Subscribe
Share
Share
Embed
"Personalization is finding that balance — you can give customers everything they want, but it doesn't make the business happy. True personalization is finding the perfect balance." – Eloise
Personalization only works when the infrastructure beneath it is built for business reality, not hype.
In this episode, Jacqueline sits down with Eloise Gillespie, Associate Director of Personalization and Martech at Optus, to unpack what it actually takes to make AI and personalization deliver at scale. Eloise brings hard-won experience from telecom, wagering, and hospitality, including managing a nine-figure bet budget and building customer-level optimization engines that won CFO approval by anchoring every decision in margin, ROI, and commercial viability.
They cover the infrastructure work that must happen before AI can deliver value, why treating data as a product changes everything, and how to make personalization credible to finance teams. If you're done with quick fixes, this one is for you.
Timestamps
01:19 — The biggest bet: betting on yourself
04:14 — How Salesforce Marketing Cloud changed everything
08:45 — Why LTV is dead (or at least overrated) and what metric actually matters
18:05 — Leading with commercial impact: a $XXX,XXX,XXX budget gap to win CFO trust
26:08 — Proving personalization to the CFO: Leading with dollars, not fluff
38:22 — System of record vs. system of context: What CDPs actually do
45:38 — Building an incentive optimization engine across bet types, margins, and interaction contexts
52:07 — AI-driven personalization: What's real today vs. what's still on the horizon
Sponsor
Brought to you by Hightouch, the leading composable CDP and decisioning platform trusted by brands like Domino's, Chime, and Aritzia. 90% of customers have a real use case live within their first week, delivering world-class personalization at scale. Learn more at hightouch.com/msom.
Creators and Guests
What is Making Sense of Martech?
Unfiltered takes on the biggest shifts in marketing technology. We spotlight what matters, who's leading (or lagging), and what's next. In Martech, clarity is power — and we're here to deliver it.
00;00;06;09 - 00;00;21;26
Unknown
Welcome to the Making Sense of MarTech podcast, where we interview leaders and put them in the hotseat. I'm Jaclyn Friedman, founder of monarch and global head of advisory for the MarTech weekly. Today's episode is all about making the case for the marketing engineer as a distinct, high leverage role that bridges data
00;00;21;26 - 00;00;25;06
Unknown
engineering and marketing to deliver profit and execution.
00;00;25;08 - 00;00;29;03
Unknown
But first, a quick intro on our guest, Jeff Lee.
00;00;29;03 - 00;00;37;25
Unknown
Jeff is a lifecycle marketing technical lead at com. He's gone from it at IBM to web and email development to high volume lifecycle at Flipboard. And now. Com
00;00;37;25 - 00;00;38;19
Unknown
where he runs a
00;00;38;19 - 00;00;48;21
Unknown
lean team and ships at massive scale across email, push and in-app. He's a leading voice for legitimizing the term marketing engineer. And Jeff, I'm so excited you're here. Welcome.
00;00;48;21 - 00;00;50;21
Unknown
Thank you for having me. It's good to be here.
00;00;50;23 - 00;00;51;20
Unknown
Yeah.
00;00;51;20 - 00;00;58;23
Unknown
So to dive in, first up, we have our rapid fire round. So five quick hits. What was your first martech platform?
00;00;58;23 - 00;01;01;10
Unknown
Well, we built it in for
00;01;01;10 - 00;01;01;19
Unknown
Tell
00;01;01;19 - 00;01;02;18
Unknown
me more.
00;01;02;18 - 00;01;04;15
Unknown
So Flipboard was.
00;01;04;15 - 00;01;13;07
Unknown
We built it on Python and, lots of scripts and data database. And it was basically like a, like an adorable or a braze,
00;01;13;12 - 00;01;14;22
Unknown
Oh, wow.
00;01;14;22 - 00;01;18;12
Unknown
less feature rich. It was really just,
00;01;18;12 - 00;01;24;00
Unknown
very difficult to work with. And then moving over to those new platforms is like a dream come true compared to that.
00;01;24;00 - 00;01;24;17
Unknown
But
00;01;24;24 - 00;01;27;09
Unknown
Okay, well, that settles the build versus buy debate real
00;01;27;09 - 00;01;29;09
Unknown
quick.
00;01;30;06 - 00;01;40;02
Unknown
All right, so what has been your biggest oops moment? So whether a campaign went sideways because of tech or data issues, or maybe it's just building your own ISP? I don't
00;01;40;02 - 00;01;40;14
Unknown
know.
00;01;40;14 - 00;01;46;09
Unknown
Okay, so this one, this one is kind of a bit like I, I had to create a whole story for this one because it was
00;01;46;09 - 00;01;48;06
Unknown
large like that. It
00;01;48;20 - 00;01;54;20
Unknown
That. You know what? In my eyes, the biggest fuck ups or the most biggest learning lessons. So.
00;01;54;20 - 00;01;58;19
Unknown
So this is. Yeah, it was just happened recently.
00;01;58;19 - 00;02;05;07
Unknown
it was kind of like crunch time where, like, we're getting ready to launch another app, like an entirely new app.
00;02;22;26 - 00;02;24;07
Unknown
Oh.
00;02;24;07 - 00;02;34;02
Unknown
It kind of dropped off from like 70% of people receiving an email within the first day to 15, 10%, something like
00;02;34;06 - 00;02;36;12
Unknown
Oh, dear.
00;02;41;14 - 00;02;42;27
Unknown
Where did we go wrong?
00;02;43;09 - 00;02;48;26
Unknown
exactly. So it turns out and like it started with, okay,
00;03;07;12 - 00;03;10;20
Unknown
Oh.
00;03;10;23 - 00;03;12;05
Unknown
Yeah. Of course.
00;03;12;05 - 00;03;17;14
Unknown
other countries where it's kind of okay to assume that you're marketed in,
00;03;17;14 - 00;03;18;05
Unknown
You are.
00;03;18;05 - 00;03;24;05
Unknown
You're automatically opted in. Wow.
00;03;24;05 - 00;03;30;23
Unknown
yeah, that was a some huge thing. So I was just looking at it. And nobody else is paying attention to this. Right. Because I just noticed and like, hey, you guys look at this.
00;03;30;23 - 00;03;33;28
Unknown
And like, did you do anything recently? Like, no, we didn't change anything.
00;03;33;28 - 00;03;41;12
Unknown
It boiled down to three lines of code on the server with, like, kind of like a weird double negative and all that. It was
00;03;48;23 - 00;03;49;05
Unknown
Oh.
00;03;55;19 - 00;03;57;04
Unknown
Right.
00;04;07;05 - 00;04;15;05
Unknown
Of course.
00;04;15;08 - 00;04;15;26
Unknown
Yup.
00;04;15;26 - 00;04;22;12
Unknown
So it sounds like a double negative is never a value, both grammatically, but also in software development.
00;04;22;21 - 00;04;27;11
Unknown
negative and like if, if something isn't already set, like it was kind of like a weird.
00;04;27;26 - 00;04;31;26
Unknown
Yeah. Not the most logical or clean way of organizing.
00;04;31;26 - 00;04;37;11
Unknown
Oh, wow. Yeah. That's wild.
00;04;37;11 - 00;04;44;03
Unknown
one line of code later, it took like a couple of weeks to kind of sort through it and like, get it deployed back up to 80%, like. And we're all
00;04;44;03 - 00;04;48;26
Unknown
Wow. Is.
00;04;48;26 - 00;04;53;10
Unknown
Yeah. Of course. And were you using a deliverability tool for that or,
00;04;53;10 - 00;04;54;17
Unknown
how was it discovered?
00;04;54;17 - 00;04;59;11
Unknown
I was just digging around inside of amplitude, kind
00;05;03;01 - 00;05;07;28
Unknown
I think new users seem to be buying less like, within
00;05;09;13 - 00;05;10;08
Unknown
Got it.
00;05;10;08 - 00;05;14;12
Unknown
sliver of people that actually grew out to be a much larger piece of the pie than we thought.
00;05;14;16 - 00;05;14;29
Unknown
Yeah.
00;05;15;03 - 00;05;22;29
Unknown
Yeah. No kidding. Okay, now let's see. What's one tool you think is completely overrated or overhyped in martech?
00;05;22;29 - 00;05;26;13
Unknown
I feel like segment the CDP
00;05;28;28 - 00;05;36;13
Unknown
I don't I'm trying not to, like, incur, you know, the wrath of, you know, millions of people,
00;05;36;27 - 00;05;41;17
Unknown
There there is no rest to be had except for truth and speaking. Truth to power is far more important.
00;05;43;24 - 00;05;47;21
Unknown
I mean, it does the job, but then you get to buy all the extra stuff to actually make it do what you want.
00;05;47;25 - 00;06;04;05
Unknown
Yeah. And you also have to pay for your data more than once. And I don't like paying. I'm a cheapskate. I want to pay only what I'm. I'm amusing. And if I already own my data, why would I do it again? But maybe that's just me. Maybe I just would like to save money. Okay. And something we ask everyone.
00;06;04;06 - 00;06;06;27
Unknown
Who is someone you admire? Whether it's professional or personal?
00;06;06;27 - 00;06;09;18
Unknown
I would say both anybody who
00;06;09;18 - 00;06;20;15
Unknown
is kind of over 40 and able to navigate this job market or, you know, be able to find a job or it's tough out there.
00;06;21;00 - 00;06;21;23
Unknown
It is.
