In this episode, Sahil Patel, CEO of Spiralize, discusses predictive conversion rate optimization for B2B SaaS companies.

What Spiralize Does:

• Spiralize offers predictive conversion rate optimization by scraping and analyzing A/B tests from 34,000 websites.
• They identify successful A/B tests and implement them for clients to optimize conversions.

Understanding A/B Testing:

• Sahil explains A/B testing using the analogy of a clinical drug trial, emphasizing the importance of statistically significant results.

Scraping and Analysis Process:

• Spiralize utilizes their proprietary software to crawl the internet, capturing visual differences in A/B tests.
• They identify common elements across successful tests to determine proven winners.

Expertise and Value Proposition:

• Sahil highlights the expertise Spiralize brings by analyzing data from thousands of websites, providing insights beyond what individual companies can achieve.
• Spiralize offers a blend of product and service to address the challenges and complexities of A/B testing.

Customer Base and Business Model:

• Spiralize primarily serves B2B SaaS companies with at least 500 employees, spending $50,000 to $70,000 monthly on marketing.
• The company operates on a performance-based model, starting with a 90-day pilot followed by a monthly fee.

Acquisition Strategy and Market Focus:

• Spiralize relies on client referrals and relationships with CMOs, leveraging the frequent job changes in the industry.
• They target the B2B SaaS market due to its relative newness in conversion rate optimization, offering more opportunities for growth.

Team Structure and Engineering Role:

• Spiralize employs a team of software engineers to improve their proprietary software and assist clients in building and running A/B tests.
• Engineers focus on implementing significant changes in user experience rather than minor adjustments, maximizing impact.

Future Vision and Goals:

• Sahil envisions CRO becoming as essential as SEO, with Spiralize leading the industry in the next five to ten years.
• The company aims to democratize conversion rate optimization, making it a standard practice for all websites.

Hello, everyone.

2

Welcome to the B2B SaaS podcast.

3

I'm your host Rupesh Javvarman.

4

Today we have Sahil Patel here with us.

5

Sahil is the CEO of a

company called Spiralize.

6

Hey Sahil, welcome to the show.

7

Sahil Patel: Hey, really glad to be here.

8

Thanks for having me.

9

Upendra Varma: All right, Sahil.

10

So let's try to understand

what Spiralize does, right?

11

And how you sort of, you

know, help your customer base.

12

Sahil Patel: Yeah, so we do

predictive conversion rate

13

optimization for B2B SaaS companies.

14

There are 34, 000 websites that run

A B tests somewhere on their website.

15

We scrape all of them.

16

We find the best A B tests.

17

We shamelessly borrow them and we

run those tests for our clients.

18

Upendra Varma: So explain more, right?

19

So what do you mean you

scrape APTest, right?

20

So I think the way I understand people

run these APTest is, hey, you, you have

21

a period, like you have a timeframe where

you're running some, showing something

22

for like 10 days and something else for

the next 10 days, and then they're sort

23

of looking at metrics and all of it.

24

Like, how do you

25

Sahil Patel: that's not a B testing.

26

That's actually not a

because it's something.

27

I'm not sure what that is,

but that's something else.

28

Upendra Varma: so what, what

do you mean maybe testing in

29

Sahil Patel: uh, so great

example of a B testing.

30

I'm going to step outside websites and

digital and everything else for them.

31

Think about, uh, like

a clinical drug trial.

32

You have a new medicine.

33

What you do is you get a representative

sample of people, men, women,

34

young, old, all those things.

35

And half of them get the drug and half

of them get what's called a control.

36

Which is a placebo and at the end and

you gotta make sure you run it from

37

like say the start of the month to the

end of the month and then at the end

38

you compare how many people got better.

39

Was there a better outcome and what you're

looking for is a difference because even

40

in a placebo about maybe 5 percent of

the people just get better on their own.

41

So you gotta try and beat what's

called beating the control.

42

Now I'm starting to use a

little statistical language.

43

That's what we're doing.

44

We're running those like a

drug trial for your website.

45

So let's say you're a B2B SaaS company

and you want to get people to sign up

46

and do a free trial of your product.

