Dana Myers, CTO of Lamarr, talks about taking LoRaWAN from curiosity to critical infrastructure. A longtime systems engineer and ham radio operator, Dana explains how his early experimentation with LoRa in the 900 MHz band evolved into deploying hundreds of production devices monitoring water tanks and wells across rural terrain.

He walks through the pivotal moment when Lamarr abandoned an expensive, Raspberry Pi-based “Combox” approach and shifted to low-power LoRaWAN end nodes, cutting costs by an order of magnitude and making the business viable. The conversation dives into what really changes as you move from one prototype to 100 and then to more than 1,000 deployed motes in revenue operation, including hardware revisions, battery budgeting, vendor selection, and the decision to stop building everything in-house.

Dana also breaks down common misconceptions about LoRaWAN, particularly the tendency to treat it like a real-time broadband network. He explains why LoRaWAN requires a mindset shift toward small, infrequent data transmissions, report-on-change logic, and simplicity at the edge. Firmware over-the-air updates, ADR expectations, and backend-driven innovation are all examined through the lens of practical deployment.

The episode closes with Dana’s direct advice to young engineers entering the LoRaWAN space: understand your customers, avoid sunk cost traps, fail fast when necessary, and design for simplicity from day one.

Dana on LinkedIn

Meter.me

• Helium Global IoT Coverage - Want to know if Helium coverage exists where you need it? Check out this map!
• Helium Foundation - The Helium Foundation's IoT Working Group (IOTWG) has generously provided support for the first 6 months of shows, please go check them out and consider using the Helium LoRaWAN as a primary or backup on your next deployment. With over a quarter million gateways deployed worldwide, it's likely that you have and can use Helium coverage.
• Support The Show - If you'd like to support the MetSci Show financially, here's where you can donate on a one-time or an ongoing basis. Thank you!
• MetSci Show - If you'd like to use our IoT or AI Data Value calculators, or you'd like to contact me, the MetSci Show site is the best way to do it.
• MeteoScientific Console - Use LoRaWAN - The MeteoScientific Console allows you to use LoRaWAN today. As long as you have Helium coverage (and you probably do, about 90% of populated areas in the world have a gateway within 2 miles), you can onboard a sensor. You can always check coverage at https://explorer.helium.com and switch to the "IoT" tab in the top right.

Today's guest on

2

MeteoScientific's The Business of LoRaWAN

is Dana Meyers, CTO of Lamarr, Inc.

3

and a longtime systems

engineer and ham radio operator.

4

Dana brings a rare combination

of RF intuition, operating system level

5

engineering experience from his years

at Sun Microsystems, and hard won start up

6

lessons, from taking a LoRaWAN product

from zero

7

to more than 1000

deployed motes in the field.

8

In this conversation,

we talk about how his background in ham

9

radio shaped his understanding of LoRa

before LoRaWAN, the technical and business

10

pivot from a $1,500 Raspberry

Pi Combox, to low power LoRaWAN and nodes.

11

What really changes when you scale

12

from one device to hundreds

in revenue operation, and why most people

13

misunderstand LoRaWAN

as a real time, high bandwidth network

14

when it's actually built for small amounts

of data sent infrequently.

15

This episode is sponsored

16

by the Helium Foundation and is dedicated

to spreading knowledge about LoRaWAN.

17

If you'd like to try Helium’s

publicly available global LoRaWAN for free

18

and support

the show, sign up at metsci.show/console.

19

Now let's dig into the conversation

with Dana Meyers.

20

Dana,

thanks so much for coming on the show.

21

Super excited to have you here.

22

Well, thank you so much, Nik

I was flattered when you asked.

23

I'm happy to be here.

Hopefully I won't let you down.

24

I think you're going to crush.

25

I thought we might start

with how being a ham

26

radio guy has informed your use of LoRaWAN

or RF in general.

27

And do you think that's something that

a lot of LoRaWAN folks should explore?

28

That is an excellent

and outstanding question,

29

and it kind of makes you wonder how

well you actually know me.

30

We met before because my initial interest

31

in LoRaWAN in 2017,

and it wasn't even LoRaWAN, it was LoRa.

