Artwork for podcast RevOps FM
Is Marketing Mix Modelling the future of B2B analytics? - Mark Stouse
Episode 2419th March 2024 • RevOps FM • Justin Norris
00:00:00 00:47:44

Share Episode

Shownotes

Every marketing team  wants attribution. But weirdly, it's often not that satisfying when they actually get it.

I led many multi-touch attribution projects as a consultant, and we got really good at implementing tools, creating taxonomies, and making sure that data was clean.

But I found that when you actually showed these reports to a C-level executive, it was usually kind of underwhelming. The data didn't always pass the common sense test.

Today's guest thinks there's a better way — Marketing Mix Modelling. It's basically the application of mathematical techniques to model relationships between different variables.

However, technology now enables it to happen faster and more cost-effectively than ever before.

Thanks to Our Sponsor

Many thanks to the sponsor of this episode - Knak.

If you don't know them (you should), Knak is an amazing email and landing page builder that integrates directly with your marketing automation platform.

You set the brand guidelines and then give your users a building experience that’s slick, modern and beautiful. When they’re done, everything goes to your MAP at the push of a button.

What's more, it supports global teams, approval workflows, and it’s got your integrations. Click the link below to get a special offer just for my listeners.

Try Knak

About Today's Guest

Mark Stouse is CEO of ProofAnalytics.AI. With over 26 years of experience in marketing communications and strategy, he has a passion for transforming GTM performance with data-driven insights and agile decision making. Prior to founding Proof, Mark was CMO at Honeywell Aerospace, CCO at BMC Software, and a marketing leader at Hewlett Packard Enterprise.

https://www.linkedin.com/in/markstouse/

Key Topics

  • [00:00] - Introduction
  • [01:15] - Clarifying the acronym “MMM”
  • [02:39] - Mark’s background and how he founded Proof Analytics
  • [07:57] - Limitations of multi-touch attribution (“MTA”)
  • [14:16] - How MMM avoids the shortcomings of MTA
  • [16:42] - The Fischer Price definition of MMM
  • [19:56] - Demand vs. brand investments and their impact
  • [24:09] - A/B vs. multivariate regression
  • [25:21] - MMM is aggregate modelling, no reliance on PII
  • [27:12] - Simple explanation of multi-variate regression
  • [30:29] - Incorporating third-party data sources
  • [31:48] - Historical ROI vs. forecasted ROI
  • [32:52] - Is MMM just for enterprise?
  • [34:51] - Marketing as a non-linear multiplier
  • [38:02] - Getting started with MMM
  • [41:18] - Updating models to include new data sources
  • [42:07] - Competition in the marketing analytics space
  • [44:41] - B2C marketing is more advanced in usage of multi-variate regression

Resource Links

Learn More

Visit the RevOps FM Substack for our weekly newsletter:

Newsletter

Transcripts

Justin Norris:

Every marketing team wants attribution.

2

:

But the weird thing is that it's often not

that satisfying when they actually get it.

3

:

I did a lot of multi-touch attribution

projects as a consultant, and we got

4

:

really good at implementing these

tools, technically creating taxonomies,

5

:

making sure that data was clean.

6

:

But I found that when you actually show

these reports to a c-level executive,

7

:

it can be kind of underwhelming.

8

:

They don't always pass

the common sense test.

9

:

People want to pick on details, and maybe

that's because the idea of dividing up an

10

:

opportunity like a pizza between different

touch points isn't actually the way.

11

:

Maybe that's not going to

give us the answers we need,

12

:

but where does that leave us?

13

:

Because we still need to know what's

working in marketing, and today's guest

14

:

thinks there's a better way, and he's

founded a company to make marketing mix

15

:

modeling or MMM available to B2B marketers

and he's gonna tell us all about it.

16

:

So I'm super excited to

welcome Mark Stouse, CEO of

17

:

proof analytics to the show.

18

:

Thanks a lot for being here, mark

19

:

Mark Stouse: Hey, thank you so much.

20

:

Justin Norris: Mark, maybe

one clarifying question.

21

:

I've seen MMM, spelled out as media mix

modeling and marketing mix modeling, which

22

:

is the right way from your point of view?

23

:

Mark Stouse: actually.

24

:

It really represents the evolution of

it over the last say, 40 to 45 years.

25

:

Back when Procter and Gamble first brought

out what was then called econometric, I.

26

:

Analysis it was advertising.

27

:

that's what it was really all about,

hence the media mix modeling reference.

28

:

As, time moved on channels proliferated,

it became still very much within

29

:

kind of B2C a reference point.

30

:

It became marketing mix modeling.

31

:

Today it is really go to market mix

modeling because it includes not

32

:

just marketing data and channels and

investments and all that kinda stuff, but.

33

:

Sales and customer success and product

data outside data externalities, you

34

:

know, the economy, your competitor

actions, really is today much

35

:

broader canvas as it should be.

36

:

Right?

37

:

That's the long and the short of why

people say that MMM means media mix

38

:

modeling or marketing mix modeling.

39

:

And actually both of 'em are

kind of A little bit passe today

40

:

Justin Norris: Maybe let's take a

step back before that and talk about

41

:

what was your professional experience

leading up to founding proof analytics

42

:

and what brought you to this direction?

43

:

I.