47

Tons of companies do that.

48

We're going to make two

versions of the page.

49

Now you definitely had that part right.

50

And then for, let's say a month,

half of everyone will randomly get

51

the normal version of the page.

52

But it's really important during that

same time cohort, that same month

53

at the same time, randomly, half the

people will get the other version.

54

And you're going to compare,

did you get more on.

55

The B.

56

Or did you get more on the A?

57

That's an A B test.

58

Upendra Varma: so my question is like,

what do you mean when you say we're going

59

to scrape 34, 000 websites and understand

60

Sahil Patel: do we scrape?

61

Yeah, so we've built our own.

62

This is our own sass product and it

crawls the Internet and it finds anyone

63

any website that's running a what's

called in browser some people client side

64

testing and that's the first thing we do

to We capture the visual image capture.

65

We can visually see the difference

between the A version of a page

66

and the B version of a page.

67

By the way, you could do this right now.

68

You could go to a website, you could

open it in two different browser

69

windows, and see, is there an A and a B?

70

Now you know they're running a test.

71

Now, anyone can do that.

72

We've automated it, so

we're doing it at scale.

73

The third thing we do, and this is

where there's some special sauce.

74

Capture and find the common

elements across all these tests.

75

Because if one company

runs one test, so what?

76

I'm not running to tell

everyone out to go copy it.

77

If we see 10 companies all run the same

test and they converge on an answer,

78

that's what we call proven winner.

79

Those are the kind of

tests we like to run.

80

Upendra Varma: Okay.

81

So I know I sort of get a sense of,

you know, what you're trying to do

82

here, but I just want to, you know,

get into any specific example here.

83

Right.

84

So what does that mean?

85

Let's say I'm, I'm a SaaS company, right.

86

And I just want to optimize my pricing

page so that, you know, more, more people

87

sort of signs up for example, right.

88

So there are going to be a bunch of

templates that I could see, you could just

89

put all of them, you know, in, you know,

stacked, or you could have them, you know,

90

sorted in vertically, horizontally, sort

of various templates that I could get,

91

right, maybe hundreds of examples, right?

92

So, yes, I could run my own A B test and

maybe could figure it out for myself.

93

So what sort of expertise are you

bringing in by analyzing these 34, 000,

94

you know, sort of companies or websites?

95

Sahil Patel: yeah, why not do it

on your own if you've yeah, by

96

the way, there's some great self

service tools you could do out there.

97

Here's the challenge.

98

First, testing, A B

testing is a cruel game.

99

Industry average, only 10

percent of tests actually win.

100

So think about that.

101

If you, if that, those are not good odds.

102

Okay.

103

If you, if you went to the casino

and you, whatever your favorite

104

game is, let's say it's blackjack.

105

You had to play 10 hands to get one win.

106

You're going to run out

of money really quickly.

107

Upendra Varma: Sure.

108

But, but, uh, okay.

109

Makes, makes a lot of sense.

110

I'm, I'm still sort of, I think

I just want to get a sense of it.

111

Right.

112

So this is still a

service that you provide.

113

Right.

114

So there's still some manual.

115

Sahil Patel: It's both.

116

Yeah, it's a, it's a

product and a service.

117

It's both bundled together.

118

Upendra Varma: Okay.

119

That makes sense.

120

Because yeah, I mean, I could see

how you like, it's pretty hard for

121

a product to do it out of the box.

122

Right.

123

So then how could you even do that?

124

Sahil Patel: Yeah.

125

Well, it's not that it's hard.

126

It's that the odds are not in your favor.

127

Number one.

128

Number two, let's say the average company.

129

You got to get a whole bunch of people in

the room and agree on what test to run.

130

They didn't even know what

the test should look like.

131

Everyone's got an opinion.

132

The design guy has an opinion.

133

The product gal has an opinion.

134

CEO walks in the room and

goes, Oh, I heard if you put

135

the CEO's dog on the homepage.

136

Conversions will be through the roof.

137

And what you end up doing

is testing based on opinion.

138

Or even worse, you get someone with

just a little bit of knowledge, and

139

they think they've got best practice.