32

I first saw the LoRa radios, you know,

33

I was selling them on breakouts

and I hey, I props I mean, here's a

34

a feather with a one of the

I really hope RF modules on it.

35

And and I thought wow spread spectrum.

36

It's chirp.

37

Oh wow 900 mag cool.

38

That's underutilized band.

39

It's a special band

as a as a radio amateur yourself.

40

You probably also understand

it has multiple users

41

and there's a lot of drama potentially

playing out there right now as well.

42

There always has been on 900 mag,

but it's more than ever right now,

43

so I initially was interested in LoRa

just as a radio nerd,

44

saying, okay, these 100 kilowatt radios

45

with kind of marginal antennas,

what can I do with this?

46

And so I actually built you can laugh,

I built an Apex implementation

47

and I'm trying to remember all the bits

I used, but it actually worked.

48

And was running on a feather.

49

I had a feather that I just built it

to send off packets under command,

50

and I drove around and I actually

had it up linking to the app sis

51

so you could go look on whatever app site

you know, on in the internet and see me.

52

And these are reports coming at a 900 mag

just range wasn't great.

53

My antenna was here at the house,

54

and I think I had about 9

or 10 miles of range across the flat.

55

I live adjacent to a a marsh.

56

I overlook a marsh.

57

I'm only about 90ft above sea level,

but the marshes sea level.

58

So it worked pretty well.

59

So that's how I get started.

60

That's

what really raised my interest in it,

61

and I had a lot of fun

experimenting with it.

62

I was very surprised

multiple times at what you could do

63

with that very low power

radio in particular, I,

64

I know that I went on a drive

one time and was driving

65

across a bridge

25 miles from here, above water.

66

It's about 200ft above the water

67

and excellent coverage all across

that bridge and 100 megawatt radio.

68

Just astounding capability.

69

It is impressive.

70

Yeah, just how far you can go

with how little energy.

71

You know, the processing game, the coding

gain, spreads spectrum in general.

72

It's all about spread your energy wide

and then you bring it back.

73

And and when you do that,

all the narrowband interferes,

74

they go away, they go away.

75

So are they are they the reduced.

76

So that's and particularly on under med

with so many users

77

that in particular

smart meters up here are utility

78

meters are all chattering away

constantly in an actual mesh network.

79

So it works very well in that presence.

80

So and maybe that's a nice segue into you

81

start off with LoRa at least in 2017.

82

And if we fast forward

83

and now you're working, doing

a lot of hardware stuff over at Lamar,

84

and you guys have more than one

LoRa radio.

85

Can you walk me through what it was like

to go from kind of 1 to 100 and maybe 100

86

to 1000?

87

Well, I'm going to start with

going from 0 to 1, okay.

88

Because that that was fine.

89

This was a case of where

so I'm a radio nerd,

90

but I'm also into cooking barbecue

and I enjoy wine.

91

And we have a wine country here.

92

And I have good friends in the wine

country here that I've known for years,

93

and I cook regularly with an association.

94

I'm actually a member of the Sassoon

Valley Vinegar Association.

95

I don't grill wine, I don't make wine,

but I'm a member of the association.

96

And through that,

I got to know a gentleman named George.

97

And George had been working for Lamar

or the predecessor

98

to Lamar Valley Internet,

for a number of years.

99

And he kept saying, you gotta come middle

for he's doing this great stuff.

100

You got to come meet over.

101

And I sure, sure, sure.

102

But I had plenty to do and all that.

103

I had a day job.

104

I was at Sun Microsystems from 1993,

right up until the acquisition

105

by Oracle in 2010,

106

You're on one of their patents, right?

One of the sun patterns.

107

I have set up a couple couple.

108

Okay. I'm working on.

109

Yeah, I spent some time in a job

of business development, even though

110

most of my time at sun was doing Solaris

operating system development,

111

I was behind getting functional ACP

into the service kernel.

112

And ironically, Intel

113

was our big partner for that back

when Intel thought we were a competitor.

114

But you know, funny how time works out.