44

:

Mark Stouse: So I was actually pretty

much a classic marketer and communicator.

45

:

I've worked across all of the different

subsets of marketing at one time or

46

:

another, and I've been a large company,

CMO, . About a little less than 20

47

:

years ago I was at HP and we were all

kind of in the middle of an existential

48

:

crisis because the then CEO of hp,

mark Hurd was a very Operations focused

49

:

and a very customer focused CEO.

50

:

And he wanted to know why there wasn't

more evidence of what marketing was

51

:

actually delivering to the company.

52

:

It was actually incredibly unpleasant.

53

:

And about the only good thing that I can

say about that whole experience that I had

54

:

and that other peers of mine had was that

it was, was highly motivating to change.

55

:

I kind of got to a point where I

said to myself, look, I either have

56

:

to do something to fix this or I

just need to like go do something

57

:

else, 'cause it's not just about

budget issues and all that kinda stuff.

58

:

It's about credibility,

am I actually doing here?

59

:

So in my particular case for whatever

reason, herd gave me a project.

60

:

To actually see what could

be done to resolve this.

61

:

And I didn't know Jack at

that time about analytics.

62

:

I was not even a math enthusiast,

and and so I, I remember I went

63

:

home that next Friday I went

through all the stuff in my garage.

64

:

. And found a couple of old math

textbooks from university and started

65

:

to read them in ways that I never

read them when I was in school.

66

:

' cause I was actively seeking an answer

and all of a sudden I got to this and it

67

:

was all about multi-variable regression.

68

:

Which is absolutely the cornerstone of

causal analytics ? It is the cornerstone

69

:

of the scientific method of inquiry.

70

:

It has so much credibility that,

it'd be hard to have more credibility

71

:

than multi-variable regression has.

72

:

Not because it is perfect, but it

is absolutely the best that we have.

73

:

And today, is actually

the bedrock of causal ai.

74

:

So I started working on this project

for Herd, and he liked it a lot.

75

:

And I guess my prize was he

set up a mentoring session.

76

:

With the CFO of hp, Bob Leman.

77

:

The whole idea was, I want you guys to

talk you from your end and he from his end

78

:

and see if we can't get to a meeting of

the mines between marketing and finance.

79

:

This whole experience it just changed me.

80

:

And before he passed away, I had

the opportunity to talk to Mark

81

:

again and thank him for it, right.

82

:

Because it was, not fun at all,

but he did me the hugest favor.

83

:

So I started climbing the

analytics ladder, right?

84

:

The stairway to heaven, so to speak.

85

:

Right?

86

:

By the time I was CMO at Honeywell

aerospace, pre any kind of

87

:

automation for analytics, right?

88

:

meant we had to hire a ton of people

in order to get the latency on the

89

:

recalculation down, the scalability up

the cost be damned at this point, right?

90

:

'cause it was so important.

91

:

And get the understandability of

the outputs to the point where the

92

:

business goes, yeah, I not only get

it and believe it, but I can make a

93

:

better decision today than I could

make before as a result of this.

94

:

And then you kind of say, there's not very

many companies that are willing to invest

95

:

seven, eight, $9 million a year just in

marketing analytics particularly in B2B.

96

:

That point you start to realize

that, that automation was gonna

97

:

be absolutely indispensable.

98

:

To solving the underlying root problems,

if Data science has has a number

99

:

of major problems in the business

context, but they can kind of be summed

100

:

up by the fact that it's too slow.

101

:

So by the time they get you the

insights, everything has moved on,

102

:

you've already had to make the decision.

103

:

The information is not comprehensible

many times by normal people.

104

:

it's not scalable.

105

:

So you end up with three or four

mega models that get updated

106

:

once, maybe twice a year.

107

:

And so not agile, right?

108

:

But big opportunity to appropriately

automate and bring ai into that

109

:

whole thing and really elevate

all of this to a new level of

110

:

accessibility and meaningfulness.

111

:

And so that's why we built proof.

112

:

Justin Norris: And so I want to

contrast MMM and regression based

113

:

analytics that you're describing with.

114

:

What most B2B marketers are

probably familiar with, which is

115

:

multi-touch attribution, that's what

pretty much everybody does today.

116

:

What is the problem with this form of

attribution from your point of view?

117

:

Or is there a problem?

118

:

Is it still okay?

119

:

Can they be complimentary?

120

:

Mark Stouse: as long as it's accurate

data, there's absolutely nothing wrong

121

:

with having a really good understanding

of the customer journey, patterns in the

122

:

customer journey, things like that, right?

123

:

And indeed, a lot of our customers

will include customer journey

124

:

data in regression models.

125

:

It's valuable.

126

:

The problem is that it is not fit for

purpose for what it's being sold to do.

127

:

We can just start with the idea of bias.

128

:

There's a lot of bias in that data.

129

:

There's a lot of bias in

the weightings of that data.

130

:

So marketing teams set their

own weightings in advance.

131

:

It is antithetical to the mathematical

principles, to say that you can

132

:

use MTA data to optimize anything.

133

:

Particularly spend

134

:

Justin Norris: I just want to

unpack that to understand why

135

:

Mark Stouse: mTA data is an effect.

136

:

You're measuring an effect.

137

:

not capturing.

138

:

The cause of that.

139

:

And more specifically,

the time lagged cause.