00;06;21;23 - 00;06;26;23
Unknown
you're in it or dealing with it, you know, keep track and you know,
00;06;26;23 - 00;06;28;07
Unknown
support
00;06;28;18 - 00;06;29;28
Unknown
Yes,
00;06;29;28 - 00;06;44;28
Unknown
I agree. Definitely. Leveraging your network and and the communities around it. I wholeheartedly agree. All right. Before we fully dive in, I want to ask one hot take because I. We're going to get serious. But I'm curious what your thoughts are on
00;06;44;28 - 00;06;52;22
Unknown
warehouse. Native solves everything. A CDP replaces engineers or LMS can safely choose audiences. Which is the most dangerous and why?
00;06;52;22 - 00;06;58;26
Unknown
I, you know, since we were talking about segment, I think the CDP replacing engineers, I don't think is true.
00;06;58;26 - 00;07;05;19
Unknown
This engineer still has to implement the CDP. It's really it's streamlined some things. It takes a few hours off of their plate.
00;07;05;19 - 00;07;11;02
Unknown
But our stuff is pretty stable. Like, once we once we get it in the into place,
00;07;11;07 - 00;07;11;24
Unknown
Yeah.
00;07;11;24 - 00;07;13;17
Unknown
it's not saving anything.
00;07;13;17 - 00;07;18;20
Unknown
I feel you on that. Okay, let's set the stage and really dive into
00;07;18;20 - 00;07;26;20
Unknown
things that. Let's define what the job needs to get done before we really get into architecture. So one sentence, zero jargon. What is a marketing engineer
00;07;26;28 - 00;07;29;14
Unknown
And what problem does it own that no one else does?
00;07;39;20 - 00;07;41;20
Unknown
No big deal. Just casual.
00;07;43;27 - 00;07;52;10
Unknown
achieve, right? There's just it's in the app. It's in the backend, it's in the data. It's in the CDP, it's in the CFP, it's in, you know, all of these things.
00;07;52;13 - 00;07;54;25
Unknown
All I'm ever thinking about is marketing. And
00;07;55;05 - 00;07;56;04
Unknown
Yeah.
00;07;56;04 - 00;07;59;17
Unknown
systems and people together to achieve,
00;07;59;17 - 00;08;02;15
Unknown
you know, whatever your marketing program happens to be?
00;08;02;15 - 00;08;09;15
Unknown
Yeah. I guess a double click on that. Are your goals aligned to engineering goals, or are your goals aligned to marketing goals?
00;08;13;08 - 00;08;19;29
Unknown
Next, if only every martech and marketing engineer team. That was always the case.
00;08;19;29 - 00;08;29;18
Unknown
Okay, so if this role vanished while you were no longer in that position and no one else backfill to you, what would break first? Speed, reliability or profitability?
00;08;29;18 - 00;08;31;19
Unknown
Profitability? Probably.
00;08;31;23 - 00;08;33;21
Unknown
Yeah.
00;08;33;21 - 00;08;34;04
Unknown
the
00;08;34;07 - 00;08;38;22
Unknown
out. If everybody all of a sudden 80% of people are opting out, you are losing money.
00;08;39;01 - 00;08;44;18
Unknown
That would be very bad. Yeah. If that was, continued and sustained. Agreed. Okay.
00;08;44;18 - 00;08;52;00
Unknown
Similar to how a go to market engineer or moms or rev ops in this current stage, like where does a marketing engineer come
00;08;52;00 - 00;08;59;25
Unknown
into play in terms of decision making on tools and and strategy? Like what is the profit lever only marketing engineering can pull?
00;08;59;25 - 00;09;07;05
Unknown
So from my. What I'm trying to achieve usually is probably a deeper product integration,
00;09;23;08 - 00;09;29;10
Unknown
And that's probably like that. That level of getting into the weeds on the product is probably a difference. There.
00;09;29;10 - 00;09;45;27
Unknown
I agree, I think also, and this is the beauty of martech, both from the engineering perspective and the non like the technical but non engineering perspective is. I see it as an access point where you're thinking holistically. Both business objectives, customer objectives and internal stakeholder objectives. And very few teams or team
00;09;45;27 - 00;09;55;17
Unknown
members are doing that or even have to because it's a little bit more one, one track minded, versus you've got a lot of customers in this capacity.
00;09;55;17 - 00;09;56;08
Unknown
Oh, yeah.
00;09;56;08 - 00;09;57;17
Unknown
Without a doubt.
00;09;57;17 - 00;10;03;28
Unknown
What new capabilities become possible when someone stitches data to a reasoning to activation?
00;10;03;28 - 00;10;05;05
Unknown
does the magic happen?
00;10;18;26 - 00;10;24;09
Unknown
there that happens where maybe if they're disconnected, you don't get right because you don't maybe.
00;10;24;09 - 00;10;28;17
Unknown
Yeah. If you're receiving an email that's not well integrated,
00;10;28;17 - 00;10;31;15
Unknown
It's a destroying process, right. Because it's destroying the experience
00;10;31;15 - 00;10;48;15
Unknown
I wholeheartedly agree, and I know it's usually on brand to think through that, but I think both the engineering and marketing, both sides of the coin should always be thinking you want one unified experience. If it's just to. Your customers are going to not be as inclined to understand what you're trying to do.
00;10;48;15 - 00;10;51;17
Unknown
And that's one of the I think missing steps in rebrands
00;10;51;17 - 00;11;03;19
Unknown
is I forget just how many emails there are and how most teams are not fully set up to have a universal email design system so that they can automatically populate and update
00;11;03;19 - 00;11;06;02
Unknown
all of their templates relatively quickly.
00;11;06;23 - 00;11;14;14
Unknown
There are, of course, some teams that have been able to modernize and do that. But if you're on legacy tech, like good luck, that's that's not how it's going to happen.
00;11;14;14 - 00;11;22;02
Unknown
Speaking of legacy tech, I am curious, what is your dream martech stack? If you could just pie in the sky, start a new.
00;11;23;03 - 00;11;24;29
Unknown
having a data warehouse.
00;11;24;29 - 00;11;25;03
Unknown
Like
00;11;25;05 - 00;11;28;18
Unknown
Oh, yeah.
00;11;28;21 - 00;11;33;18
Unknown
I love Databricks, yeah, they're both great.
00;11;33;18 - 00;11;40;24
Unknown
within the that, like, notebook environment was great. To be able to do those things, especially with the AI stuff nowadays.
00;11;41;01 - 00;11;41;22
Unknown
Yes.
00;11;41;22 - 00;11;51;01
Unknown
You can talk to the AI to produce your notebook, to create your segment and do some interesting, transformations on, like, whatever the data is.
00;11;51;03 - 00;11;52;08
Unknown
Feed that into
00;11;52;08 - 00;11;53;11
Unknown
our, our
00;11;53;11 - 00;12;02;06
Unknown
engagement platform, whether that's, braze, iterable, mowing edge. You know, any one of those three? I feel, you know, I'm sure there's other ones, but,
00;12;02;06 - 00;12;05;14
Unknown
very capable systems that can do what,
00;12;05;14 - 00;12;10;00
Unknown
like, deliver the messages and be able to see what's going on. And then on the.
00;12;10;00 - 00;12;14;28
Unknown
Probably like an amplitude, like an analytics suite to see what's going on.
00;12;15;01 - 00;12;25;07
Unknown
Do some maybe testing. Analysis of what people are doing and able to actually be able to sync back in. So we do use some amplitude to send things back into the KPIs
00;12;25;18 - 00;12;26;17
Unknown
Oh, nice.
00;12;28;04 - 00;12;29;15
Unknown
that can talk to
00;12;29;15 - 00;12;30;15
Unknown
the platforms as well.
00;12;30;15 - 00;12;33;20
Unknown
So I use Google Sheets, Google Apps Script a lot.
00;12;34;02 - 00;12;34;24
Unknown
Oh, nice.
00;12;34;24 - 00;12;35;17
Unknown
It's not like
00;12;35;17 - 00;12;39;11
Unknown
perfect, but like if there was something that was a little bit more,
00;12;39;11 - 00;12;44;11
Unknown
Wouldn't Airtable be a little bit more perfect? Or do you mean even a step change from there?
00;12;51;25 - 00;12;56;07
Unknown
even just a, like a bare metal python box. I can, like, receive and and
00;12;57;03 - 00;13;00;19
Unknown
That's very, you know, unfortunate like an engineering answer. But
00;13;01;04 - 00;13;11;11
Unknown
No, it's a good answer. And I almost would implore you to look into a composable CDP, because I think it will change the way you consider CDPs in the tech stack, at least currently.
00;13;11;11 - 00;13;11;29
Unknown
Sure.
00;13;11;29 - 00;13;14;05
Unknown
Because of that data centricity in the warehouse.
00;13;14;05 - 00;13;16;05
Unknown
I'd be curious to hear your thoughts as a follow up.
00;13;16;05 - 00;13;17;28
Unknown
But since we're getting nerdy and talking
00;13;17;28 - 00;13;23;24
Unknown
about the tech, I want to pop the hood on comms lifecycle engine.
00;13;23;26 - 00;13;28;23
Unknown
How big is the team? What is your stack look like? What's missing? You name it.
00;13;55;13 - 00;13;56;20
Unknown
It's hard to find.
