to talk through this, I had
the the privilege of having Bill

Habib on.

He's the CMO of Demand Science
with more than 25 years of

executive experience.

Bill's built and led high-impact
marketing teams at both

established technology leaders
and fast-growing innovators.

He's played a pivotal role in
category creation.

I worked with him at Barrisent
when he was working through that

category.

Go to market strategy for
multiple companies, Oracle,

Bullhorn, Data Robot, Barrisend,
and he's helped guide the

creation of multiple unicorn.

He's also been a marketing
leader at three companies that

were acquired at 10X valuation.

So very distinguished guests to
have on.

And you guys, you were at the
center of this.

You and I actually talked, hey
Bella, it's it's great to have

you.

SPEAKER_00: Great to see you
again.

Seth Marrs: Yeah, I mean, you
and I had this conversation

before we we did this.

Just we were talking about it,
and and I've always really liked

like you're very thoughtful in
the way that you think about

things.

And you could see it in the
report and how you put it

together, the way you sequence
this, the bytes around the top

five, like all this stuff.

So I'm super excited to jump in.

Let's let's just start with a
lot of times when people do

these surveys, they do a general
survey where they think the ant

where they're trying to get the
answers they want versus like a

very target, I have a
hypothesis.

Like which one of this was for
you?

Did you know the data mirage at
the beginning and you were

trying to dig in, or were you
looking in general at the

problems that are happening
around the business, things that

are working in marketing, and
you came to this conclusion?

SPEAKER_00: Yeah, it's a really
great question.

And so the answer is a mix of
both.

I knew from 25 years of
marketing experience that people

were dealing with problems with
data and signals, with content,

with activation, with um now
with AI.

How do I incorporate that in my
processes?

How do I continuously innovate?

Because a lot of the value comes
in the iteration loops and the

speed at which you could do
that.

So I've seen it and lived it
myself.

I've built big stacks, Martex
stacks, and found I didn't get

the value of them.

So a lot of these were problems
I've personally seen and

experienced as a marketing
leader.

And I'm in lots of CMO forums
hearing from my peers who all

talk about these.

So we had a theory that these
problems were there.

And so we went out to test some
of those, plus other areas to

see, hey, is there anything else
that we've missed?

All around the lines of like
everybody's got to drive higher

performance from marketing.

Where are there gaps that are
potentially sitting in sort of

um plain sight that we could
help marketing leaders close and

help them drive more revenues?

So we had a thesis, but we also
went kind of broad with our

questions to see that the thesis
get proven and were there other

areas or insights that we
missed.

Seth Marrs: Oh, cool.

So you kind of knew it was gonna
go there.

So you asked the right questions
on it, but then you expanded out

just in case you had other
things and it kind of pulled the

air awesome.

SPEAKER_00: Exactly.

Um, for example, we asked a lot
of questions about AI and

marketing.

Um, we didn't report out all of
the findings because a lot of

people have seen some of those
things already, but the things

that we did find that aren't out
there, we uh we reported back

because you thought those are
pretty innovative.

Yeah, that's like AI-driven
content isn't working very well.

And 71% of people said AI-driven
content is hurting their brand.

I'm like, this is not
surprising, but it was

interesting.

We didn't know what was going to
come back on that.

So let's dig into that a little
bit, right?

Seth Marrs: Like 72% say it's
hurting their brand.

Like, does that mean they
shouldn't use it?

Are they using it wrong?

Like, what on what that that was
kind of a shocking stat in there

because everybody now is just
pushing out AI-driven content.

I don't know.

I would say most content these
days is at the very least going

through AI for a second look.

Like, what what do you do with
that?

Like, how how it is do you stop?

Like, what is it?

SPEAKER_00: Yeah, it's a great
question.

And and especially if you
consider people are using AI to

generate the content so that
ChatGPT and other LLMs consume

the same content.

And so you've got chat and LLMs
creating the same content

they're gonna consume.

And at some point, you've got to
get out of this sort of circular

loop.

Um, so what do you do with this?

I believe, first of all, the
human element to this thing is

super important in two respects.

Um, one, um, to what extent are
you using AI as sort of the idea

generator and the first draft
generator as opposed to the

final arbiter?