140

:

So how are you going to optimize the

money that you spent the past at this

141

:

point, based on what you think you're

seeing from effects in the present when

142

:

you have no idea what the time lag is?

143

:

Right?

144

:

So there is this assumption,

that the time lag is near zero,

145

:

but that's just so not the case.

146

:

Mm-hmm.

147

:

I B2B, the time lags can be extensive.

148

:

What do I mean by extensive?

149

:

Two, three quarters.

150

:

we get into brand investments, it's

double that, exactly are you optimizing

151

:

and how do you know which part of your

budget back in time you needed to optimize

152

:

based on this effect in the present?

153

:

Right?

154

:

Justin Norris: If we take an MTA

scenario, just to make it like a

155

:

little bit more clear, let's say

we have an opportunity and we have

156

:

three people in the buying committee

that are part of that opportunity.

157

:

And between those three people,

there were sacred brown numbers, 30

158

:

different touch points that they had

with different marketing initiatives.

159

:

And those could be like, they

downloaded eBooks, they attended

160

:

webinars, they clicked on ads.

161

:

And so the, traditional way is like,

we're gonna take those 30 touchpoints

162

:

and we're gonna take that opportunity.

163

:

Let's say it's a 3 million opportunity,

and we're just gonna divvy it up.

164

:

And the way that we

divvy it up could vary.

165

:

We may pick some touch points, is being

more influential or not influential?

166

:

And I totally agree with you there,

that it's kind of like arbitrary that

167

:

they give you different models that

you can choose from and it's like, pick

168

:

the one you want, like on what basis?

169

:

Mark Stouse: I don't think

that marketers understand how

170

:

transparently untrue the premise of

MTA is to everyone else but them.

171

:

Justin Norris: I wanna get to the

reality of the situation because this

172

:

is the, the industry that we're in.

173

:

Mark Stouse: The reason why it

went through adoption is that most.

174

:

Marketers don't have enough knowledge

of math to be able to recognize

175

:

when something doesn't work.

176

:

And number two, it was presented to

them as a way of mining data that they

177

:

already were generating in their stacks.

178

:

So it was easy in quotes, right.

179

:

To do this.

180

:

And I've been around it long enough

and talked to enough marketers about

181

:

it, everybody was shocked when they

went into their first meeting with

182

:

MTA data, Thinking that they were

just gonna be greeted as conquerors

183

:

and victors and to your earlier

point at the top of the show, right.

184

:

It was anything but that.

185

:

it was very innervating

for a lot of marketers.

186

:

You can actually do it in

Excel if you only have to do

187

:

it on a very limited basis.

188

:

But if B2B would just adopt B2C analytics

and also I would just add market research

189

:

priorities they would see a very rapid

transformation of their credibility and

190

:

their precision and everything else.

191

:

And.

192

:

can always modify it as needed,

but the core principle is ready to

193

:

go and has been a very long time.

194

:

Justin Norris: So if I recap your

perspective as I understand it, touch

195

:

points, the customer journey data,

what the person did and when that can

196

:

be interesting, that can be valuable.

197

:

The notion of then doling out credit

to those touch points for revenue

198

:

outcomes, not mathematically,

statistically defensible.

199

:

Mark Stouse: would just add this, right?

200

:

MTA at the end of the day

is about pattern matching.

201

:

It's about identifying large

scale, repeating patterns

202

:

in the customer journey.

203

:

And then the assumption is that if a

bunch of people are doing the same thing

204

:

over and over again, it must be causal.

205

:

That is a complete fallacy, the

same math is used for example,

206

:

and in fact, pattern matching too

is used to study climate change.

207

:

Some are causal, some are

absolutely not causal.

208

:

They happen all the time,

but they don't mean anything.

209

:

You can't just assume that

because, a thousand people do it.

210

:

Okay.

211

:

One of your channels, one of your feedback

loops, you know, it just keeps happening

212

:

that ipso facto means that it's causal.

213

:

Justin Norris: I've always had this

unease as well where you a great

214

:

example from earlier in my career,

we had a self-serve SaaS app, what

215

:

would be called product led today.

216

:

And everybody that signed up for

that app got a welcome email.

217

:

So from that point of view, if you're

looking at touchpoints, like the

218

:

welcome email is very influential.

219

:

It's very important but,

but everybody got it.

220

:

So just because people got it and

some people opened it doesn't mean you

221

:

could necessarily infer that that was

driving the outcome in a particular way,

222

:

. So everything we've been discussing

is kinda like the current state.

223

:

Let's contrast it now with MMM and

why does it not have these different

224

:

shortcomings that we've described?

225

:

Mark Stouse: Number one, let's start with

the shortcomings that it does have, right?

226

:

It is very dependent, as is

every analytic of any kind.

227

:

Very dependent on data quality, There's

that old saying about gigo, right?

228

:

Garbage in, garbage out.

229

:

Everyone is going to be

victimized by gigo equally.

230

:

Beyond that though.

231

:

It has a lot of major advantages.

232

:

It is a lean data mathematical process.

233

:

So you do not have to have big data

or even a lot of lean data order

234

:

to run these models accurately.

235

:

And that is actually a huge factor.

236

:

One of the things that CDOs chief

digital officers or data officers are

237

:

grappling with right now on private

LLMs is that they big data stocks.