00;14;02;13 - 00;14;13;13
Unknown
So the four of us, and then we work with the other teams content team to, to do a lot of the writing for the emails on the news, like the weekly newsletters that are going out. So mostly, you know, we're not
00;14;13;13 - 00;14;14;28
Unknown
copywriters or anything like that.
00;14;15;06 - 00;14;20;17
Unknown
Yeah. That's a very lean team for. How many emails are you sending a month or a year?
00;14;26;20 - 00;14;27;08
Unknown
Well,
00;14;27;08 - 00;14;28;08
Unknown
yeah.
00;14;29;07 - 00;14;34;19
Unknown
Yeah, that's a very small team. If we're just doing, you know, emails per person.
00;14;34;19 - 00;14;35;24
Unknown
That's impressive.
00;14;35;24 - 00;14;41;02
Unknown
Yeah, it's a lot. I mean, I think we did. I was looking at it before, I think we probably we did about
00;14;41;02 - 00;14;44;17
Unknown
200 or 240 distinct campaigns. Right.
00;14;44;20 - 00;14;46;01
Unknown
Wow.
00;14;46;01 - 00;14;53;25
Unknown
them into our automations, but like 240 individual templates with different text on them.
00;14;54;01 - 00;15;00;02
Unknown
So basically, an email a day on business days. That's insane.
00;15;02;17 - 00;15;07;26
Unknown
And so you mentioned that you're on the consumer side of the business. Is there like a B2B side of the business? Two or.
00;15;07;26 - 00;15;12;20
Unknown
Yeah. So we have an enterprise business that does, the health alarm health.
00;15;13;26 - 00;15;17;02
Unknown
or oriented, like, more, clinical
00;15;17;07 - 00;15;23;13
Unknown
Yeah.
00;15;23;16 - 00;15;24;10
Unknown
That makes sense.
00;15;24;10 - 00;15;25;25
Unknown
employer bases. Yeah.
00;15;25;25 - 00;15;41;26
Unknown
Understood. So on the consumer side, what's your tech stack? And I'm selfishly curious because I love B2B to see and B2C to be plg companies where you have both signs and how actually you kind of need them to be both distinct but sometimes combined. And so I'm
00;15;41;26 - 00;15;43;21
Unknown
curious, the insights there.
00;15;49;15 - 00;15;50;03
Unknown
Yeah.
00;15;50;03 - 00;15;55;15
Unknown
We don't even talk to them. Right. They are very careful about you guys will never see this data now
00;15;55;15 - 00;15;56;26
Unknown
Good.
00;15;58;11 - 00;15;59;19
Unknown
or something. GCP.
00;15;59;19 - 00;16;03;08
Unknown
But then on the consumer side, we have
00;16;06;05 - 00;16;10;28
Unknown
I still use mode and Redshift. I think we're moving off of that towards red. Databricks.
00;16;11;04 - 00;16;15;19
Unknown
Oh, good. I mean, I love redshift, but it's time to time to centralize.
00;16;15;19 - 00;16;20;06
Unknown
It's pretty darn good.
00;16;20;06 - 00;16;22;25
Unknown
I think we paid up, ponied up for the genie thing I
00;16;22;25 - 00;16;25;04
Unknown
You know, I don't blame you.
00;16;25;09 - 00;16;25;26
Unknown
Yeah.
00;16;29;26 - 00;16;34;29
Unknown
Iterable catalog. I think this is like a really important piece of what what I do.
00;16;34;29 - 00;16;47;14
Unknown
It's such an important feature. And expand. Especially. I don't know if you've ever had to work in a legacy tech where it's all relational databases. I feel like this is the next frontier and version of being able
00;16;47;14 - 00;16;49;09
Unknown
to do it.
00;16;49;09 - 00;16;52;02
Unknown
I like especially. Yeah, I'd like to be able to
00;16;52;02 - 00;16;56;16
Unknown
prevents us from having to spin up a lot of other servers like,
00;16;56;16 - 00;16;57;07
Unknown
Exactly.
00;16;57;07 - 00;17;00;08
Unknown
that engineers wouldn't, probably wouldn't spin up for us.
00;17;03;03 - 00;17;05;26
Unknown
I mean, I think that's the bane of everyone's existence, but also,
00;17;05;26 - 00;17;12;04
Unknown
Well, it's good to hear that you essentially are working with the the dream tech stack you outlined in the first place.
00;17;12;06 - 00;17;17;21
Unknown
Yeah. I mean, I guess the thing that we're missing is that composable CDP that you're right, like, I haven't had a chance to
00;17;17;21 - 00;17;18;12
Unknown
play with those.
00;17;18;26 - 00;17;20;15
Unknown
We will talk after this.
00;17;20;15 - 00;17;23;14
Unknown
I've got some ideas for you.
00;17;23;14 - 00;17;31;05
Unknown
Yeah I know like I've tried to I've, I've tried to simulate one basically. And in our own it using Databricks. Right. Like
00;17;31;05 - 00;17;31;27
Unknown
wrote the thing and
00;17;31;27 - 00;17;33;01
Unknown
I can sync the stuff.
00;17;33;03 - 00;17;34;09
Unknown
But there's obviously I'm sure
00;17;34;09 - 00;17;35;09
Unknown
there's other tools that
00;17;35;20 - 00;17;35;29
Unknown
I
00;17;35;29 - 00;17;44;13
Unknown
have a feeling it's a it's a bit of a preview to a question I'll ask later on in terms of observability, but we will, I digress. We will. We'll get there.
00;17;44;13 - 00;17;51;11
Unknown
So with such a small core team, that's both impressive because, I mean, I've seen teams
00;17;51;11 - 00;17;52;13
Unknown
very large
00;17;52;13 - 00;17;58;02
Unknown
doing less in terms of personalization, customization and actually output.
00;17;58;05 - 00;18;02;29
Unknown
I'm curious, how do you sustain these billions of sense without burning the engine
00;18;02;29 - 00;18;05;22
Unknown
entirely? Like what are the latency
00;18;05;22 - 00;18;08;16
Unknown
targets? How do you guys ensure guardrails are
00;18;08;16 - 00;18;19;09
Unknown
are you paying attention? Deliverability? It sounds like you just explore and have fun, you know, tinkering as we all do. But are there certain systems and hygiene and processes you guys have in place to accommodate?
00;18;19;09 - 00;18;23;26
Unknown
So for deliverability, we we work with the deliverability service.
00;18;24;02 - 00;18;24;24
Unknown
Which one?
00;18;27;08 - 00;18;28;10
Unknown
Oh, nice.
00;18;28;10 - 00;18;32;21
Unknown
team. They're, you know, having a deliverability expert is
00;18;32;21 - 00;18;35;01
Unknown
Makes a difference.
00;18;35;01 - 00;18;36;01
Unknown
People forget
00;18;36;01 - 00;18;46;16
Unknown
you need to have your all of your pipes running. Otherwise your kitchen's not going to work. You're not gonna be able to to make the dream happen.
00;18;53;21 - 00;18;59;03
Unknown
So we we kind of built in, like, our our, our internally, our team calls it the Sorting Hat
00;18;59;05 - 00;19;01;11
Unknown
Ooh, getting Harry Potter up in here.
00;19;23;00 - 00;19;24;17
Unknown
I.
00;19;28;15 - 00;19;29;23
Unknown
Smart.
00;19;29;23 - 00;19;35;01
Unknown
And, you know, but still allowing for, like, the blasts to sneak in and that kind of stuff.
00;19;35;01 - 00;19;44;07
Unknown
So we want to do like a global shout out, but hold out or anything, but it's like, really is the automation stuff that before we did the Sorting Hat, we never knew, like how many messages might you receive? You might
00;19;44;07 - 00;19;49;11
Unknown
be receiving five different messages from five different automations, and they're all overlapping. And,
00;19;49;11 - 00;19;51;04
Unknown
didn't want to do a global like Max
00;19;51;04 - 00;19;53;25
Unknown
because sometimes things need to get there.
00;19;53;25 - 00;19;54;21
Unknown
Right. But,
00;19;54;21 - 00;19;55;09
Unknown
That's true.
00;19;55;09 - 00;20;04;20
Unknown
the Sorting Hat I think is fundamentally changed. Kind of like our, our ability to, like, be confident that we're not over messaging
00;20;05;03 - 00;20;08;18
Unknown
Yeah. I love the naming of that, too,
00;20;08;18 - 00;20;11;03
Unknown
because
00;20;11;03 - 00;20;30;02
Unknown
I. I love it because I think we all think about that. And I'm like, okay, how do we waterfall suppression list? How do we prioritize campaigns based off of categorization or things like that? And, I definitely if you're open to it, I would love to ask my questions about the Sorting Hat because it sounds really cool and I
00;20;30;02 - 00;20;31;17
Unknown
think folks could learn
00;20;31;17 - 00;20;31;24
Unknown
more.
00;20;31;24 - 00;20;45;05
Unknown
We've got the Sorting Hat. I'm super intrigued. Who and how do we decide someone is in Slytherin? Gryffindor, Hufflepuff, you name it. How do we do this or what? What is the behind the scenes? Some of the filtration system.
00;20;45;05 - 00;20;46;07
Unknown
So
00;20;49;28 - 00;20;52;08
Unknown
Who's not currently in something right now.
00;20;55;13 - 00;20;57;08
Unknown
Okay, so we're using a score.
00;20;57;08 - 00;21;01;19
Unknown
Yeah. Or it's like a yeah it's a priority. There's a priority to it. It's like
00;21;01;24 - 00;21;02;10
Unknown
Nice.