115

So without digressing too much there,

I had a day job in

116

and I was busy and and over sounded great,

but I, you know, it just sounded.

117

Yeah, well, eventually I was lured in

118

Yeah.

119

Late 2019, mid 2019, late 2019.

120

And I started advising

and I did this on spec

121

just to help out my friends, you know,

thought it was interesting people

122

that I mean, on one hand I got to get paid

for the things I do that had value.

123

On the other hand, I don't demand to get

paid because oftentimes I,

124

you know, you got to give to

125

the community and you have to become

there's a lot of fun work to do.

126

Yeah. Yeah.

127

So there's a

there's a definite trade off there,

128

although I'm definitely not on Reddit

doing young people's homework for them.

129

You know, like, you know, you know,

you embedded or are embedded, right.

130

So anyway, you know, for.

131

Yeah. Yeah.

132

You know, well, so I started

attending the meetings and at the time.

133

But meter me what became meter me

what they were doing was building a

134

they called it a comm box

and it was a Raspberry Pi.

135

And I don't know

if over describe this to you.

136

Maybe he did.

So I don't have to rehash that.

137

They're building this big heavyweight

thing with a battery in it.

138

And solar and RF link. Right.

139

So probably ubiquity

or something like that.

140

And by the time you had you counted up

just the cost of parts, the bomb,

141

the bill of materials

was like $1,500 to read a water meter.

142

And they were using breakout boards

and fly wires and jumpers.

143

And the first long conversation

I had with over was,

144

you're doing this wrong,

you'll never succeed doing this.

145

And he told me, because, well, I tried

15 years ago or 20 years ago and I failed.

146

I said, yeah, and this will fail to.

147

And because it's too expensive

by an order of magnitude,

148

I said,

you know, you're thinking about this.

149

I mean, you know, here he is

very successful deploying now wireless ISP

150

high speed, low latency,

a lot of bandwidth,

151

a very, very high quality

SLA service level agreement.

152

Yeah.

153

And I'm saying, yeah,

you don't need that for what we're doing.

154

Let's go. The other way.

155

You only need to take a measurement

156

maybe every two hours

or maybe when it changes or.

157

Right.

158

So let's go that way and that.

159

First there was some sunk cost going on.

160

So here we are talking to you.

161

This is a business conversation

fundamentally

162

although we're going to nerd out.

163

And so number one sunk cost.

164

Yeah we've already invested in this thing.

165

We already had done it at that.

166

So anyway after a while

167

a few months go by

and I did lend them a hand

168

and did some engineering for them

to try to improve this comm box,

169

and in particular a controller

based on an Esp32 to do all the real time

170

things the comm box had to do

without using any real power,

171

and then the comm box could wake up

and interrogate the Esp32 and

172

and that that was successful,

but it was still not the right thing

173

at some point over says to me, well,

by the way, I would point out

174

Valley Internet, they deploy

a lot of RF point to point links,

175

you know, broadband

highly familiar with 900MHz networking

176

and how bad it was,

how difficult it was unreliable.

177

It was how expensive.

178

Everything about it was a turn off.

179

So when I said 900MHz,

you can just seal it.

180

We're, you know, shut down.

181

And I said,

no, no, no, no, no, it's better than that.

182

So anyway, the 0 to 1, I built

183

a simple mode out of or,

you know, end node by remote.

184

I built one out of that kind of stuff.

185

And for folks who are watching two things,

you know, for reviews mode,

186

which I haven't heard before,

187

maybe it's just my stunning ignorance

once again surfacing.

188

I would normally call it a sensor or

an end node, but you guys call it a mode.

189

Is there some terminology

that I'm missing, or is that

190

that's actually coming from the

the LoRaWAN, a LoRa Alliance terminology?

191

That's where I actually learned that from.

192

So back in the official again,

193

my ignorance can be stunning

in the official stuff,

194

which will put you right to sleep

if you read it. Right.

195

Yeah, maybe that's my excuse.

196

Certainly got to read some specs

once in a while.

197

And that's where I got that from in. Okay.

198

But the only ones, it's an end

node. It's not the sensor itself.

199

It's not the thing

actually taking the measurement.