Um, and so are you generating
stuff that's not having enough

human involvement?

But secondly, how do you make
sure this is both um consumable

by the machines, but consumable
and relevant to the human?

Um so even if if if chat or
another LLM um consumes it or

you show up on AI overviews, is
it the right content for your

personas that's gonna resonate
with their pains and problems?

And because that's what's gonna
actually cause that content to

convert into initial meetings
and subsequent stages in the

pipeline.

So I think there's a piece of
this that's simply make sure

your humans are involved in this
process pretty actively.

And are you really are the
people driving this, do they

really know what your target
personas want and need and what

their visceral pains and
problems are, what's gonna

resonate with them on a tactical
basis for their jobs and an

emotional basis?

And machines just aren't so good
at that.

Seth Marrs: And so no, like what
you were saying around it, it

kind of makes me think.

Do you remember way back?

I mean, uh both of us would
remember this, like when the

when the internet started and
people people were doing like

really weird stuff.

Like they would just put a
string of keywords into their

website because they wanted it
to come up in the different

search site.

It kind of feels similar in that
you're trying to placate the

engine rather than the person.

Yes.

Inevitably, one backfires
because the person doesn't care

about that, and two, eventually,
all these tools are gonna just

wipe that out because it's it's
actually the raw, it's cheating.

SPEAKER_00: It's cheating.

It's it's like the old, like you
mentioned, it was the keyword

stuffing in the metadata tags or
at the bottom of website pages,

but in the meta the meta tags
that you were you were stuffing

in there.

And yes, at some point this is
gonna run its course.

I I I firmly believe, and we
adopt AI and use it all all the

time in our content and other
things.

I'm a massive believer in it.

It makes me more efficient, it
makes my team more efficient,

makes every marketing team more
efficient.

But you've got to build into
this process, making sure

someone knows who's my target
persona, what jobs do they have

that that have to get done, what
matters to them for success in

their job.

And how are you making sure your
content is provocative and

insightful for them to bring new
insights that help them do their

jobs better?

And like, I don't know that
machines can do that very well

today.

You still have to have humans
involved to make sure that all

fits.

Seth Marrs: Yeah, I mean,
because a lot of the biggest

campaigns that work that really
drive interest are not the same.

Like they the way they stand out
is by being different than any

different before.

So if you don't have somebody
trying to look for the

differences, or even if you're
using eye to prompt properly

with your expertise to find the
differences, yeah, it's the same

old crap and everything just
looks gray rather than, and then

the one or two people who are
really making it stand out stand

out almost even more.

SPEAKER_00: That's exactly
right.

And and what we learned from the
from the finding, which we

didn't go into as as a as a core
theory, but we learned from it

was when, you know, on the on
the revenue topic, when people

and marketing leaders are trying
to drive more pipeline, almost

any sales or marketing leader
can emphasize with with I think

this statement, oh crud, we're
halfway through the quarter and

we're short on our pipeline
target.

Like, okay, like everybody can
what do people do?

They generate more content, they
buy more data, they advertise

more.

Um, and and but what's happening
is they're buying, they're

dialing up more of the same
instead of saying, gosh, do I

have a connective thread between
my signals, my content, and my

activation?

It in ways are gonna make this
really efficient, or do I just

dial up more of what I was doing
and hope that's gonna yield more

pipeline?

Seth Marrs: Yeah, so this takes
us into another part of it.

And you and I had this
discussion around this.

The like you talked about 80
seven, 80, 80, and this kind of

goes into what you were talking
about with 87% of marketers

chase fandom signals and only
26% become opportunities.

And what I told you before on
that, I'm like, that actually

sounds not too bad because when
you're sitting at the top end,

it is very hard to understand
what a real signal is.

And even a signal that's real at
the top end, customers are

fickle, right?

Like, I may be interested at
this point, but my interest dims

as I go.

So you're going to have this
natural flow through.

But I think it, I think the
point that you're making with it

goes back to what you've just
been talking about.

Like, what why like is is that
like when you think about that

challenge, like is there a way
to really keep that all the way

through or more than like I
thought 26%.

I'm like, oh, that's not so bad.

But you looked at it, you're
like, no, like there's there,

there's a there's a real
opportunity here.

SPEAKER_00: Yes, yes.