00;21;02;10 - 00;21;11;24
Unknown
So you're you're at zero. So find anybody who's a zero. And then they come into this hat. And then we have like your waterfall. Right. Probably the top
00;21;23;17 - 00;21;28;24
Unknown
Right. So if you're ready to buy and whatever signals where you happen to be using, you get the score of ten. We'll say
00;21;28;24 - 00;21;31;20
Unknown
give Jacqueline a score of ten and then toss her into this workflow
00;21;39;13 - 00;21;40;28
Unknown
or you could be
00;21;40;28 - 00;21;45;04
Unknown
Okay.
00;21;45;06 - 00;21;47;28
Unknown
Got it.
00;21;49;19 - 00;21;50;18
Unknown
That makes sense.
00;21;50;18 - 00;21;53;04
Unknown
down the chain. We probably have
00;21;53;04 - 00;21;55;07
Unknown
the chain is probably,
00;22;03;05 - 00;22;03;21
Unknown
Yeah.
00;22;03;21 - 00;22;10;24
Unknown
you're not dormant? And then the last ones are usually just catch all this like, hey, it's just here's a blog. Like,
00;22;12;27 - 00;22;16;10
Unknown
Yeah. Of course. Yeah.
00;22;16;10 - 00;22;17;13
Unknown
is probably the,
00;22;17;13 - 00;22;19;26
Unknown
the best thing that we can do because
00;22;19;26 - 00;22;21;08
Unknown
For sure.
00;22;21;26 - 00;22;28;23
Unknown
going to click on the you're not going to listen to something unless you can read it and like learn a little bit more about what you know with the,
00;22;28;23 - 00;22;30;15
Unknown
product is trying to do for you.
00;22;30;15 - 00;22;47;29
Unknown
I love that. And in terms of the scoring methodology, is this something homegrown? Is this something you guys are building elsewhere with a third party? How are you scoring? Because I think scoring to most folks is a B2B concept, when in reality, I think it's actually one of the things that B2B
00;22;47;29 - 00;22;57;29
Unknown
and B2C should learn from each other, because there's there's pros and cons of the signals, whether it's from a graphic techno graphic or actually engagement, literally behavioral based.
00;23;16;05 - 00;23;19;14
Unknown
You know, on a personal score that would be interesting. Like like okay, yours.
00;23;19;14 - 00;23;23;11
Unknown
It might be a combination of scores, right? I like it, but that would be
00;23;23;11 - 00;23;25;02
Unknown
probably the next level thing. All
00;23;25;29 - 00;23;28;20
Unknown
Hey, you gotta start somewhere.
00;23;28;20 - 00;23;29;13
Unknown
Okay. Awesome.
00;23;29;13 - 00;23;32;21
Unknown
Oh. So thank you for the the sidebar with Sorting Hat because I think one
00;23;32;21 - 00;23;34;03
Unknown
no one thinks like that.
00;23;34;06 - 00;23;39;25
Unknown
And I really like the the turn of phrase and concepts. And now on your team came up with that. So we can give
00;23;39;25 - 00;23;46;00
Unknown
credit to.
00;23;46;03 - 00;23;54;25
Unknown
Oh, even more impressive. You know I get it.
00;23;54;25 - 00;23;59;09
Unknown
Well y'all, maybe she gets a shout out later on of the day.
00;23;59;09 - 00;23;59;29
Unknown
That's also
00;23;59;29 - 00;24;02;08
Unknown
in your eyes, there is
00;24;02;08 - 00;24;04;19
Unknown
always going to be, to your point,
00;24;04;19 - 00;24;07;11
Unknown
a V1, the NVP,
00;24;07;11 - 00;24;10;17
Unknown
platform tool, product solution, like a sorting hat.
00;24;10;17 - 00;24;15;25
Unknown
Like where do you think duct tape and paperclips belong and where do they eventually just become tech debt?
00;24;18;00 - 00;24;20;03
Unknown
Everyone does. And and there's no
00;24;20;03 - 00;24;25;01
Unknown
shame in duct tape. It's only a shame if all of a sudden it's no longer working. So I guess
00;24;25;01 - 00;24;25;08
Unknown
Yeah.
00;24;25;17 - 00;24;31;01
Unknown
What is one of the first signals of a crack in the system that like. Okay, no, we need deadbolts now.
00;24;31;07 - 00;24;32;00
Unknown
Yeah. I mean,
00;24;43;04 - 00;24;47;27
Unknown
again, then it's probably needs to be systematized, right? Like and
00;24;47;27 - 00;24;48;20
Unknown
Yeah.
00;24;48;20 - 00;24;50;12
Unknown
with engineering backing.
00;24;50;12 - 00;24;51;18
Unknown
It. There's
00;24;51;18 - 00;25;00;11
Unknown
some systems that I might add did with duct tape like five years ago, and they're still there just because it didn't achieve kind of like
00;25;00;11 - 00;25;05;06
Unknown
that mission critical thing where it's like, hey, if this breaks, there's going to be consequences, right?
00;25;05;06 - 00;25;05;11
Unknown
Yeah.
00;25;06;05 - 00;25;11;05
Unknown
of them don't ever need it and some of them need it for like, almost like the next water.
00;25;11;05 - 00;25;13;27
Unknown
Right? But they have to be a quarter because they're not going to work
00;25;13;27 - 00;25;15;00
Unknown
on it the same way in
00;25;15;27 - 00;25;30;11
Unknown
Totally. Yeah. I think, a lot of folks immediately just assume. Go straight for the. And I'll be all. I actually really prefer starting with duct tape, because I want to prove out the process before we commit time resources and actually automating or
00;25;30;11 - 00;25;38;11
Unknown
repeatable components to it. And so, I'm a huge fan of like, we're going to do it manually to see if it's even worth our time.
00;25;39;04 - 00;25;46;11
Unknown
I mean, and with the new tool, Zapier, any at the end or wherever you'd be like, people are probably able to automate them anyway, right?
00;25;46;16 - 00;25;47;26
Unknown
Yeah,
00;25;47;26 - 00;25;51;16
Unknown
yeah. It's true.
00;25;51;16 - 00;25;53;27
Unknown
and, all that kind of stuff. There's where
00;25;55;18 - 00;26;02;05
Unknown
For sure. And if you're using a middle management tool, just make sure you have your data protection addendum in place, because
00;26;02;05 - 00;26;03;20
Unknown
otherwise
00;26;03;20 - 00;26;11;24
Unknown
we got other issues to deal with. But we've kind of alluded to this question vendor versus homegrown platforms like
00;26;11;24 - 00;26;16;12
Unknown
since I know now that you had an in-house tool before, like
00;26;16;12 - 00;26;20;26
Unknown
what did you learn from a user experience perspective that
00;26;20;26 - 00;26;23;22
Unknown
the pros and cons of it were and,
00;26;23;22 - 00;26;29;23
Unknown
do you ever foresee an opportunity, especially with AI building homegrown again?
00;26;29;23 - 00;26;37;14
Unknown
For that system. Probably wouldn't like a for, like an email system. There's a bunch of other stuff besides UA that goes into it, or it's not.
00;26;37;14 - 00;26;38;09
Unknown
Not a
00;26;38;15 - 00;26;38;20
Unknown
If
00;26;38;20 - 00;26;41;08
Unknown
I could insert, sitcom like
00;26;41;08 - 00;26;44;05
Unknown
applause right here, I would, because
00;26;44;05 - 00;26;45;20
Unknown
I think,
00;26;45;20 - 00;26;50;27
Unknown
a lot of I don't want to say engineers, but I would say engineering leadership believe that
00;26;50;27 - 00;26;57;16
Unknown
just because they're builders and really great at building software means they can build all software regardless of their expertise.
00;27;03;18 - 00;27;07;24
Unknown
When there's a bounce, you need to. You should honor it. Right. And then the
00;27;08;01 - 00;27;09;09
Unknown
Yes.
00;27;09;09 - 00;27;12;03
Unknown
a bounce. Right. To understand, to take this person off. Because
00;27;12;26 - 00;27;14;13
Unknown
didn't handle bounces like we just
00;27;16;08 - 00;27;18;08
Unknown
we didn't have like the proper
00;27;20;12 - 00;27;33;09
Unknown
those are the things it's more of those like domains of like email deliverability. You're going to get hurt, right? Like the engineering might not be aware of those, and it might take a little extra time to kind of get that stuff in place.
00;27;33;09 - 00;27;33;24
Unknown
Right.
00;27;33;24 - 00;27;34;14
Unknown
Yeah. What you
00;27;34;14 - 00;27;49;23
Unknown
think are are silly little metrics are actually kind of the nuts and bolts holding the whole system together. And I do think that's a concern in general, not just within martech, but in general. Like, if you haven't done the job, if you don't actually know what you're talking about, you shouldn't be leaving it.
00;27;49;23 - 00;27;52;24
Unknown
And that's at all levels.
00;27;52;26 - 00;27;53;23
Unknown
You need to at least have
00;27;53;23 - 00;28;03;13
Unknown
some familiarity enough to know, like OOP, I need to outsource because I know someone knows more about this, and I'm a huge proponent twofold of not building in-house. One,
00;28;03;13 - 00;28;11;17
Unknown
yes, it's very cool that you have it on your resume that you worked with an in-house platform, but that's not exactly the most transferable of skills versus working with
00;28;11;17 - 00;28;20;00
Unknown
a well known platform, whether it's braze or Iterable or engage, as you said, like those are marketable traits and skill sets,
00;28;20;00 - 00;28;23;04
Unknown
that recruiters and hiring managers can understand to
00;28;23;04 - 00;28;24;23
Unknown
translate the skill set.