200

Oh it's the radio.

201

That's that's sending and receiving okay.

202

It's it's the it's the radio

and the radio has the migrated computer

203

that does that.

204

So anyway, I built one

and we bought a micro ticket.

205

You know we play nine

you know which is there.

206

You know I don't have one.

207

Do I have one here somewhere.

208

But I'm not going to leave her out.

209

Yeah. It's a, it's a podcast after all.

210

And we put one of those up on a mountain

211

peak near here where they have

one of the Valley Internet radio sites.

212

Really easy to do.

213

Connect it up.

214

There's power,

there's there's Ethernet boom.

215

And I'm about 10

216

or 11 miles as a crow flies from that.

217

And not particularly friendly terrain

either.

218

All the way.

219

It's on a high peak, but there's still

some terrain that might be an issue.

220

And I went down there,

take a little parking lot near

221

here,

and I fired up this thing and connected,

222

and I called up over and said,

hey, look, I'm connected.

223

And he said, oh crap.

224

Oh, right. This was this was too good.

225

But it meant that now

226

he really did have to make this big pivot

that we've been trying to avoid.

227

And the next thing that we discussed

was battery life.

228

And I knew

229

and I was confident and comfortable

with how the battery life would work.

230

He was not.

231

We started with solar charged lithium

232

ion cells, you know,

initially with pouch cells, although.

233

So cylindrical cells are a better choice,

you know, overall.

234

Yeah. He says it's not going to last.

235

And I said, let's just do an experiment.

236

And I, I said, I'll just build a test

bench.

237

A test bench that behaves

just like a real mode

238

with a real 420 pressure sensor.

239

The way we measure the tank

was with the barometric sensor

240

that lives at the bottom of the tank,

so just the pressure of the water,

241

you can turn from that, that and the

the dimensions of the tank.

242

So I built one in

and I said it's got to be realistic.

243

And so I made it

take a measurement periodically.

244

And like every I don't know, five minutes

because I implemented report on change.

245

This was a completely functioning sensor.

246

And in mode your mode,

this was the whole whole kit and caboodle.

247

And I charge the battery up and I put it

in a shoe box and I put it in a closet.

248

And every week

we'd have this, this meeting

249

where we talk about, oh, yeah,

the battery is it, you know, 92.2%.

250

And oh, now it's it, you know, 89%.

251

And after about five weeks

he said, this is going to work.

252

Let's go.

253

Can you just build one more of those

and I'll try it out on my tank up here

254

and I'll right there.

255

We've gone from 1 to 2.

256

We've gone to our first deployed sensor

it right there.

257

I just made a mistake.

258

What did I do?

259

I took the science experiment in the lab

and I deployed it to the field.

260

You got to be careful with that.

261

And so it happened after that,

of course, is what we deployed ten more.

262

And now we actually had ten more science

experiments.

263

Okay.

264

And we did a better job of building them

and making them rugged and

265

and George and the team, you know,

they were out finding the right enclosures

266

and trying to make them adequately

weather resistant and 60 close.

267

Yeah. Yeah, yeah, yeah. Exactly.

268

So you know,

and there was a real learning curve there.

269

And but we fell into the number one

trap of a small company

270

that thinks that they're doing something

new and innovative is you figure

271

you got to build all this stuff yourself.

272

And so that's one of the first lessons

I have here is you probably don't.

273

And if you do, if you do build something

yourself, you have to really add value.

274

It has to be not just a little bit,

not just a little better.

275

You really need to be an order

of magnitude more functional or better

276

or easier to deploy, or something

that shows up in your bottom line.

277

So anyway, we were deploying these were

278

There were some issues with

279

the maturity of the RSA stack at the time.

280

And yeah, I decided

281

I wanted to get more control

over the firmware stack.

282

And right about there, their AK 3172,

which is based on the micro,

283

the single chip Stm32 well, x x,

284

which, you know, combines both the radio

285

and and the MCU on a single chip

reduces the cost.

286

Really a great, great thing.

And I loved it.

287

And it made it really easy

because they had a WIS

288

WIS block module that swapped right out.