And here's the um here's the
extra layer of of thought here.

And I'm a marketer who was
involved for the last 15 years

in pushing intent companies and
others to get those signals.

Who's clicking on my stuff,
who's clicking on my

competitor's stuff?

Like, what are the indicators of
somebody potentially being in

market so I can point my
machinery toward going against

them?

Here are the two problems and
the challenges, and why that

that that that delta stum kind
of jumped out at us.

First of all, what's a signal?

Um, and all of the companies in
the intent space today are also

kind of making, you know,
raising this sort of question.

Is a signal um somebody from um
NVIDIA clicked on one piece of

content or a few people clicked
on pieces of content about

supply chain?

Does that mean NVIDIA is in the
market for supply chain

software?

No.

If what you do without having
the context and provenance for

that is simply tell your selling
team, hey, NVIDIA, you're seeing

signals for NVIDIA around supply
chain, and then your sales force

goes against those.

What's gonna happen when they
when they um these flame out and

they don't they don't turn
anything?

At some point, people are gonna
say, This is the marketing

organization that cried wolf
telling me to go chase these

signals that aren't really
signals, they're just clicks,

they're anonymous clicks from a
company on certain topics, and

that's what you know is part of
what's leading to that 89%.

People have gotten really
committed to chasing these

signals that we thought were
intent, which are nothing more

than clicks on topics.

So that's problem one, and it's
a phenomenal um a misuse and

waste of resources if you're
because then you direct your ABM

campaigns against the companies
that have the signals.

Seth Marrs: Yeah.

SPEAKER_00: Um are they signals?

They have to be validated, they
have to have provenance, you

need to know the context of
those and how important they

are.

So that's problem A, and that's
solvable with provenance and

greater context and sort of
opening up the black box of what

these are so you can better
evaluate them.

But I would volunteer, there are
multiple types of signals.

There are the clicks, there are
click signals on on content, and

those are sort of well
established, but there are other

signals.

Um MA activity, um, investment,
whether it's VC investment or PE

investment, um, new hires,
especially at an executive

level.

Um, you joined Sandler not all
that all that long ago last last

summer.

I joined Demand Science last
August.

We were about a month or two.

A new executive hire joins for
senior hire, even um a revenue

potential CRO, CMO, or people at
the VP level, they often are

brought in because there are new
ideas that they're going to

bring in.

And and so um executive hires
can be a signal.

Market growth, um, um MA
complexity.

What's the complexity in their
sales stack or their Martech

stack or their supply chain
stack?

So people can use tools to know
the complexity, risk,

compliance, there are um um a
variety of signals, firmographic

uh is is yeah, people do that
for 30 years.

Yeah, even technographics,
people have been doing for a

while.

But these other signals could be
greater indicators of a

likelihood or propensity for
somebody to buy or be in market.

And are you looking at those
signals as well?

And how do you incorporate all
those into your engine for ABM,

for account scoring, for
building your book of business

and sales?

Those are signals too.

Seth Marrs: So I if I
understand, I mean, that makes

total sense to me because what
you're saying, and and I agree

completely, is don't take the
onesie twosies aggregate as a

marketing team, aggregate them
all together and turn them into

a unified signal and pass that
signal because signal A, B, C,

and D together about that
account says yes.

So pass that one, not the five
signals that you got that

individually mean nothing.

They go to sales and they have
five actions versus I've

actually thought we we've pulled
together an algorithm that looks

at these things, and when we see
these five, six, seven things

happening, it's important.

Now you're past it.

So instead of 89%, it may it I
totally, yeah, that makes total

sense.

SPEAKER_00: Bingo.

And here's how people, here's
how you can monetize this

actually.

So this isn't just sort of
theoretical.

So I'll give you two, I'll give
you two examples.

Um uh I talked to um uh a head
of demand gen just two days ago

who is using these kinds of
signals, and he's driving um an

ROI of um of CAC to LTV of one
to seven one to seven.

So they spend a dollar in
customer acquisition costs, they

are getting seven dollars in LTV
and with a very small BDR team

and and an agile marketing team,
in large part by using these

kinds of signals.

Seth Marrs: Yeah, the aggregate
being smart at the top end, it

allows because I mean the most
expensive resource in that whole

thing is your salesperson.