00;28;24;23 - 00;28;45;05
Unknown
And also, why build something if you're not going to have a product owner and an actual ability to upgrade it, and so much that it's something you could even sell externally, if it's going to be a new business division that you can white label, go for it. But building an ESP is not at or ESP.
00;28;45;05 - 00;28;48;05
Unknown
Yep. Whatever you wanna call it is not as easy as people think.
00;28;48;08 - 00;28;57;03
Unknown
Yes, it's just a front end UI of a back end system, but there's a lot more that goes into it, and it's much more complex than just HTML and CSS.
00;29;01;17 - 00;29;04;12
Unknown
like? Being able to drop in a person
00;29;04;12 - 00;29;08;17
Unknown
that's able to understand that system, especially. You know, I guess that's probably part of the reason why
00;29;17;06 - 00;29;18;05
Unknown
But do they.
00;29;22;04 - 00;29;35;19
Unknown
I hear what you're saying. I push back only for the fact that folks who worked with the platform when it was exact target, as in, like when it was in its original state, still don't know the full platform and still don't understand the quirks. And nothing has changed in 20
00;29;35;19 - 00;29;37;00
Unknown
years. So
00;29;37;00 - 00;29;38;19
Unknown
that's where I struggle with.
00;29;38;22 - 00;29;47;24
Unknown
If you can't fully become an expert in a platform, it's because the platform is ungovernable and and not in a positive sense.
00;29;48;13 - 00;29;51;21
Unknown
that's more of a legacy issue, thankfully,
00;29;51;21 - 00;29;52;22
Unknown
anything else.
00;29;52;22 - 00;30;11;21
Unknown
Brought to you by our sponsors. If there's one thing that's followed me at every stop in my martech career, it's trying to get good data into the hands of marketers. That's why I'm so excited to tell you about our sponsor, High Touch, the leading composable CDP, an AI decisioning platform. Companies like Domino's, chime, Aritzia, and PetSmart trust high touch to power their data.
00;30;11;23 - 00;30;30;05
Unknown
And here's the kicker. 90% of customers have a real use case live in production within their first week. That means you can implement a world class CDP in months, rather than the usual years long headache. That's why top brands choose high touch to personalize every customer interaction at scale. See what high touch can do for you at high touch.com/miss.
00;30;30;07 - 00;30;31;28
Unknown
And now back to the hotseat.
00;30;31;28 - 00;30;35;29
Unknown
If your martech budget was cut by 50%,
00;30;35;29 - 00;30;37;01
Unknown
what would you keep?
00;30;37;01 - 00;30;44;21
Unknown
What would you kill or what would you rebuild in-house? And ultimately, what is the risk of reducing your tech budgets?
00;30;45;03 - 00;30;46;08
Unknown
And I feel like it's a tricky answer
00;30;46;08 - 00;30;51;22
Unknown
because you guys have a very lean and non redundant tech stack. From what I've heard.
00;31;01;05 - 00;31;01;20
Unknown
Yeah.
00;31;01;20 - 00;31;12;20
Unknown
let's call it 50%. Right. If I had to because the downstream effects of that 50% sending 50% less is right. Higher engagement on what we are sending.
00;31;12;28 - 00;31;14;24
Unknown
Yeah. Cheaper. All the things.
00;31;15;04 - 00;31;21;02
Unknown
Yeah. Like less load on the downstream systems. Right. Because like, probably email alone
00;31;21;02 - 00;31;22;16
Unknown
costs.
00;31;22;16 - 00;31;29;19
Unknown
I think the engineering team was almost like we account for 60% of the spend on amplitude.
00;31;29;19 - 00;31;31;22
Unknown
So there's a downstream there. Right. And then
00;31;31;22 - 00;31;32;23
Unknown
Yeah. Of course.
00;31;32;23 - 00;31;38;08
Unknown
another processing cost involved with, the data warehouse, the data, because
00;31;38;08 - 00;31;40;04
Unknown
Yes.
00;31;41;27 - 00;31;48;20
Unknown
just if we just have the, the volume and try to be, like much more vigilant about the, the like,
00;31;48;20 - 00;31;50;02
Unknown
list quality,
00;31;50;05 - 00;31;50;22
Unknown
Yep.
00;31;50;22 - 00;31;54;27
Unknown
then we almost wouldn't sacrifice anything.
00;31;54;27 - 00;31;56;12
Unknown
That's the right answer. Which
00;31;56;12 - 00;31;57;27
Unknown
I think very few,
00;31;57;27 - 00;32;14;07
Unknown
frankly, would recognize that as the right answer. Like, oh, okay, we can do suffer without X, Y, or Z. But no, actually you're right. It's it's actually proper hygiene. It's let's be super clear on who we want to talk to, how we want to talk to them. And as well it cleans up the whole process.
00;32;14;10 - 00;32;15;13
Unknown
And
00;32;15;13 - 00;32;25;28
Unknown
interestingly, I feel like most data bricks and snowflake instances are actually just completely comprised of marketing data, whether it's, in a lake and, you know, unstructured or not. Depends. But just if you think
00;32;25;28 - 00;32;35;06
Unknown
about it, if you're doing paid ads, if you're doing emails, especially if we're talking, if you've got hundreds of millions of customers like that is going to be
00;32;35;06 - 00;32;42;18
Unknown
some of the most costly and expensive components of that annual budget.
00;32;42;23 - 00;32;50;02
Unknown
Yeah. I mean, it's hard, right, to throw them away because, like, well, we worked really hard to get these people,
00;32;50;02 - 00;33;07;19
Unknown
Of course, this is a this is a scenario that isn't going to happen, but I think it's really helpful advice as folks are facing budget cuts as well as changes. It's like, okay, where can we cut without removing quality? And I think that is the most
00;33;07;19 - 00;33;17;19
Unknown
prudent approach as well. Like let's just be super clear on who our ICP is and who our customers who want to talk to us are like, do we need to keep dormant folks?
00;33;17;19 - 00;33;19;19
Unknown
What is our retention policy?
00;33;19;19 - 00;33;22;15
Unknown
Maybe it's really starting in that capacity to,
00;33;22;15 - 00;33;26;00
Unknown
But great answer. I really like that one.
00;33;26;00 - 00;33;31;24
Unknown
zooming out, if a CMO wants more revenue and that's the capability they're looking for in the next quarter to hit revenue
00;33;31;24 - 00;33;34;16
Unknown
targets, like who or what should they hire?
00;33;34;16 - 00;33;40;07
Unknown
Where should that person or item live like does I help in this or does it add noise.
00;33;40;07 - 00;33;42;11
Unknown
Kind of. How would you go about,
00;33;42;11 - 00;33;44;00
Unknown
advising a CMO here?
00;33;47;10 - 00;33;53;29
Unknown
What would Jeff do? A new version of Wwjd?
00;33;53;29 - 00;33;55;22
Unknown
revenue. Well.
00;34;14;23 - 00;34;23;08
Unknown
segment or I, you know, did that person, you know, maybe you don't usually have multiple KPIs or something like that, but like, you know, which.
00;34;23;08 - 00;34;31;10
Unknown
Like understanding from the like when they landed on the marketing and then hitting getting into the hitting the product to be able to,
00;34;31;10 - 00;34;33;04
Unknown
To address that
00;34;33;04 - 00;34;36;10
Unknown
like across the full funnel experience seems
00;34;37;09 - 00;34;38;18
Unknown
something useful.
00;34;38;22 - 00;34;43;06
Unknown
Yeah.
00;34;54;22 - 00;34;57;17
Unknown
Yeah.
00;34;57;17 - 00;35;01;11
Unknown
actually, you know, because of the privacy loss stuff, like, can you actually do it?
00;35;01;13 - 00;35;12;19
Unknown
But, you know, once they come in from a if they're coming off of a web form, you can definitely say like, hey, I know you came here because, you know, you wouldn't say creepy like that, but
00;35;13;14 - 00;35;20;18
Unknown
Well, it depends on who you ask. Yeah, maybe ask for advice. Don't be creepy. That is the goal.
00;35;21;13 - 00;35;24;26
Unknown
Just just based on, like, here's what you like.
00;35;24;28 - 00;35;33;24
Unknown
And a semblance of you came in here through this ad, you filled out this thing and you came into the app. So you're probably you want to do that?
00;35;33;24 - 00;35;35;08
Unknown
Let's carry it all through.
00;35;35;08 - 00;35;40;11
Unknown
Follow through on the promise. Right. That was made by the ad. And then,
00;35;40;11 - 00;35;44;10
Unknown
you know, if they are not responding to, to be able to switch off of it,
00;35;44;10 - 00;35;45;08
Unknown
is that
00;35;45;08 - 00;35;46;15
Unknown
kind of a weird answer.
00;35;46;15 - 00;35;47;08
Unknown
If I'm,
00;35;47;08 - 00;35;58;13
Unknown
Like, hearing the through line of what you're saying, it is personalize and retarget accordingly and suppress and re segment as quickly as possible
00;35;58;13 - 00;36;05;23
Unknown
to reduce the amount of time and or money spent on any given individual that is coming through the funnel.