289

And so we were able to

and I wrote the firmware for that.

290

I actually read the whole design

to a Gen two and probably

291

three weeks worth of engineering,

and I even tested it.

292

Good. That's other thing

that we'll talk about that. Yeah.

293

And so we ended up

deploying hundreds of those hundreds.

294

I think we get

we probably deployed easily 500 of these.

295

We had them in revenue operation.

296

We have most of them are still out there

still in revenue operation.

297

But almost as soon as I realized what I,

298

you know, created in the lab

and then helped it out into the world,

299

you know, I was just reading

in Scientific American about reverse DNA.

300

You know, about reverse DNA. Oh, no.

301

Tell me so if you reverse the order

of DNA, you'll get the same thing.

302

But if it's like, a bacteria,

your body won't recognize it.

303

None of your immunity

will work against it.

304

So I built this reverse DNA thing,

and I pushed it out of the lab.

305

And now everyone's infected,

and their bodies can't fight it.

306

Well, anyway,

I think that's a different podcast.

307

Yeah, I know, I know, okay,

but I thought that.

308

I think that it is.

309

It is interesting though

since Chris. Awesome. Yeah. Yeah.

310

So the point is, is I immediately said

we have to find vendors.

311

And that was a real challenge for us

early on was finding vendors

312

that had the right functionality and the

right features hit the right cost point.

313

You know,

I I'll name check a few miles site.

314

You know, one of our partner vendors

let me still have a lot of stuff.

315

So we still have a lot of my site

in the field.

316

They have really improved,

you know, from the initial generations.

317

We started with dramatically improved

the product

318

and I can't complain about that.

319

And then of course I have more.

320

Oh, more to go. Oh, yeah.

321

Drug, you know, you know, I found very

low cost, but the quality is good.

322

And I think you said

they had some excellent documentation.

323

Some of the best documentation

out there. Yes, absolutely.

324

They use wiki page for documentation,

325

and it's extensive and comprehensive

and understandable.

326

And they're great in in fact, that was an

this is a

327

you know, for the podcast people

I'm holding up or for the audio listeners,

328

I'm holding up in scientific 53,

which is a

329

a mode or node made by your gyno,

their bank.

330

They were like 60 or 70 bucks.

331

I don't know what the tariffs

have done to them since,

332

but they make the source code

available for that

333

and you can use that is your jumping off

point to build anything you want to build.

334

It's it uses a different chip.

335

It's like an Stm32.

336

Well but it's the Chinese version

with the SSR or something.

337

Yeah. The arm star.

338

Yeah. And it's our 66.

339

I want to like 66 or 1 right in the MCU.

340

And that is like a cortex M4.

341

But it was actually

when ARM China was going rogue.

342

It's actually the IP that they produced.

343

So but works great.

344

You know great battery life.

Very much a fan of that.

345

Something else that you and you can I'll

let you tell people what I'm holding up.

346

You know what that is now,

this little field tester.

347

Yeah.

348

Yeah, that looks like and 1007 and one,

which worked out great.

349

I got a couple of those

around here as well.

350

And would you say there's

something that folks

351

get wrong about LoRaWAN

pretty consistently.

352

You hit a little piece of it

when you're talking about over

353

coming in from his background,

I'm thinking, okay, 900 a no go.

354

Yeah.

355

Is there anything else you talk about

whether it's ADR or frame

356

counters or whatever,

power budgeting, etc.?

357

Oh, is so the biggest thing that people

get wrong about LoRaWAN consistently,

358

and it's a thinking cap or a mindset

frame of mind, a context.

359

And that is LoRaWAN is slow.

360

It's small amounts of data. Infrequently.

361

It is not, you know, again, it's not that

362

that wireless ISP, high bandwidth,

low latency,

363

the you can just ping stuff

and know that it's online all the time.

364

You know, with LoRaWAN you got to be okay

not knowing for two hours

365

whether it's a mode or an end

node is working or not.

366

But at the same time,

you tailor the operation of it.

367

Now to how quickly the thing

that you're reading changes the tank.

368

Most of the tanks, so initially are in

probably still for water tanks.