238

:

They don't have enough training

data and they don't have enough

239

:

operational data to run private LLMs.

240

:

Everyone's been so understandably

fascinated with what they can do

241

:

with public LLMs that they haven't

really thought it through in terms

242

:

of the limitations on private stuff.

243

:

Right?

244

:

regression does not have this problem.

245

:

If you say, oh, I don't have enough

data to run regression analytics.

246

:

not even remotely able

to do any kind of ai.

247

:

And I think that a lot of c-suites

now Are starting to say, your

248

:

creation and maintenance of a high

quality data pool that's relevant to

249

:

your function is a core competency.

250

:

And if you tell us that you don't have

the right data, you don't trust your

251

:

data that's on you, The other thing is

you have to have some basic capability

252

:

on the human side of the equation.

253

:

So with proof, for example, we have

simplified the whole thing, dramatically.

254

:

Automated it significantly

in the right places.

255

:

you don't need a full-blown

data scientist to run it.

256

:

You need a competent data

analyst and you really don't

257

:

even need that person full time.

258

:

That's probably a half FTE, right?

259

:

Justin Norris: Could we give a

Fisher Price version definition of

260

:

what MMM is like, just so people

can conceptualize it in their minds.

261

:

Mark Stouse: we live in a multi-variable

world there's tons of potential

262

:

causes everything that you control

and everything that you don't control.

263

:

That's kind of a really

super easy way to bucket it.

264

:

And these all have interactions with

each other across time and space

265

:

that produce particular outcomes.

266

:

Causally speaking this is a

probabilistic calculation.

267

:

So outside of certain physical laws like

gravity, There is no deterministic outcome

268

:

that's possible to determine, right?

269

:

You'd have to know everything

that there is to know about what

270

:

contributes to a particular outcome

to get to a deterministic answer.

271

:

That's just not the way that

operates, particularly when you're

272

:

talking about human behavior.

273

:

So this is the same math that's

being used routinely to study climate

274

:

change, to study epidemiology.

275

:

To study economics, it captures time lag.

276

:

So in the end, the report will tell

you historically the stack rank

277

:

of everything that you control and

don't control, that's in the model.

278

:

It's relative effect on this outcome.

279

:

It will then forecast all that

into the future so that you can

280

:

then make different choices.

281

:

And because of automation,

we've sped that up to the point

282

:

where you can run multi-variable

regression exactly like GPS.

283

:

say that you're recalculating the model

every week, new data comes in and is

284

:

presented to the model automatically,

it automatically recalculates.

285

:

And you see how the present now

is comparing with the forecast

286

:

that you have for the same period.

287

:

if there's a growing delta, right?

288

:

You see, hey, okay, I

need to make some changes.

289

:

Stick with the GPS analogy.

290

:

need to reroute.

291

:

Or man, this is great.

292

:

It's just totally tracking.

293

:

And it will tell you how long, longer

it's going to take for you to reach your

294

:

objective, your goal, your destination.

295

:

So there's a countdown.

296

:

Regression is a huge part of

the actual GPS that's on your

297

:

phone, you use every day.

298

:

And if you stop and think about

it for a second, most business

299

:

questions, and indeed a lot of life

questions are navigation questions.

300

:

Where am I?

301

:

Where do I want to go?

302

:

What's the best way to get there?

303

:

I have enough time and resources

to achieve my destination?

304

:

Am I gonna run outta

gas before I get there?

305

:

I gonna run out of time

before I get there?

306

:

Am if I have to be there at

nine o'clock for a meeting?

307

:

And the GPS says, you're

not getting there until 11.

308

:

gonna have to make some

choices, that's part of it.

309

:

also captures all the headwinds

and tailwinds that may be speeding

310

:

you up or slowing you down.

311

:

So in many ways, it it gives you

the answer to your questions.

312

:

Justin Norris: Take a practical example

along the lines of one of those questions,

313

:

let's say a company spends a hundred

thousand dollars a month on display

314

:

ads as an act of faith because those

can be notoriously difficult to track.

315

:

They're not always resulting

in a click, but they could have

316

:

impressions that could have an impact.

317

:

So they're spending that a hundred

K month over month, and the CFO

318

:

challenges the CMO and says, why.

319

:

Are we spending this 100 K every month?

320

:

What's it doing?

321

:

And that's the question

the CMOs trying to answer.

322

:

How would MMM, like what does it compare

that spend versus a particular outcome?

323

:

And try to show a causal

relationship between them.

324

:

Mark Stouse: yeah.

325

:

Multiple variables, right?

326

:

So it's highly context oriented.

327

:

All of these models are.

328

:

Seek to capture as much of

a known context as possible.

329

:

what a lot of marketing teams figure

out using exactly that scenario,

330

:

is they've been trying to justify

that display ad budget in terms of

331

:

demand gen, in terms of, performance

marketing, And that is just not the

332

:

way that that typically plays at all.

333

:

That is a brand reputation investment.

334

:

know, you kind of think about

marketing expense as being

335

:

essentially two big chunks, right?

336

:

Branded demand brand is easily two to

three times time lagged in its effect.

337

:

Than demand is.

338

:

It also sticks around a lot longer.