00;36;23;25 - 00;36;28;04
Unknown
AI to be able to say, hey, this person is interested in this
00;36;28;04 - 00;36;34;02
Unknown
So, almost a more automated sorting hat. Or a more complexly automated one.
00;36;36;28 - 00;36;38;08
Unknown
Another thought,
00;36;38;08 - 00;36;44;27
Unknown
I ask these questions aloud also for myself. The other one, I think, is reduce, pay tax, spend,
00;36;44;27 - 00;36;54;23
Unknown
because that's the most expensive thing, but also increase, investment in retention, nurture and like surprise and delight campaigns.
00;36;54;23 - 00;36;59;22
Unknown
There's a lot of options.
00;36;59;22 - 00;37;00;25
Unknown
retention
00;37;00;25 - 00;37;01;18
Unknown
and.
00;37;01;18 - 00;37;05;20
Unknown
Yeah, that was a good amount because it's very relevant to like, that, that full funnel kind of view.
00;37;05;26 - 00;37;06;03
Unknown
Yeah,
00;37;06;03 - 00;37;16;03
Unknown
well, it blows my mind. They're like, yeah, we'll spend 80 million on paid. I'm like, okay, why don't you spend a third of that instead on resources internally
00;37;16;03 - 00;37;17;18
Unknown
that
00;37;17;18 - 00;37;21;18
Unknown
have a better lifetime value? But I digress.
00;37;21;18 - 00;37;25;09
Unknown
And it could be just like the company and the and maybe this is the problem is,
00;37;25;09 - 00;37;29;08
Unknown
like, we have the these multiple teams, right? You have the life cycle team and you have your team.
00;37;29;08 - 00;37;31;08
Unknown
Right. We're on the same org,
00;37;31;08 - 00;37;32;23
Unknown
same same boss.
00;37;32;23 - 00;37;33;20
Unknown
You UAE paddock
00;37;33;20 - 00;37;36;17
Unknown
And then I don't know if it's technically possible, but like,
00;37;36;17 - 00;37;39;28
Unknown
they're doing something that's very distinct from.
00;37;39;28 - 00;37;43;28
Unknown
They're quite siloed apart from us. Even though we're the same group.
00;37;43;28 - 00;37;44;17
Unknown
They're just
00;37;44;17 - 00;37;49;13
Unknown
like, they're just siloed just in terms of, like, the data, like, doesn't quite get through.
00;37;49;13 - 00;37;58;29
Unknown
So, I mean, maybe this was like, if I'm a CMO trying to probably try to maybe that's what it was like, merge those two together. Ish. Like to to work a little bit better
00;37;59;17 - 00;38;00;11
Unknown
Okay.
00;38;00;11 - 00;38;01;26
Unknown
together and
00;38;01;26 - 00;38;06;26
Unknown
then, use those to kind of fuel. What did you say? Retention.
00;38;07;10 - 00;38;08;11
Unknown
Yeah.
00;38;08;11 - 00;38;09;28
Unknown
surprise the like.
00;38;10;00 - 00;38;11;26
Unknown
I'm not sure how to phrase that, but, like.
00;38;11;26 - 00;38;14;20
Unknown
If you were to advise a CMO on
00;38;14;20 - 00;38;18;17
Unknown
how to increase revenue next quarter because they're receiving pressure from
00;38;18;17 - 00;38;27;10
Unknown
upward places like the board or the CEO and revenue targets like, what should that CMO do that you're able
00;38;27;10 - 00;38;28;22
Unknown
to do within a quarter?
00;38;28;22 - 00;38;39;09
Unknown
Because you probably can't build tech that quickly. But I'm curious if you had a quick answer on this is how we could do it with no new resources, what would you recommend?
00;39;11;22 - 00;39;22;20
Unknown
either in the product or on the lifecycle side, if, you know, if we can get that signal, I think I can do some of those things for it based on whatever signals they might be receiving from the ad, because I've heard
00;39;25;15 - 00;39;31;27
Unknown
it would be nice, like especially with the internal capabilities. And I find that
00;39;31;27 - 00;39;33;04
Unknown
matching
00;39;33;04 - 00;39;39;02
Unknown
like trying to match the message with the intent that they came on will improve metrics.
00;39;39;02 - 00;39;41;22
Unknown
So I'd probably be something like that. It would involve working
00;39;47;24 - 00;39;49;09
Unknown
That makes sense.
00;39;50;27 - 00;39;52;21
Unknown
hopefully move, move the needle there.
00;39;53;01 - 00;40;11;17
Unknown
Yeah. For sure. And I guess building off of that, like, what tips do you have for the non-technical marketers, the life cycle folks, the paid acquisition folks for wanting to work with marketing engineers more in terms of like, how can they level set and have more informed conversations together?
00;40;24;11 - 00;40;30;18
Unknown
And that's kind of like where I'm working. Right? The as the marketing engineer, I'm playing across all of those
00;40;30;18 - 00;40;32;27
Unknown
companies or silos,
00;40;32;27 - 00;40;33;23
Unknown
whatever you want to call them.
00;40;34;05 - 00;40;34;25
Unknown
For sure.
00;40;34;25 - 00;40;45;22
Unknown
You know, have a curiosity about, hey, what's the product doing? What is the UI team doing? Are they doing something that's working? Can we use that? Right? Can I repurpose that? Can I borrow from that?
00;40;45;22 - 00;40;51;24
Unknown
And you know, as a if you're non-technical, it's like at least thinking about those things. Some of that stuff is actually
00;40;51;24 - 00;40;53;22
Unknown
could be it might sound hard,
00;40;56;01 - 00;40;56;25
Unknown
Yeah.
00;40;56;25 - 00;41;06;27
Unknown
one field that has to carry through and then some, you know, on the other hand, some things that you think like might be really easy, like, oh, let's just change the color of a button might actually be really super hard. Right.
00;41;07;01 - 00;41;07;24
Unknown
Yeah.
00;41;07;24 - 00;41;10;24
Unknown
Dark mode. Man. The inverse.
00;41;10;24 - 00;41;13;29
Unknown
So like super like super kind of this was really just.
00;41;13;29 - 00;41;25;26
Unknown
I think, you know, getting out of like, oh really? I got a, I got to write the next email and I got to do the next thing. I was like, well, how is that email going to affect what's happening on the other, like on the next team downstream, or where
00;41;25;26 - 00;41;28;22
Unknown
did it come from? Upstream. Like you're trying to understand, like
00;41;28;22 - 00;41;31;16
Unknown
what is this ecosystem of like that's surrounding,
00;41;31;16 - 00;41;33;05
Unknown
you as the marketer.
00;41;33;05 - 00;41;34;01
Unknown
Yeah, I love that.
00;41;34;01 - 00;41;44;10
Unknown
that is a such a truth in that if everyone can understand, we're all working towards the same goals. It's all part of the same ecosystem. Everything is connected. Everything is systems.
00;41;44;10 - 00;41;47;10
Unknown
How do we design a system? And this is where I left
00;41;47;10 - 00;41;54;19
Unknown
systems thinking in particular. But how do we design a system that is both scalable, helpful and benefits everyone?
00;41;54;19 - 00;41;57;25
Unknown
Because truly we don't need different fields for different teams.
00;41;57;25 - 00;42;06;27
Unknown
We should have the same fields and be able to leverage them differently for multiple different types of use cases, depending on your business, your segments, you name it.
00;42;06;27 - 00;42;08;22
Unknown
There really shouldn't be a difference.
00;42;08;22 - 00;42;09;11
Unknown
And so
00;42;09;11 - 00;42;12;21
Unknown
I love that as a helpful tip. And so
00;42;12;21 - 00;42;14;25
Unknown
that's in terms of folks already
00;42;14;25 - 00;42;23;14
Unknown
inside an organization, if you're trying to hire for a marketing engineer, like what skills or instincts and cross influences are you looking for for that?
00;42;23;14 - 00;42;24;21
Unknown
The strongest candidate.
00;42;36;05 - 00;42;45;03
Unknown
something I did, like where we're talking about the budgets. Right. Something that happened here has downstream four levels away, something that happened a couple levels upstream.
00;42;45;03 - 00;42;46;08
Unknown
It might exist.
00;43;05;02 - 00;43;10;22
Unknown
Right. So I think the marketing engineering fits more into like this,
00;43;10;22 - 00;43;15;07
Unknown
like slow, like this fast or big part of the team. And then fast and
00;43;16;22 - 00;43;21;09
Unknown
just one team's moves at a different pace because they had to like think about
00;43;21;09 - 00;43;23;16
Unknown
different things. Right. So the slow org it might be engineering.
00;43;23;16 - 00;43;31;04
Unknown
They're they're worried about redundancies and load balancing and all this other stuff fast. Org is like like experiment and whatever. So I think marketing engineer is like
00;43;31;04 - 00;43;35;00
Unknown
that. That fast brain mindset is like, I'm just going to do a bunch of stuff, throw it away,
00;43;35;00 - 00;43;38;00
Unknown
and be ready just to be wrong. Like all the time.
00;43;38;00 - 00;43;38;12
Unknown
Yeah.
00;43;38;27 - 00;43;45;01
Unknown
to be able to, like approach it with, like, okay, even if I'm going to be wrong, maybe there's a little chunk of it I can use for the next thing
00;43;45;01 - 00;43;46;09
Unknown
to keep going fast.
