Concept to Launch in 18 Months: Building Hardware, Software and AI at Startup Speed with Ben Brown, CTO of Canopy

[00:00:00] [SPEAKER_01]: Welcome to the Convergence Podcast. I'm your host Ashok Sivanand.

[00:00:06] [SPEAKER_00]: It's not just what they say, it's what you see. They want you to disable a notification,

[00:00:12] [SPEAKER_00]: but what you see is that their workflow doesn't necessitate them to constantly go in and turn

[00:00:17] [SPEAKER_01]: it on and off throughout the day. On this show, we'll deconstruct the best practices, principles

[00:00:23] [SPEAKER_01]: and the underlying philosophies behind the most engaged product teams who ship the most

[00:00:28] [SPEAKER_01]: successful products. This is what teams are made of. On this episode, I am joined by my good friend Ben

[00:00:41] [SPEAKER_01]: Brown, the Chief Technology Officer at Canopy Security. Now you might have seen Canopy's

[00:00:48] [SPEAKER_01]: devices floating about. They help secure the contents of your vehicle, specifically the

[00:00:54] [SPEAKER_01]: often-time precious cargo that sits in the beds of our pickup trucks. This is a super huge problem

[00:01:00] [SPEAKER_01]: for people like trades folks especially who carry a combination of expensive tools, equipment and

[00:01:08] [SPEAKER_01]: materials in the back of their trucks as they commute from job to job. And having that cargo

[00:01:14] [SPEAKER_01]: lost, stolen or damaged can be a real disrupter to their businesses. We'll get to hear from

[00:01:21] [SPEAKER_01]: Ben some of the stories that inspired him to double down on this venture. With respect to Canopy

[00:01:28] [SPEAKER_01]: as a company, we've seen a number of ways that large companies and enterprises stay relevant

[00:01:35] [SPEAKER_01]: by harnessing the trust in their brand and the span of their distribution channels to put

[00:01:41] [SPEAKER_01]: more innovations and products in front of their customers. Things from corporate incubators and

[00:01:48] [SPEAKER_01]: accelerators through to mergers and acquisitions and other forms of partnerships. In Canopy's case,

[00:01:55] [SPEAKER_01]: this was an idea that was born from within Ford Motor Company whose F-Series pickup trucks

[00:02:01] [SPEAKER_01]: continue to be the largest selling pickups in the country. So you can imagine they have a lot

[00:02:07] [SPEAKER_01]: of insights that come from their customers. And they spun Canopy out as a separate entity

[00:02:12] [SPEAKER_01]: for a variety of reasons. One of which I'd be willing to bet is that this team needed to go from

[00:02:18] [SPEAKER_01]: idea to launch within 18 months. That might seem like a long time. I think it's particularly impressive

[00:02:25] [SPEAKER_01]: when you take into consideration that this product combines hardware, software and AI ML models

[00:02:33] [SPEAKER_01]: that sit on the device. Not to mention the fact that they need to comply with really stringent

[00:02:39] [SPEAKER_01]: automotive grade and FCC requirements. About Ben, their CTO and our guest, he was a key part of

[00:02:47] [SPEAKER_01]: spinning Canopy out of Ford. And prior to that, he actually joined Ford when he was still in high

[00:02:54] [SPEAKER_01]: school. And we'll get to hear from him in the variety of roles that he got to play there.

[00:02:59] [SPEAKER_01]: I have personally got to work with Ben and I find him to be an extremely curious and thoughtful

[00:03:04] [SPEAKER_01]: leader. He can switch context between casting a vision, getting his team excited, engaged and

[00:03:13] [SPEAKER_01]: executing against that vision while also being able to roll up his sleeves and write debug code

[00:03:19] [SPEAKER_01]: and anything else that the team needs to ship product and make progress. This gives him a super

[00:03:25] [SPEAKER_01]: unique perspective both at a macro and micro level. And you all know that I'm really curious

[00:03:32] [SPEAKER_01]: about shipping hardware, software, IoT products. And we get to hear a lot of Ben's stories and

[00:03:40] [SPEAKER_01]: insights from that very unique perspective that I believe he has. So I'm really excited for you

[00:03:44] [SPEAKER_01]: all to check this out. On the episode, we'll get to hear about the different approaches

[00:03:50] [SPEAKER_01]: that his team at Canopy took to validate and test the various assumptions that they had

[00:03:57] [SPEAKER_01]: as they were progressing towards launch. This included everything from technical integrations

[00:04:03] [SPEAKER_01]: through to customer assumptions around the depth of the problem through to onboarding flows.