339

:

The effects of brand investment

doesn't deteriorate very quickly

340

:

unless there is a major scandal of some

sort, something like that, that all

341

:

of a sudden breaches the trust wall.

342

:

But absent that right, it hangs

around The halflife is quite extended,

343

:

whereas the Halflife on demand

investment is highly perishable.

344

:

Lasts maybe a couple months and

then it's gone, so the snows melt

345

:

a lot faster with demand than brand.

346

:

You have to then say to finance,

never discussed time lag with you

347

:

before specifically, but we all know.

348

:

That marketing takes

time to have an effect.

349

:

That simple statement is insufficient

because if we don't know the

350

:

time lag, we will never know.

351

:

The ROI, so we now have the ability

to say that this investment over

352

:

here doesn't really drive demand,

doesn't really pay off on this side.

353

:

There isn't a quick return a demand

perspective, but from a brand perspective,

354

:

this is how it is improving average

deal size and average deal velocity.

355

:

Those are the two big ones that we see

again and again on brand investment.

356

:

It is grease on the wheel of the deal.

357

:

That's what brand

reputation really is, right?

358

:

It makes people buy more than

they otherwise would and buy

359

:

faster than they otherwise would.

360

:

that That is key in B2B because

it's a higher cost, higher

361

:

risk by decision to begin with.

362

:

And then today, if we layer in all the

risk factors that people are worried

363

:

about it's even more important,?

364

:

I mean, If you really want to understand

how trusted you are by your customer

365

:

base, look at your average deal velocity.

366

:

If you really want to understand how

confident people are in your product's

367

:

ability to generate massive amounts of

value for them, look at their average

368

:

deal size, particularly year one, right?

369

:

Certainly year two, your, a

lot of people, just as a matter

370

:

principle today, are buying small

in year one and testing it out.

371

:

If you see a major uptick in

year two and deal size, that is a

372

:

major vote of confidence in you.

373

:

you don't, you need to

really find out why that is.

374

:

'cause I guarantee there's a reason.

375

:

Justin Norris: Looking at our

example, let's say the dataset that

376

:

you had, the company had always

been running those display ads.

377

:

Do you need a period where those display

ads are not there in order to demonstrate

378

:

the impact of having them or not?

379

:

Or is there a way, even though they've

always been there, to somehow demonstrate

380

:

the impact that they're having?

381

:

Mark Stouse: I think that the answer to

that question is very much contextual.

382

:

if we were talking about a correlation

analysis, meaning this ad buy right

383

:

against revenue, Then yeah, you would

definitely need some interruptions

384

:

to be able to essentially create a

385

:

Justin Norris: Oh, control, that's

the word I was looking for too.

386

:

Yep.

387

:

Mark Stouse: If you're talking about

multi-variable though, where a lot of

388

:

things are changing all the time around

it, it's not necessary that's actually

389

:

one of the things that's really, really

cool about MVR multi-variable regression

390

:

is that because you're bringing in so

many different variables and because

391

:

everything changes, particularly the

stuff that you don't control you're gonna

392

:

see exactly what you're talking about.

393

:

You're gonna see implicit AB test.

394

:

Evolving across time.

395

:

Justin Norris: We're

talking about limitations.

396

:

And one of the things we're talking

about was being an aggregate

397

:

versus tracking individual users.

398

:

Mark Stouse: Yeah, so one of the things

that's really important to say about

399

:

multivariable regression slash go to

market, mixed modeling, et cetera, right?

400

:

Is that it is looking at all the

causal relationships in aggregate.

401

:

So it's not possible to say that all

these things cause this one identified

402

:

person to take the steps that they took.

403

:

That's not what go to market

mix modeling is all about.

404

:

It's about looking at a population wide

trend that's causal for a lot of people

405

:

one of the things that companies really

appreciate though about this is that.

406

:

There's no PII in proof

or in, a regression model.

407

:

So you don't have the security

concerns that you would have in

408

:

MTA or any of the other touch based

approaches, which are trying to tie.

409

:

Broad investments, programmatic

investments to impact on a particular

410

:

human being or particular company.

411

:

And to do that, you have to

pierce the veil on identity.

412

:

And that's a problem.

413

:

Justin Norris: So it's

looking at the big picture.

414

:

So cookies, not an issue,

GDPR, not an issue.

415

:

It's really comparing

bulk data sets and the

416

:

Mark Stouse: In fact, the only GDPR

thing that we ever have to encounter is

417

:

the login information for users of proof.

418

:

So we keep that secured, But in

terms of the data that's in the

419

:

computations, it's a non factor.

420

:

It's just non issue.

421

:

Justin Norris: Is there a way to

explain that multi-variate regression

422

:

to someone who maybe they've got high

school math, first year university

423

:

math, 'cause maybe that's a challenge.

424

:

As you said, marketers, some of them

are more quantitatively oriented.

425

:

Not all of them are.

426

:

And they're gonna need to trust,

this data, so they'll need to be able

427

:

to make sense of how it's produced.

428

:

You know, If it's just a kind

of computer and it spits out an

429

:

answer like, yep, do X or do Y.

430

:

Mark Stouse: They're already effectively

thinking in this way because they're

431

:

spreading out their investments

across all kinds of different

432

:

channels, So they're acknowledging the

multi-variable world in which they live.