140

What they're really doing is

testing based on instinct.

141

They go, Oh yeah, I've been, I've

been doing digital for 10 years.

142

I know we should make all the

143

Upendra Varma: So in this

particular scenario, right.

144

So how would you help that particular

customer to get out those top, you

145

know, and number of tests to run.

146

So what exactly would you do

and how exactly do you do that?

147

Sahil Patel: Yeah, we'd come in and say,

Hey, we've looked at 30, 000 websites.

148

First of all, do you think

anyone in the room has personally

149

looked at 30, 000 websites?

150

No, it's not possible.

151

Second thing is that we've clustered

them and we've looked at just

152

the companies in your vertical.

153

Number one, we've looked at just

people that ran tests on homepages.

154

And we've looked at just the

companies that offer a free trial.

155

And we've identified these seven tests.

156

And we know how often they win and

how much lift they should produce.

157

Lift just means how many more conversions

you're getting compared to your control.

158

Upendra Varma: But how

do you get that data?

159

The whole conversion data, that's

still an outcome that, that only

160

the not destiny company sort of can,

161

Sahil Patel: Ah, yes, yes.

162

So I love that question.

163

How do we know the stats?

164

Because the scraping of the Internet,

what we're looking for signals.

165

There's a ton of noise.

166

gives us valuable signals.

167

When we find something that repeatedly

wins, we then run it for our clients.

168

We're the largest test in the world.

169

Tremendous scale.

170

Because of that, we've run

tens of thousands of tests.

171

We've been doing this for

more than eight years.

172

When we run tests, that's

the first party data.

173

We know the exact statistics.

174

Of the performance of each test.

175

Upendra Varma: So that's the sort of

expertise that you're bringing in, right.

176

That a typical SAS company wouldn't have.

177

Right.

178

So that makes a lot of sense.

179

I talk about the

180

Sahil Patel: Plus, we've got

the in house service team.

181

So we've got the in house designers,

UX, UI, copywriters, full stack

182

engineers to actually build the test.

183

Upendra Varma: Yeah.

184

So I'm going to come to that.

185

Right.

186

So I just want to get a sense

of your customer base today.

187

Right.

188

So who are you primarily serving to

and like how many active customers

189

Sahil Patel: Um, yeah, so, um,

we've served thousands of customers.

190

I have to be careful here.

191

I can't, I can't reveal kind of our

exact, uh, customer account, but the

192

typical customer is a B2B SaaS business.

193

They've got at least 500 people.

194

And a great indicator is that they're

usually spending around 50, 000 to 70,

195

000 a month on marketing to get traffic.

196

Could be SEO, could be search engine ads,

could be social media ads to get traffic.

197

Why do we use that?

198

It's because it's a signal They've got a

high value conversion and they've got a

199

maturity about their marketing function.

200

Is the first problem most

companies have to solve is get

201

people to come to your website.

202

You can't really, you can't do

what we do until people are there.

203

Our job is to take the traffic you already

have and get more of it to convert.

204

But if you don't have any

traffic, we can't really help you.

205

Upendra Varma: Makes sense.

206

Can you just sort of, you know,

give a very broad range, right?

207

Whatever you're comfortable with.

208

I've been talking about 100

customers or 500, 000 customers.

209

What's that number look like?

210

Or is it just five or 10 handpicked

companies that you work with?

211

So that we could sort of, you know,

212

Sahil Patel: yeah.

213

Upendra Varma: the scale

at which you're operating.

214

Sahil Patel: you this.

215

Yeah.

216

At any given time, we're

working with about a hundred.

217

Upendra Varma: Yeah, that

makes a lot of sense.

218

Yeah.

219

Uh, and talk about the

business model here, right?

220

So, how does this work, right?

221

If I'm a B2B SaaS company

looking at your service, right?

222

So, how do I engage with you?

223

Is it a one time thing that I

run, figure something out, and

224

then I just go, or like, do you,

like, how do you keep me, right?

225

So, how, how long does a typical

customer stick with you, and like,

226

how does that engagement model work?