So if you pummel them, it
doesn't even include the fact

that they what you talked about
earlier, they lose interest.

That's exactly right.

Yeah, that makes total sense.

Interesting.

SPEAKER_00: Yeah.

You know, different example.

Um, a very big cloud company, I
I have to leave them nameless,

was spending like 600 some odd
dollars per account to qualify

from from marketing.

When they better leverage both
signals and um a much more

tightly integrated kind of
multi-channel activation

approach, they lowered that cost
for marketing to acquire an

account from like 600 a change
to 100 a change.

It's like a 400% improvement by
better signals and better

picking those signals against
their ICP and then activating

those accounts in a in a very
tight closed loop sort of way.

This is what we what we saw in
the report is sort of

repeatedly, if people weren't
able to connect these threads,

then their waste was higher.

Um and they're they realized it
and they said, My revenue

opportunity potential if I did
this well is enormous.

They actually recognize there's
a there's a weak spot here.

It's costing me for not doing
this well.

Like 25% of their marketing
budgets are spent on stuff that

looks okay in a dashboard.

They look at the dashboard, they
say, Oh, my impressions are

good, my stage one pipeline
first meetings look good.

If you if you measure M2Ls, my
M2Ls are equivalent, look good.

Um, but then it's not
converting.

And then so the CMO has to go.

Seth Marrs: Well, this takes us
into that that what you have

talked about in the report
around the around the confidence

paradox.

Yes.

Which so like you you you talked
about it in 89% of marketers

trust their data, but only 43%
uh say the metrics look good but

don't convert.

Yeah, so 43%.

So you you've got this, like you
what you you're basically

talking about that gap.

Like I when I looked at that,
I'm like, okay, well, if I've

got good data, shouldn't that
flow in?

So it felt to me like we're
measuring the wrong things.

And I think what what you're
saying from the previous

conversation is you shouldn't be
measuring the individual signals

that come in.

You need to build aggregated
signal signals that that say I

mean to a certain extent, isn't
this like what what you're

hearing with people like Terry
Fleurerty, Amy Hawthorne, Kerry

Cunningham talking about buying
groups and how do I get a how do

I get signal aggregation into a
place where I can understand not

just the account stuff that's
going on, but I attribute it

into individual opportunities
and deals that you could push

through to a salesperson.

SPEAKER_00: Exactly.

So Kerry's talking about, you
know, their their most recent

market research talked about how
um early the buying journey

actually starts.

And if you're just chasing the
5% of the red market, what's

happened is you left the other
95% of your ideal customer

profile to poke around and learn
this on their own.

And then by the time they get to
the project stage, the RFP

stage, whatever, they've already
shortlisted the one, two, or

three vendors that they most
want to go after.

So you got the buying group
dynamic.

You also got the dynamic of are
people forming opinions really

early before the signal, you
know, um, you know, the strong

signal shows up.

So so what do I think is going
on with that um that particular

um finding for the report?

People said they have a
reasonably a rather high

confidence level in the state of
their data and analytics.

I feel that I have the right
people in place, the right

systems and processing in place.

In general, I feel pretty good
about the state of my data and

analytics.

But when pressed further to say,
how often do you have stuff

showing up on your dashboards
that looks good that doesn't

turn into pipeline?

It was like 43% or something
like that said these early

metrics might look good.

And that's not surprising.

These are sort of um
top-of-funnel sort of metrics.

It could be website visits,
which are fine.

But are they high intent leads?

Um are they MQLs if people are
still measuring those, or um

early stage pipeline like stage
one?

You and I talked about this when
we first met a year and a half

ago.

How often are companies, um,
organizations still measuring

sort of early stage pipeline
that means uh you know a meeting

has been accepted or fulfilled?

Um, and but it's not really
qualified yet.

And you can't build a forecast
off that.

Um, you could build a pipeline
model that says here are early

stage indicators, but you
shouldn't build a forecast off

it or pat yourself off the back
with vanity metrics about

impressions or website traffic
or things of that nature.

What really matters for revenue
leaders, if I I'm a marketing

guy, but what is from every
revenue leader is how many

qualified opportunities are we
generating that I can predict

will close at a certain
predictable win rate?