433

:

What they are not necessarily

acknowledging is the fact that it's not

434

:

a list of one-to-one correlations, My

investment in display ads versus revenue,

435

:

my investment in social versus revenue,

my investment here versus revenue.

436

:

That's not it.

437

:

It's a tapestry that weaves back and forth

with different time lags associated with

438

:

hmm.

439

:

Literally the only way capture that

is with multi-variable regression

440

:

or econometrics or, marketing

mix modeling, go to market mix

441

:

modeling, right there, it's all

442

:

the same thing,

443

:

Justin Norris: the fir, the

first thing that you mentioned

444

:

Is what we're conditioned to look for.

445

:

Like we were conditioned to say like, I

invested a dollar here and it produced

446

:

two, like these definitive statements.

447

:

I think if I'm interpreting you

correctly, you're saying reality is not

448

:

that straightforward most of the time.

449

:

Mark Stouse: The final output

can be that straightforward.

450

:

Okay?

451

:

But the causality elements on

it, are going to vary with the

452

:

wind, so a great example of this

is uh, Johnson Controls, right?

453

:

So even before Covid really became a

story, uh, their analytics started to

454

:

say in terms of forecasting, right?

455

:

for reasons we don't understand.

456

:

All these different channel investments

that we have historically made are

457

:

showing that they're not gonna be as

effective in six months as they are now.

458

:

A lot of these were physical events.

459

:

, this was not due to marketing

data in these models.

460

:

This was all about external data

was already up these early signals,

461

:

they decided, you know what?

462

:

We are gonna fly by instrument.

463

:

We're not going to fly by what we can see.

464

:

And so they started removing investments

early a lot of these things and they were

465

:

able to claw back quite a bit of money.

466

:

And then all of a sudden, real

life proved the analytics accurate.

467

:

And so they were very happy about that.

468

:

And then finance came later and said, Hey,

so sorry, we're gonna have to cut you.

469

:

30 or 40%.

470

:

I can't remember exactly

what it was, but it large.

471

:

And they said, let us show you something.

472

:

Right?

473

:

And they modeled the effects of those

kinds of cuts over the next three

474

:

years, and it was so detrimental

to the business that finance took

475

:

a lot less and went elsewhere to

get whatever they needed, right?

476

:

So this is an example, not only of

how a marketing team used it to avoid

477

:

waste, but also to avoid cuts that would

be ultimately bad for the business,

478

:

Justin Norris: and the sort of external

data you just described, those are things

479

:

that's not specific to any one business.

480

:

It's common to the macro environment

that many businesses are working in.

481

:

Like if I come to you, do I need to

bring that external data on my own?

482

:

Or do you have sort of global

external data that you can match

483

:

up with my business specific

data and include in the model?

484

:

Mark Stouse: So we are not a

data provider, but we can and do

485

:

all the time help people locate

data sets that are usually free.

486

:

That kind of stuff is usually

free, either from the government

487

:

or major financial institutions

or universities, things like that.

488

:

So that's not a problem.

489

:

And it's great.

490

:

Usually just fantastic data.

491

:

I mean, It's been totally scrubbed,

so that's not a big deal at all.

492

:

In fact, we live in the golden

age of data availability.

493

:

Even if it's not free,

someone is measuring it.

494

:

Purely speculatively in the belief

that somebody is gonna wanna buy it.

495

:

And so today with a credit card,

you can subscribe to all kinds of

496

:

data sources, cost effectively.

497

:

Particularly given how important

it is to maximizing the investment.

498

:

The other thing that I would

just point out is ROI is really

499

:

important, but you only know RROI.

500

:

Looking backwards, right?

501

:

ROI is a historical assessment.

502

:

What is really important is

forecasted, ROI, and then the

503

:

comparison between it and actuals.

504

:

So again, this is very much like

public companies issue guidance

505

:

and then they issue regular

updates against that guidance.

506

:

That is exactly what the C-suites of many

companies are, demanding today, right?

507

:

It's like an investment

uh, to start a new company.

508

:

The first question the investor's

gonna ask is, what do you expect

509

:

this to do in year 1, 2, 3, 4, 5?

510

:

And what's the basis for that projection?

511

:

If you're just extrapolating from hope

and best wishes and all that kind of

512

:

stuff, that's not much of an argument.

513

:

But if you're doing regression based

analysis, that starts to mean something.

514

:

Justin Norris: And listening to the

examples that you described, it does feel

515

:

like a very enterprise oriented solution.

516

:

If I work at a company, 400

people, 50 million a RR, is this

517

:

something that can work for us, or

do I need to be of a certain size

518

:

threshold for it to be useful?

519

:

Mark Stouse: No, actually

it totally scales.

520

:

The reasons for investing in it are

going to be different in a small,

521

:

medium, or enterprise type business.

522

:

But it totally scales.

523

:

And the cost is totally approachable,

even for mom and pop, right?

524

:

So the reasons for doing it as

a small, and let's say the lower

525

:

half of the medium sized business.

526

:

Are mainly because if you make a bad

investment in whatever, It has a almost

527

:

immediately negative effect on cashflow.

528

:

Too much risk, so they are

modeling to avoid that.

529

:

We do have some small customers and

that's their main reason for doing it.

530

:

If you are an enterprise and

you're spending, I don't know, $200

531

:

million a year on marketing, right?