227

Sahil Patel: Yeah, it's

interesting because it's a, it's

228

a hybrid business tech enabled.

229

Um, it's not like a typical SAS product

where we sign you up and we hope you stick

230

around for five years and we create a long

term contract and try and lock you in.

231

It's just not the nature of what we do.

232

Uh, by the way, I ran a SAS business

for 11 years before I ran Spiral Eye.

233

So definitely know that model really well.

234

All of our clients start

with a 90 day pilot.

235

And the fee is a performance fee based

on how much conversion lift they pay.

236

And if we get them no

lift, they pay nothing.

237

We get a little lift, they pay a little.

238

And to earn our full fee, we have to get

them at least 30 percent conversion lift.

239

It's a good number.

240

You know, if you go to your average chief

marketing officer and you tell her or

241

him, we'll increase your pipeline by 30

percent on your highest value conversions,

242

that tends to get their attention.

243

Upendra Varma: Yeah.

244

Sahil Patel: So that's the 90 day pilot.

245

After the pilot, it's

a simple monthly fee,

246

Upendra Varma: Yeah.

247

And then click.

248

Okay.

249

And then based on like how

long they stick with you, they

250

just keep on paying you, right?

251

Sahil Patel: Yeah, if we're getting

the lift, they're happy to pay.

252

I like, there's a good phrase, when

you find gold, you keep digging.

253

Upendra Varma: And just, just

to get a sense of, right, so for

254

example, if a customer were to

stick with you for 12 months, right?

255

So how much would they pay you on an

average, I know there's going to be

256

a spectrum and all of it, but what

257

Sahil Patel: there's no spectrum.

258

Everyone pays the same amount.

259

Actually, it's on our website.

260

It's 25, 000 a month.

261

Upendra Varma: A month.

262

Okay.

263

So basically you said that for

the first 90, 90 days that you

264

do the pilot with them, right?

265

So I think it's, they could just

pay you nothing if you don't

266

showcase any good results, right?

267

But that changes after

that 90 day pilot, right?

268

That makes

269

Sahil Patel: Yeah.

270

So at the end of the pilot,

we add up the results and then

271

they pay us a success fee.

272

And then the vast majority go,

wow, this thing really worked.

273

Let's keep going.

274

Upendra Varma: sort of help us understand

your sort of acquisition strategy, right?

275

So where you're finding all of these

clients and like, how does it all

276

work for you, your own marketing.

277

As

278

Sahil Patel: Oh yeah.

279

You know, I just sit back on the

couch and people call us now.

280

I'm just kidding.

281

Um, look like any other business,

early stage growth business, you gotta,

282

you gotta work multiple channels.

283

So first for the, I mean, the

best source of, of business client

284

acquisition is, uh, client referrals.

285

Some, we do great work for a

chief marketing officer and he

286

or she tells their peers CMOs.

287

They talk all the time.

288

Number one.

289

Number two, we sell into the

marketing office on average.

290

These people change jobs every 18 months.

291

And if they're happy with us, they

get to their new job and they call us.

292

So that's the, um, those are the two

biggest sources of business for us.

293

We also are always trying out different

channels, you know, outbound channels,

294

inbound channels, advertising.

295

Our business is very much, it's a niche.

296

So, you know, kind of running ads.

297

Um, we haven't gotten kind of a lot.

298

There's not a lot of.

299

We haven't seen kind of great traction

on typical advertising channels.

300

Upendra Varma: know, why

just B2B SaaS, right?

301

So is there, you know, why just

this particular market, right?

302

So what, why, why that positioning

303

Sahil Patel: Yeah.

304

Um, there's two reasons.

305

The biggest reason is that

it's an untapped market.

306

So e commerce discovered CRO

probably about 10 years ago, and

307

it's a very competitive space.

308

And B2B SaaS is much newer to the game.

309

So there's more low hanging fruit and

there's more market share to be gained.

310

Upendra Varma: makes

a lot of sense, right?

311

So I talk about your team today, right?

312

So I see you, you've got a lot

of engineers out there, right?

313

So what do they do in this setup, right?