369

It's report on change once

you change a certain amount,

370

and I think we sample

every 2 or 5 minutes for that.

371

Now that we've gone to Modbus

based sensors,

372

the power consumption

is dramatically lower with 420 sensors.

373

That was a really big battery budget hit.

374

When you think about it,

a 420 sensor can use a quarter of a watt

375

just to take a reading.

376

It's it's drawing a quarter of a what?

377

Yeah, yeah.

378

It's,

you know, 20 milliamps at 12V minimum.

379

It's 240mW right there.

380

And it's actually more than that.

381

So every time you turn the sensor

on, you're a quarter of what it was

382

hard on the budget.

383

Whereas my Modbus considerably better

low voltage, low current.

384

You're talking

385

tens of milliwatts maybe and much quicker

measurement of the reading.

386

So once you've done that.

387

But what people get, you know, wrong is

they think they have to know all the time

388

what something is.

389

You really, really need to know

when it changes.

390

And for some tanks, water tanks

that are out in the middle of a field

391

somewhere, you might only care about it

once a day.

392

So you have to go into with a mindset

393

of how often do I need to know

and approach it that way?

394

That is number one area of Siem.

395

Are people saying, well,

I just installed it.

396

I'm not getting readings yet.

So was it changed yet?

397

No, I would actually say that's the first

kind of structural thing you have to do.

398

The other thing

399

too, is kind of a mistake

we almost made, and I, I favor simplicity.

400

You'll figure that out about me

at some point early on.

401

Or if you said, well,

how do we do firmware updates?

402

I said, you know what

403

firmware is a component

soldered on to the printed circuit board?

404

That's our model and goes, well, well,

how does that work?

405

You know what?

406

If we wanted it,

407

I said most of the features

408

and most of the value we had

is in the network itself, in the back end.

409

Right?

410

We just have a sensor out in the field

that's getting data and, you know,

411

whatever rate.

412

Sure, there are parameters

we like to change, like how often

413

you take the reading

or what the threshold of change is.

414

So those things ensure and you can do down

links to to tune those in the field.

415

That's fine.

416

But but overall

the firmware tested a little more upfront.

417

And after that most of the new features

and most of the innovation

418

you're going to do

will be in your back end.

419

And what you do with the data.

420

That's been our approach.

421

So don't feel let's make it stable

and then let it rip.

422

Right now

there is a firmware update over the air

423

spec for LoRaWAN, which, if I understand

correctly, is a multicast

424

type approach is kind of like the way

cable boxes or satellite

425

box is used to get firmware

426

updates a little bit

at a time over a period of time.

427

Right? And sure, we've not deployed that.

428

It actually requires a a fancier network

than we've deployed.

429

I haven't even talked about our network

yet. That's okay.

430

I mean, you guys are doing

some really cool stuff

431

with updates and onboarding

and all the rest of it up for is

432

is telling me, right, right.

433

So that's another podcast you can go into.

434

I can tell you all the depth about that

you'd like, but the biggest

435

or the biggest adoption problem we've had

436

is it's not you or it's not me

that are installing these things.

437

Most of the time it's somebody who is more

438

accustomed

to installing a pump motor, right.

439

And getting the red wires where they go

in the black wire in the right way.

440

Right.

441

And not that they're not smart,

they're just not trained.

442

Different area of expertise.

443

Yeah, absolutely.

444

I mean, I don't know the first thing

about drilling a well, but

445

or maybe I know the first thing now,

but I didn't got to start digging.

446

Yeah. Yeah.

447

Well you have to drill at home

when you have to put out the anyways.

448

Yeah. Maybe have drill another.

449

Well this is sensor too.

450

Anyway, there

there's a lot of fun stuff I've learned.

451

But the point being, you know,

we don't have to we, we found

452

we could not require our,

our field partners,

453

our channel to actually know

454

all the technology that so many more.

455

So you've configured a lot of things.

456

You open the book up

and it says set this to that.

457

Well, what's your uplink to, you know,

and and here's your Debbie.

458

Why am I right.