532

:

you spend it wrong, what the CFO

is most concerned about is things

533

:

like opportunity cost, right?

534

:

That's the comparison that's going on

in their mind all the time, particularly

535

:

today, is what are all the different ways

that I can spend this dollar and what

536

:

are the most effective, or what's most

likely to give me the biggest impact?

537

:

That's the competition.

538

:

That is the Game of Thrones,

particularly right now in budget season.

539

:

budget season right now is

sort of year round now, right?

540

:

Because everything is so pressed.

541

:

So if you are a marketing leader, you

are in competition to show that money

542

:

spent with you is a better return.

543

:

Than if they spend it in

it or HR or whatever, right

544

:

you also need to really understand

this, you marketing by definition,

545

:

is a non-linear multiplier of areas

of business performance that are

546

:

linear, one of which is sales.

547

:

So in simple language, your leverage,

you are bringing huge amounts of

548

:

leverage to sales performance that

sales cannot create for itself, right?

549

:

not a ding on sales, it's just the nature

of the reality of the situation, what

550

:

do I mean by linear and non-Linear, okay.

551

:

It means that if I were to go

to my CRO and double goal for.

552

:

For 2024, the first conversation

that they're gonna want to have

553

:

with me is about essentially

doubling their Salesforce, right?

554

:

Because the relationship between

revenue coming in and the cost of that

555

:

revenue in the form of sales team is

it's known it's a linear function.

556

:

And that's because it's the

collective performance of a

557

:

lot of individual performances.

558

:

Okay?

559

:

So a bell curve your sales team's

performance is gonna be on a bell curve.

560

:

It's just fundamentally linear.

561

:

The whole reason why modern marketing

was created in:

562

:

to bring non-linear leverage to

that whole equation, even in B2C.

563

:

So what does that mean?

564

:

It means that because of the way

marketing is, were to have the

565

:

same conversation with the CMOA,

we're doubling the revenue goal.

566

:

Might have to increase marketing spend

by 25%, maybe 20%, there's already

567

:

a ton of leverage built into it.

568

:

And if you want an easy understanding

of ROI for marketing, it is the

569

:

extent of that multiplier, so

we'll talk about it this way.

570

:

You are basically saying that

marketing's mission is to help

571

:

sales more product to more people.

572

:

That's revenue faster.

573

:

That's cash flow from

revenue more profitably.

574

:

That's margin impact than

sales could do by itself.

575

:

That's the key phrase.

576

:

The extent to which that is

true is the ROI to the business.

577

:

So does that actually

look like in real life?

578

:

Well, and a lot of really,

really great B2B marketing go-to

579

:

market kinds of operations.

580

:

ratio is somewhere between 10 and 20 x.

581

:

So that means that if you take marketing

away entirely, which I think would be a.

582

:

Something that not even the most draconian

CFO would ever contemplate, but let's

583

:

just say it is the ultimate AB test.

584

:

So we're gonna completely

shut down marketing.

585

:

You're gonna see a massive fall off.

586

:

It may take a year, but you're gonna

see a massive fall off in sales

587

:

productivity that is going to reveal

the extent of the marketing multiplier.

588

:

It's like literally guarantee able because

sales create the leverage for itself.

589

:

It's just not the way it works.

590

:

Justin Norris: If a company is

getting started with this solution

591

:

what would the process be like?

592

:

What data sources would we need to bring?

593

:

How much time does it take to build

the model, that sort of thing.

594

:

Mark Stouse: the very first step is, and

this is part of the onboarding for us,

595

:

is that we sit down with the customer

and we say, look, what are your top

596

:

questions that you most want to know

The answer to this is a mixed audience,

597

:

usually of marketers and business leaders.

598

:

And doesn't matter what your job is.

599

:

Nobody has any problem rattling

off their list of questions.

600

:

Those questions then generate what's

called a model framework, could easily

601

:

analogize that to a recipe card., So this

is gonna be a punch list of data types

602

:

that you're going to need to be able to

supply to the model to compute the answer.

603

:

then it's gonna outline the model

itself, algorithmically speaking.

604

:

When you actually make the dish

from that recipe, that is the model.

605

:

Usually the way this actually

works is it's very fast.

606

:

It's usually a matter of

weeks, like less than a month,

607

:

that kind of timeframe where.

608

:

The analyst and the business user.

609

:

Could be a marketer, could be

finance guy, could be whatever.

610

:

Are collaborating in the tool

on a minimum viable model.

611

:

And at some point business user

says, that answers my question.

612

:

We need to put this model in production.

613

:

You hit the big red button,

it goes into production.

614

:

After that, it's pretty autonomous.

615

:

There's kind of some DevOps type work,

you you know, you have to maintain the

616

:

model and all that kind of stuff, right?

617

:

But for the most part, it is

doing its thing on an automated

618

:

basis and you're getting whatever

cadence is right for your business.

619

:

Daily, weekly, monthly, you're

getting an update on demand.

620

:

Justin Norris: the.

621

:

. Analyst that you just described, is

that someone that you're supplying

622

:

from your team or is this kind

of a bring your own process?

623

:

Mark Stouse: we We have a large partner

ecosystem that, we can recommend from.