00;43;46;09 - 00;43;50;02
Unknown
Yeah. That's super interesting. I'm curious what that podcast is because,
00;43;50;02 - 00;43;57;28
Unknown
I definitely have come across the, like, slow thinking versus quick thinking. And there's pros and cons to both. And actually I'm thinking of. So Malcolm Gladwell, one of
00;43;57;28 - 00;44;01;27
Unknown
my favorite podcasts, he does revisionist history and there's an episode on
00;44;01;27 - 00;44;02;27
Unknown
lawyers and Supreme
00;44;02;27 - 00;44;15;26
Unknown
Court justices. And so and taking the Lsat and those who go to the top tier, the Harvards, the Yale Law schools versus your affordable public school
00;44;15;26 - 00;44;24;00
Unknown
university, and the difference of who actually goes on the Supreme Court versus who is actually a top person, that certain law firms.
00;44;24;00 - 00;44;26;14
Unknown
And there's a direct correlation to
00;44;26;14 - 00;44;36;07
Unknown
those who get the highest scores on Lsat being very quick at responses versus those who get lower. However,
00;44;36;07 - 00;44;48;09
Unknown
in terms of how those think, it's like a rabbit versus hare concept, and that's what he uses. It is do you think quickly or do you think methodically, slowly and fully? And both are very important.
00;44;48;16 - 00;44;48;24
Unknown
And
00;44;48;24 - 00;45;07;26
Unknown
that's where the conversation is like, what do you want in a Supreme Court justice? Do you want someone who is a quick reaction, willing to see what happens versus what are the potential ramifications if other variables change? What are the downstream effects? And so I'm very curious what the podcast was you're referencing, because that was a
00;45;07;26 - 00;45;15;04
Unknown
big one for me.
00;45;15;10 - 00;45;16;25
Unknown
Yeah.
00;45;19;19 - 00;45;19;25
Unknown
I
00;45;19;25 - 00;45;21;10
Unknown
oh, cool. Okay.
00;45;25;25 - 00;45;27;00
Unknown
Yes, it has.
00;45;27;00 - 00;45;32;07
Unknown
over there has. I thought it was interesting. I was like, oh, that's exactly what we're kind of doing.
00;45;32;07 - 00;45;36;06
Unknown
And just he put it into the words. I was like, okay, now it makes sense.
00;45;36;19 - 00;45;37;13
Unknown
Yeah.
00;45;37;13 - 00;45;38;27
Unknown
remember what episode
00;45;38;27 - 00;45;39;25
Unknown
It's
00;45;39;25 - 00;45;50;24
Unknown
all good. All I think I missed that one because I love Lenny's podcast. I'll be sure to check it out. And also you need to check out your table because I think you'll be surprised it is Google Sheets, but super powered.
00;45;50;24 - 00;45;52;09
Unknown
And it's
00;45;52;09 - 00;45;54;02
Unknown
pretty darn cool. I have to say.
00;45;54;02 - 00;46;04;07
Unknown
From an org design perspective, because you're in this unique position of being a marketing engineer, you sit between marketing, product and data and probably a lot more in between.
00;46;04;07 - 00;46;06;01
Unknown
what teams should you actually be on? Do you
00;46;06;01 - 00;46;16;08
Unknown
think that would make the most sense? And also like, who should own the KPIs? Like, should you have the same KPIs as a marketer or product person or typical engineer?
00;46;56;29 - 00;47;14;10
Unknown
Understood. All right. We've kind of alluded to it in different capacities. But where do you see I work in really, truly run end to end and life cycle today. And where do you also see a marketing engineer needing to stay in the human loop, especially as we think through like
00;47;14;10 - 00;47;14;20
Unknown
audience
00;47;14;20 - 00;47;17;09
Unknown
selection, measurement, holdout, kill switches?
00;47;17;09 - 00;47;22;26
Unknown
Do you have any examples of where a black box selection failed you and you put a guardrail into place as a result?
00;47;41;28 - 00;47;43;13
Unknown
Of.
00;47;44;28 - 00;47;58;27
Unknown
That one's a hard one. That's actually always a question I ask. When I'm hiring someone. I explicitly ask, what's your biggest fuckup? One to know if we can speak the same language and be honest with each other, but also if they have the humility and awareness to recognize we all make
00;47;58;27 - 00;48;04;23
Unknown
mistakes and.
00;48;04;29 - 00;48;06;14
Unknown
Always bad day.
00;48;26;06 - 00;48;26;14
Unknown
Yes,
00;48;26;14 - 00;48;27;28
Unknown
sir.
00;48;27;28 - 00;48;30;13
Unknown
for the AI to realize that it didn't work right.
00;48;30;13 - 00;48;30;21
Unknown
Like,
00;48;30;21 - 00;48;38;06
Unknown
The whoops emails always outperform. And that should never be your strategy. But yes they,
00;48;38;18 - 00;48;40;03
Unknown
they always outperform. And it's very
00;48;40;03 - 00;48;40;24
Unknown
embarrassing.
00;48;41;02 - 00;48;43;13
Unknown
that one was a bad one. And it's just
00;48;43;13 - 00;48;44;25
Unknown
I could see that that was where
00;48;47;15 - 00;48;48;06
Unknown
Yeah.
00;48;48;06 - 00;48;58;14
Unknown
That's where I really strongly believe in a glass box solution when it comes to I want to know the reason, the logic and the entire thought process behind every decision.
00;48;58;14 - 00;48;59;03
Unknown
Yeah.
00;48;59;03 - 00;48;59;18
Unknown
That's where
00;48;59;18 - 00;49;14;08
Unknown
I, at least for myself, I am hesitant to not always to have a human in the loop. Like we're just, I don't know if we actually ever will be in a place where we can recognize an anomaly like that, because anomalies will always happen.
00;49;14;08 - 00;49;17;05
Unknown
And that's and daily life. And that's in marketing.
00;49;18;00 - 00;49;22;11
Unknown
I mean, it gets math wrong all the time to like two plus two was five was like well how can I trust that. Like.
00;49;22;18 - 00;49;28;29
Unknown
That's true. You know, like I'm bad at math, but Excel doesn't do me dirty.
00;49;28;29 - 00;49;29;16
Unknown
Awesome.
00;49;29;16 - 00;49;46;16
Unknown
I guess to, to really bookend our conversation. I know martech engineers, I know marketing engineers, but it's still usually not it's own dedicated discipline. So I want to hear more from you about how do we make martech or marketing engineers a
00;49;46;16 - 00;49;55;10
Unknown
thing. And I you're the most important, backbones to particularly large programs.
00;49;55;10 - 00;49;58;09
Unknown
Like, it could be like a an awareness thing,
00;49;58;14 - 00;50;05;09
Unknown
Yeah. Well, good thing you're on this podcast and you've been on Phil's podcast. It's a
00;50;05;09 - 00;50;08;22
Unknown
good thing.
00;50;08;25 - 00;50;11;09
Unknown
Yeah.
00;50;11;09 - 00;50;15;19
Unknown
GTM and that seems like maybe it's like a
00;50;15;19 - 00;50;17;25
Unknown
should there be a definition of
00;50;17;25 - 00;50;26;10
Unknown
GTM engineer means it's B2B set, like a B2B sales kind of thing, whereas a marketing engineer is more consumer.
00;50;26;10 - 00;50;27;01
Unknown
Right.
00;50;27;01 - 00;50;27;27
Unknown
Also, like, what's different?
00;50;27;27 - 00;50;31;22
Unknown
Trying to go to market engineer and what was previously a Salesforce developer?
00;50;31;28 - 00;50;38;26
Unknown
I don't know, I think to your point, actually, it is a we need formal definitions of what these roles are as they evolve.
00;50;38;26 - 00;50;42;08
Unknown
And I mean, this is across marketing, and I think it's really more of a
00;50;42;08 - 00;50;50;06
Unknown
marketing problem than anything else is we're really bad at marketing ourselves as an we don't know how to define who does what because every company does it differently.
00;50;50;09 - 00;50;52;22
Unknown
And as a result, it impairs the
00;50;52;22 - 00;51;11;24
Unknown
rest of us and all of our partners because they don't know what you do. And I mean, the amount of times I say what I do and people like, oh, you're such a great marketer because I do have a marketing background, but it's not what I do anymore. Like I am in charge of the piping, I'm in charge of the data flow and that is a distinctly different discipline.
00;51;11;27 - 00;51;22;08
Unknown
But it it's almost like we've done a disservice to ourselves in this department and organization that we can hopefully right those wrongs and bring it to the forefront.
00;51;22;08 - 00;51;27;15
Unknown
Yeah. And it could be maybe engineering. Not the right word. Marketing something. But,
00;51;27;28 - 00;51;34;26
Unknown
I mean, if you're actually an engineer, I think you should be an engineer. But and
00;51;34;26 - 00;51;36;24
Unknown
that's where I think engineering on its own
00;51;36;24 - 00;51;45;23
Unknown
side of things really has things clarified. And it's actually also a really interesting distinction, I have noticed in the martech space where martech
00;51;45;23 - 00;51;51;05
Unknown
means something different in Silicon Valley than it does everywhere else in the world, everywhere else in the world.
00;51;51;05 - 00;51;52;10
Unknown
It is not an engineer,
00;51;52;10 - 00;51;53;09
Unknown
but in
00;51;53;09 - 00;52;05;26
Unknown
San Francisco. And this is obviously a generalization, but those are martech engineers specifically. And I'm curious if that's something you've noticed as well, because there's head of martech across the globe, but not so much in the US.