Every quarter that has to go up
by 2%, 3%, 5%, however, whatever

your growth rate is, right?

Whatever your growth rate is.

And so I think the learning for
us in this was people felt they

had a reasonably high comp, but
rather high level of confidence

in the general state of the data
analytics.

But when it comes down to it,
when you're running the demand,

you've got a lot of stuff that's
showing up green on your

dashboard.

Yeah.

But this still turns into red um
when it when you get to the

point of did I generate the
required pipeline to hit the

revenue targets.

Seth Marrs: Gosh, and I think
this leads in, I want to ask one

question, one more question
around the report.

A really kind of scary metric.

And I think you're talking to it
here a little bit.

It's 85% of marketers spend more
time fixing problems and

creating programs.

That was kind of disheartening.

But when w w what you've talked
about so far kind of leads to

it, right?

Like you're creating all these
vanity metrics and putting them

in place, but you really don't
believe in them, and then you

end up having to go fix them on
the back end.

Like, how do companies like one
recomm like what recommendations

would you give?

Because it also is a capacity
opportunity.

If you're spending 80 if 85% are
spending that much time fixing

stuff, if you could redirect
that into something like signal

aggregation.

SPEAKER_00: Eh bingo.

Um yes.

So um, and let me just
acknowledge as a CMO, I've been

one who sat there saying, Oh, we
got this problem.

Let's get more data.

Maybe we should get more tools.

You and I talked about this.

The rate at which AI-powered um
sales tech tools, Martech tools

are coming out.

You can't even keep up with
them.

It's it's it's it's every week
there's another tool, and which

one should you buy?

And and so what people do is I
buy more tools, I buy more data,

I get more signals, I generate
more content, I run more

campaigns, each in their
respective style without a

connective thread that unifies
all those.

So, what do you do about this,
especially this problem of

people spending lots more like
more time fixing than actually

creating?

So, what are things what are
things that the survey revealed

um the validated things that
I've seen or experienced?

So a few things.

One, um, buying more tools does
not result in more ROI.

Um, people who had bigger tool
stacks, as it's as the as the

Martex tool tool stack got
bigger, their their ROI went

lower.

They more of them questioned
their IRI after they bought more

tools.

Slim down your stack and think
about it in layers.

What do I have in my signal
layer, my content layer, my

activation layer?

Um, focus on quality of signals
over volume.

And so take that holistic view,
not just first party and

third-party data, but and not
just clicks, but what are the

real drivers of opportunities in
my space from a business

perspective?

Do I have those signals captured
well?

And am I acting on those not
just once to create an account

list, but continuously in a
narration loop that's going to

keep um adjusting these based on
new things happening?

So one of them is data, you
know, one of them is leaner,

leaner stacks, verified signals.

Um, better attribution came out
in in spades.

So you have to be transparent in
your end-to-end attribution so

you know what works.

That's its own kind of can of
worms, as anyone in sales and

marketing know.

Um drive your content off data,
not what somebody's thinking,

and and not just what the LLM
tool tells you, you know, should

be found, but um create content
that's informed by verified

buyer um um insights and
assumptions.

Um you can use AI to help you
know what some of those are and

and use AI generated templates,
but like make sure this stuff is

sort of personalized and
relevant.

And then like, how can you make
your systems operate in a much

more efficient way?

So you replace the time years,
like some of that's connectivity

via API, some of it is leaner
stocks.

Do I people said they have tons
of redundant capabilities across

multiple tools in their Martex
stack?

Seth Marrs: Well, which creates
complexity that creates wasted

time, but together.

SPEAKER_00: Yeah, and so um, and
lastly, I'd add my own dose to

it, which is um do less and
obsess.

Like what's really gonna move
the needle and and do that

really well across the signals,
the content, and the activation,

um, and and have the right
metrics in place that are

really, really oriented around
pipeline, what's really um

contributing to the building of
qualified pipeline.

So I double down on those
things, and um um, but it's it's

connecting all those things that
also were often siloed.

Um, so that's that's what I the
survey said that, and and I hear

the same thing from from CMOs
that I talked to.

Seth Marrs: Fantastic Bill, it's
so great to have you on.

I love the insects.

I always learn something when I
talk to you.

So yeah, it really is.