624

:

We also occasionally can do it ourselves

on a managed service basis that is

625

:

actually increasingly popular as

teams get thinned out and they know

626

:

that they need this capability, but

they don't have the bandwidth or the

627

:

expertise to manage it internally.

628

:

And so that's actually extremely popular

these days and very cost effective.

629

:

the key thing is that there's lot of.

630

:

You could easily spend 2020 5K

upfront in time, not in licenses.

631

:

Okay.

632

:

But in time to get everything

set up but once you've done that,

633

:

right, again, the models, unless

you need more models, right?

634

:

The models are doing their thing, right?

635

:

You don't have ongoing major, investment,

know, every month gotta redo the

636

:

whole thing from the ground up.

637

:

That's, exactly the kind of thing

that proof was built to eliminate,

638

:

Justin Norris: if you add a

new, you add a new channel, do

639

:

you have to update the model?

640

:

Or the model can be flexible

enough to just say like, oh, you're

641

:

doing LinkedIn advertising now.

642

:

That's fine.

643

:

We can just incorporate that as we go.

644

:

Mark Stouse: you could

do it either way, right?

645

:

I think be the best practice is that

you clone the model that you have and

646

:

then you add the new data streams to it

that you're maintaining the integrity

647

:

of the original model for comparison.

648

:

And yet you are updating

it that way, right?

649

:

And then at some point you're gonna

dispense with the oldest version

650

:

of the model altogether, So it's

not like you just see a massive

651

:

proliferation of models across time.

652

:

But you do need to do this in

an orderly way so that everybody

653

:

doesn't lose reference points.

654

:

Justin Norris: you mentioned

competitors and I've seen competitors

655

:

popping up in the market, even . A

multi-touch attribution vendor

656

:

that's adding MMM to their mix.

657

:

What's your outlook on how easily other

vendors could recreate these capabilities?

658

:

Do you feel that you have a fairly strong

competitive moat around the offering

659

:

you've built, or will other people be

able to jump into this environment and

660

:

offer similar things fairly easily?

661

:

Mark Stouse: So it's really important

to say this, the competitive advantage

662

:

that anybody has is not in the math.

663

:

So if people propose that they have

some kind of Super Cal Flagal algorithm,

664

:

you need to be very careful about that.

665

:

' cause the odds also are that it's

not transparent, it's a black box.

666

:

And also.

667

:

If it is transparent, you probably

don't have the ability to evaluate it.

668

:

In our particular case, the IP that really

matters is in how we automate it, how we

669

:

scale it, how we make it consumable and

approachable and understandable and how

670

:

we do it at a particular price point.

671

:

And I think that this is the other

thing that is highly relevant today,

672

:

not just in this area, but across

SaaS, is that prices are coming down.

673

:

Uh, We're gonna see a fundamental

change in SaaS pricing.

674

:

The days of, you know, every year

having a pricing increase are over.

675

:

It's just done.

676

:

We are very well fixed

competitively speaking.

677

:

Do have a, I think, a really solid moat.

678

:

So there are a lot of competitors today

that have some great products, but these

679

:

are products that were conceived of

and written by and for data scientists.

680

:

and, And a lot of 'em

are gorgeous by the way.

681

:

Like, graphically, they're gorgeous, but

if you expose them to a normal business

682

:

user, a marketer, sales leader, whatever

gonna stare at that screen and go, I have

683

:

not a clue in the world what that means.

684

:

Like how do I make a better

decision based on that?

685

:

You've just now thrown a

lot of friction into it.

686

:

It's taking more time.

687

:

You've gotta spend a lot of time

translating the data science outputs

688

:

into something that is usable.

689

:

And we don't do that, we built it with

the end user in mind violating any.

690

:

Data science principles.

691

:

That's our big advantage.

692

:

Think one of the biggest things I can

say about this, just to sum up, is

693

:

that there's a reason why large B2C

marketing teams, CPG uh, hotel and

694

:

hospitality, retail, whatever, right?

695

:

They control all four pss of marketing and

they have, at worst, a defacto authority

696

:

over and responsibility over the p and l.

697

:

Of their product with their marketing.

698

:

There's a reason for that, they are

using econometrics slash Go-to market

699

:

mix modeling slash MMM, using that to

optimize and to understand causality

700

:

and to optimize based on that causality.

701

:

And they're also investing a ton of

money in market research, which a lot

702

:

of that ends up in the models, right?

703

:

So if B2B marketers want to have the

same attributes their B2C brothers and

704

:

sisters, gonna have to do what B2C does.

705

:

This is one of those situations where it's

like gravity, You can disagree all you

706

:

want to with gravity, and if you throw

yourself off a building, it's not gonna

707

:

end well, the same thing is true here.

708

:

This is a mathematical principle at work.

709

:

It's mathematical law of gravity

with quotation marks around it.

710

:

And so everything that I've

said today, I've, and I really

711

:

try really super hard to do.

712

:

This is not my opinion.

713

:

I'm just representing a level of fact

that people can either accept or reject,

714

:

but it doesn't change the truth of it.

715

:

Justin Norris: We will include a link to

your website so people can uh, check it

716

:

out, learn more, look at your resources.

717

:

And I'm excited to see where this goes.

718

:

Mark, thank you so much

for chatting with me today.

719

:

Mark Stouse: You're welcome.

720

:

Thank you.

Links

Chapters

Video

More from YouTube