314

I mean, you're, you're more

like a marketing agency, right?

315

You know,

316

Sahil Patel: Uh, we're not, we're not.

317

Yeah.

318

I wouldn't describe us that way.

319

Uh, we're a tech enabled service.

320

Um, that's why we have a

lot of software engineers.

321

We have our own SAS product.

322

We bundle that with the service team.

323

Upendra Varma: so, okay.

324

Sahil Patel: a bit of a hybrid business.

325

Upendra Varma: Yeah, just talk

a bit more about that, right?

326

So what are these engineers

spending all of their time on it?

327

Are they building on and improving

your existing software product?

328

Or are they doing something else

for your client, for example?

329

Sahil Patel: So we do.

330

We do have.

331

So I think you've picked up

on something really good.

332

So we do have a big team of

engineers that are building our

333

software product, making it better.

334

Um, then we also have engineers

who are working with our clients to

335

build and run A B tests for them.

336

Upendra Varma: Okay.

337

And what sort of tests are these

that you would need a good software

338

engineer to come in and build?

339

So what sort

340

Sahil Patel: Oh, I love that.

341

You asked that.

342

It's a great question.

343

Defender.

344

So, um, Here's what I think most people

think of A B testing for a website.

345

They think, oh, okay, I'm going

to take a page and I'm going to

346

change the color of the button.

347

I heard that if you make the buttons

orange, more people will click.

348

Now it turns out, but you're,

you're welcome to try that.

349

Those tests are largely a waste of time.

350

I call those meek testing,

M E E K, meek testing.

351

Meek is weak, doesn't really

move the needle for anyone.

352

And you're really just wasting your time.

353

What we do are big swing tests.

354

We're reimagining the entire user

experience on a particular page,

355

the overall composition of the page,

the imagery, the copywriting, the

356

product claim, the form design,

the layout, the flow, all of

357

the social proof, testimonials,

all of those things together.

358

Um, Now, to do that, you got to write

HTML code to do that number one.

359

Number two.

360

Once you find a go ahead, please.

361

Upendra Varma: So I'm just trying

to understand that you take over a

362

company's marketing website or their

website and then you sort of rebuild

363

it for yourself until something clicks.

364

Sahil Patel: I wouldn't

describe it that way.

365

So we're certainly not overtaking

taking over anyone's websites.

366

We are giving them a plug in tool.

367

And by the way, I just want to make

sure you keep an eye on the clock.

368

I know we've we've booked.

369

So we've got I've got about

two or three minutes left.

370

Um, we're giving them a plug in

tool piece of JavaScript code that

371

will actually trigger the test.

372

And it automatically gives half of the

audience one version of the page, half the

373

audience, the other version, and then it

keeps track of what happens, who converts.

374

Um, so that's how we're actually

doing now inside the testing

375

consoles where we build the test.

376

Upendra Varma: Yeah.

377

Makes sense.

378

And so I have one last question.

379

What's the vision here?

380

So where do you see company in

the next two to three years?

381

What's the big plan and what are

the next milestones that you're

382

sort of You know, aiming to fit.

383

Sahil Patel: We're going

to take over the world.

384

Our goal is that, in my vision, is that

CRO, conversion rate optimization, is

385

roughly where SEO was 20 years ago.

386

20 years ago, some people did SEO.

387

It was leading edge.

388

Not everyone even was told that

you needed to do it, much less

389

what was the right way to do it.

390

Today, I can't imagine a B2B, much

less any website not doing SEO.

391

So today there's a small sliver

of companies that do CRO.

392

In the next five to 10 years,

I think everyone will do it.

393

Not everyone will do it with us.

394

That's fine.

395

Um, we think we're, we're the best option.

396

If you're serious, you want to swing big.

397

Upendra Varma: That makes a lot of sense.

398

Thanks, Sahil.

399

Thanks for taking the time to talk to me.

400

Hope your skills finalized to

much, much greater heights.

401

Sahil Patel: Oh, you're welcome.

402

Thank you.

403

And you're, you're a great host.

404

Really enjoyed the conversation.

405

Upendra Varma: Thank you for that.