459

And put in the, you know, whatever lens

and you know, your network key and

460

and that was an app key and that one needs

an API and this one needs a join you.

461

It's a it is like a

462

right.

463

Like please

could you just make this simple?

464

It was crucial

to take that out of the loop

465

and make it where

someone could just with a phone app,

466

but the phone app, ideally

with no connectivity?

467

No, because a lot of our

I mean, that's the point of LoRaWAN.

468

You don't have any, you know,

cellular connectivity where you are,

469

although that may get better with,

you know, the Starlink non-terrestrial.

470

But anyway, so that's been the big,

471

big push to drive adoption,

make it possible to be adopted.

472

And that came through super strongly.

473

The for conversation hasn't gone public

at the time that you and I recording this,

474

but it comes through really strongly

for folks who have listened to it

475

that he's just hellbent on

making sure this is super easy.

476

He calls it friendly flying,

but just making it make it easy.

477

We have to trademark that as well.

478

Yeah, yeah.

479

You got you got more important things to

to protect.

480

Yeah. Let's see. Let's wrap this thing up.

481

You've been doing doing this

for I've been in game for a long time.

482

If someone, some young pup is coming along

483

and getting into LoRaWAN,

what advice might you give them?

484

Well, that's a good question.

485

And it was probably more

486

the generic advice for almost any field,

almost any path you go into.

487

Do your homework, take your time to learn,

learn who your customers are,

488

learn what your application is.

489

It's easy to come in thinking what it is

490

and driving aggressively toward a goal

that you have in your head,

491

and then find out maybe

it's not the same one your customers have.

492

I mean, I think we experienced that

a little bit in this case.

493

So so number one, take the time to get

to know your market, your customers.

494

You know,

I don't want to be careful about this.

495

Don't be afraid to fail

but don't set out to fail.

496

A friend of mine, Jeff Rothschild,

I don't know if you know him, Jeff.

497

You and I worked together decades ago,

and he went on to found some companies

498

and did Veritas among them,

and did very well

499

and went on to

he was the adult supervision on a

500

a round of funding to Facebook because

he was part of Excel partners at the time.

501

And so he had Jeff ends up Zuckerberg

hit it off,

502

and Jeff ended up retiring from Facebook.

503

A second retiring second time,

this time from Facebook.

504

And Jeff will tell you, you know, as

a serial entrepreneur himself, he'll say,

505

don't be afraid to fail and fail fast.

506

So and I would I pass that advice along

and he's been way more successful.

507

Am I have that?

508

Don't be afraid to fail, but do your best

509

not to to start with and pay attention.

510

Be honest with yourself

that this isn't working.

511

I'm not achieving these goals.

512

Yeah, the sunk costs

don't just keep if I just work harder,

513

if I just work longer,

or if I just do more of the same thing,

514

that's not working well,

I'm sure that'll work out right.

515

So you have to be honest

with yourself on that.

516

I thought so, and that's in any business,

you know, it will color

517

when we've been successful so far,

we'd like to be a lot more successful.

518

And that's exactly the way

we approached it.

519

Ripping.

520

Well, Dana, I appreciate you taking time

on a Friday afternoon.

521

I know you had to clear this

with the higher authority.

522

So thank you for,

making time for wrapping.

523

She's gone now anyway.

524

Yeah. My my pleasure.

525

That's it for

this episode of The Business of LoRaWAN.

526

If you want to go deeper

and actually deploy devices,

527

the MeteoScientific

console is the fastest way to do that.

528

And honestly, it's

also the best way to support the show.

529

When you use the console, you're not just

listening, you're participating

530

in the same real world LoRaWAN work

we talk about here every week.

531

You can get started with the free trial

at MeteoScientific.com.

532

Huge thanks to the sponsor of this show,

the Helium Foundation,

533

for supporting open LoRaWAN

infrastructure worldwide.

534

Check them out at Helium.Foundation

and if the show has been useful.

535

A quick rating or review on Apple Podcasts

or wherever you listen.

536

This really helps

537

people find it and helps the show grow

so we can help more people.

538

I'm Nik Hawks with Meteo Scientific.

539

I'll catch you on the next episode.