00;52;05;26 - 00;52;10;24
Unknown
It's getting more prevalent, getting more understood, because it really is the, in my eyes,
00;52;10;24 - 00;52;15;26
Unknown
highest cost reduction team, the most performant team.
00;52;15;26 - 00;52;20;26
Unknown
But also it's kind of like it in that it's in the back of the office and not the most,
00;52;20;26 - 00;52;25;11
Unknown
understood or celebrated except for when things go wrong and blamed.
00;52;30;00 - 00;52;30;26
Unknown
missing.
00;52;30;26 - 00;52;33;06
Unknown
like, it does require a certain budget, right, in order
00;52;33;09 - 00;52;35;23
Unknown
Yeah.
00;52;38;12 - 00;52;39;20
Unknown
It's not cheap.
00;52;39;20 - 00;52;45;10
Unknown
A year yeah.
00;52;45;10 - 00;52;59;01
Unknown
Well, and also, it's it's a double edged sword of growth. Like, if you're growing and you have a large audience, it's a privilege and a curse. You're automatically have to pay more. And you're likely going to both get upcharge if you don't negotiate very well and
00;52;59;01 - 00;53;00;18
Unknown
don't know how to. So there's a
00;53;00;18 - 00;53;08;08
Unknown
lot of, it's interesting you think it'll get easier, but actually sometimes it's a lot harder because mistakes cost more and mean more.
00;53;09;28 - 00;53;13;16
Unknown
let's push it as a real thing. That's what I'm trying to do here. It's just
00;53;13;19 - 00;53;15;01
Unknown
Hell, yeah. Well.
00;53;15;01 - 00;53;16;26
Unknown
people that want to do the same thing.
00;53;16;26 - 00;53;20;17
Unknown
just find that common taxonomy and go
00;53;20;17 - 00;53;25;17
Unknown
Yes, I am a taxonomy nerd, so yes, I am on the same page.
00;53;26;08 - 00;53;26;12
Unknown
Okay.
00;53;26;12 - 00;53;27;09
Unknown
It closes out.
00;53;27;09 - 00;53;29;04
Unknown
I'm thinking through a couple of different
00;53;29;04 - 00;53;48;17
Unknown
takeaways, one of which is like ensuring you audit where your data lives so that you know who's owning what. So really like thinking from a risky or racy standpoint so that, you know, one engineer owns versus a marketer versus a designer versus a this it's really helpful just to kind of have those
00;53;48;17 - 00;53;49;22
Unknown
designations.
00;53;49;22 - 00;54;03;13
Unknown
But similarly stable eyes before you scale like one, it's costly if you don't. But if you have those standardized schemas and if you have the rigor and hygiene built in long run, you're saving money.
00;54;03;13 - 00;54;04;10
Unknown
You're reducing
00;54;04;10 - 00;54;06;27
Unknown
headache and mistakes.
00;54;06;27 - 00;54;14;25
Unknown
also it just means you're able to input and bring on more folks. If you're changing platforms every couple of years, it makes things really hard.
00;54;14;25 - 00;54;21;15
Unknown
And, it doesn't mean you shouldn't migrate or upgrade and evolve if it's not necessary. But,
00;54;21;15 - 00;54;25;11
Unknown
makes things more complex and complicated, and then
00;54;25;17 - 00;54;26;07
Unknown
similarly,
00;54;26;07 - 00;54;33;10
Unknown
I think marketing needs to have engineers. And engineers also need to engineer their marketing for themselves too. So it's like
00;54;33;10 - 00;54;39;16
Unknown
modularized things, making things scalable, testing, observability, alerting and
00;54;39;16 - 00;54;40;20
Unknown
building for reuse.
00;54;40;20 - 00;55;05;17
Unknown
And that's for both platforms and also marketing each other, making it make sense internally and externally. And I think that's also the biggest takeaway is how do we make this real? Like this is now the second conversation I've had on this podcast previously with Neha Sikh about what is a marketing engineer or martech engineer. And I think there are not enough folks that have been given the voice to
00;55;05;17 - 00;55;06;28
Unknown
explain what it is.
00;55;06;28 - 00;55;07;14
Unknown
And
00;55;07;14 - 00;55;12;03
Unknown
would love to be challenged on this or not, but in my experience, oftentimes the martech engineers
00;55;12;03 - 00;55;13;16
Unknown
are the
00;55;13;16 - 00;55;18;16
Unknown
cast asides, as in they get assigned this and they try to get out of it as quickly as possible because they want to build
00;55;18;16 - 00;55;29;13
Unknown
something else versus, you know, they have a deep passion for. This is specifically the section, the type and component of a work.
00;55;29;13 - 00;55;36;10
Unknown
And, philosophy of work I want to do. I'm curious if you have a push back on that or not, because that's just something I've anecdotally noticed.
00;55;36;10 - 00;55;40;25
Unknown
Okay. So actually, that's the great point. And I was literally just talking to my friend about this over
00;55;41;01 - 00;55;48;18
Unknown
Yeah, I'm glad I'm not crazy.
00;55;48;20 - 00;55;50;02
Unknown
Oh, nice.
00;55;54;26 - 00;55;57;09
Unknown
Which is, excuse my French
00;55;59;08 - 00;56;01;11
Unknown
But like, on the other hand,
00;56;01;11 - 00;56;10;13
Unknown
The way we're like, we it's like it kind of pans out or works out where it's like there's. You end up with four platforms, right? There's
00;56;10;13 - 00;56;11;28
Unknown
iOS,
00;56;24;10 - 00;56;25;13
Unknown
Yeah.
00;56;25;13 - 00;56;42;19
Unknown
email there. It's still customer facing. It actually can drive more because of the retention is out, it goes outbound and it's bringing people back in. It's just like that's the fourth platform that's, you know, just because email is kind of old and clunky
00;56;47;12 - 00;56;52;19
Unknown
emails are driving multiple times more traffic
00;56;52;23 - 00;56;54;08
Unknown
Oh, yeah.
00;56;54;08 - 00;57;01;17
Unknown
Right. Or more revenue direct because I've just you come in an email, I'm landing you directly on whatever it was I wanted you to do
00;57;01;21 - 00;57;02;25
Unknown
Yeah, it's
00;57;02;25 - 00;57;06;09
Unknown
for sure. It's far more targeted than just general web.
00;57;06;09 - 00;57;11;24
Unknown
Yeah. So it's just it's that extra platform that's just kind of.
00;57;11;24 - 00;57;14;17
Unknown
the castaways. But it's probably, you know,
00;57;14;17 - 00;57;14;29
Unknown
00;57;14;29 - 00;57;16;29
Unknown
more performant. All right.
00;57;16;29 - 00;57;17;08
Unknown
Yeah,
00;57;17;18 - 00;57;27;17
Unknown
I want to no longer let marketing engineers and just those who work on the marketing side from a technical perspective, be castaways, because it's
00;57;27;17 - 00;57;29;02
Unknown
just as if not
00;57;29;02 - 00;57;36;10
Unknown
one of the most important roles, especially if we're thinking about scale. You can't really do it without it, or at least not properly.
00;57;36;10 - 00;57;41;23
Unknown
And last question that we ask everyone who is someone we should have on the podcast
00;57;45;11 - 00;57;51;26
Unknown
Yes. I do not see you. I've met her, but I would love an email introduction.
00;57;53;07 - 00;57;53;27
Unknown
She's a legend.
00;57;53;27 - 00;57;58;27
Unknown
She's like a legend, but then she decided to go off in the mountains and snowboard, like, most of the times.
00;57;59;00 - 00;58;03;28
Unknown
What? Well, I also know she also has a previous life as a musician. She's got a very
00;58;03;28 - 00;58;05;28
Unknown
fascinating, life
00;58;05;28 - 00;58;07;13
Unknown
story.
00;58;07;13 - 00;58;12;28
Unknown
she's actually kind of that forward and think I was like, hey, I need a marketing engineer. So she's out of this two years ago.
00;58;13;06 - 00;58;14;11
Unknown
I love it.
00;58;14;11 - 00;58;17;01
Unknown
came in. So it's probably has some perspectives,
00;58;17;01 - 00;58;17;19
Unknown
Yeah,
00;58;17;19 - 00;58;18;16
Unknown
beyond.
00;58;18;16 - 00;58;19;03
Unknown
you need
00;58;19;03 - 00;58;33;02
Unknown
to bring her out of retirement and snowboarding for maybe an hour. I love that. Well, Jeff, thank you so much for being here. Where can folks find you or get to know you? You name it.
00;58;33;02 - 00;58;41;25
Unknown
Yeah. I'm on LinkedIn. Don't post a ton, but I post a little bit more, especially just for fun these days.
00;58;41;28 - 00;58;42;02
Unknown
It
00;58;42;02 - 00;58;45;02
Unknown
can be entertaining. I also that.
00;58;45;02 - 00;58;51;11
Unknown
interesting to see the conversation that have in there. That's probably the best places. Reach out on LinkedIn and, you know, have
00;58;51;17 - 00;58;52;01
Unknown
Perfect.
00;58;52;01 - 00;58;53;29
Unknown
Awesome. Well, thank you so much.
00;58;53;29 - 00;58;55;20
Unknown
Yeah. Thank you. She's great.
00;58;55;20 - 00;59;06;25
Unknown
I.