[00:04:09] [SPEAKER_01]: And I think it's really interesting to hear the approaches that they took.

[00:04:13] [SPEAKER_01]: We also get to hear about the lessons from organizing and reorganizing this multi-stack

[00:04:18] [SPEAKER_01]: team that spans hardware, firmware, AI ML, cloud, web and mobile. And the lessons learned

[00:04:26] [SPEAKER_01]: across the iterations they took to get the most effective organization of their teams.

[00:04:32] [SPEAKER_01]: And finally, how his team is using generative AI to expand the scope of their product and ship

[00:04:39] [SPEAKER_01]: software faster for everyone that is curious about hardware and the integration of hardware

[00:04:46] [SPEAKER_01]: and software and road mapping and delivery associated. Make sure to also go back in our

[00:04:52] [SPEAKER_01]: archives and check out the episodes that we did with Ben Foster, particularly the second of those

[00:04:57] [SPEAKER_01]: two episodes where we get to hear a lot of Ben's stories about what it took to ship an IoT wearable

[00:05:05] [SPEAKER_01]: called WOOP. Subscribe to the podcast to get future episodes as soon as they're published.

[00:05:10] [SPEAKER_01]: If you find this helpful, give the podcast a five-star rating on your podcast app

[00:05:14] [SPEAKER_01]: or hit that like button on YouTube. Without further ado, here is Ben. Enjoy the episode.

[00:05:26] [SPEAKER_00]: I actually never thought of myself as being in a CTO role from the onset. I started out at Ford

[00:05:33] [SPEAKER_00]: some number of years ago. I won't give the actual years and age myself, but

[00:05:38] [SPEAKER_00]: I started out when I was just out of high school actually. I started with Ford Motor and

[00:05:42] [SPEAKER_00]: the Deathside Services role. I had been working at a gym at the time of all places and had a good

[00:05:48] [SPEAKER_00]: friend that worked there, his partner at the time worked at Ford and she came and said,

[00:05:56] [SPEAKER_00]: like, hey, there's this role at Ford Motor Company. You seem to be a techie interested in

[00:06:00] [SPEAKER_00]: computers you should apply for. I'm thinking no way am I going to get hired at big blue Ford

[00:06:05] [SPEAKER_00]: Motor Company. Seriously, give me your resume. We'll give it a shot. So I did and lo and

[00:06:10] [SPEAKER_00]: behold, I got an interview. I ended up getting interviewed by three managers in there and they

[00:06:15] [SPEAKER_00]: asked me a number of questions and I had basic knowledge of Windows 95 at the time

[00:06:21] [SPEAKER_00]: and some things that I had done in high school. During the interview, I thought it went okay,

[00:06:27] [SPEAKER_00]: you never really know, you second guess yourself after the interview, but at the end of it,

[00:06:33] [SPEAKER_00]: they asked me if I had any questions and I had researched for it at the time and I asked

[00:06:38] [SPEAKER_00]: them, they had made X millions or billions of dollars at the time and asked them what they

[00:06:44] [SPEAKER_00]: attributed Ford's success to at that time. The lead interviewer kind of leaned back and laughed

[00:06:51] [SPEAKER_00]: or remember vividly and said, guys, can you help me answer this question? Because he didn't have

[00:06:58] [SPEAKER_00]: an answer and they gave some answer and I don't recall what it was but after the fact I was

[00:07:04] [SPEAKER_00]: told that was actually what landed me the job was digging deep and wanting to understand and learn

[00:07:09] [SPEAKER_00]: more about the company and some of the details there that set me apart from the other candidates.

[00:07:15] [SPEAKER_00]: It's kind of been the journey and kind of the premise of my journey throughout my career.

[00:07:21] [SPEAKER_00]: So I've been in roles everything from starting in the desktop services to the hosting side

[00:07:28] [SPEAKER_00]: of things and servers and setting those up mergers and acquisitions. I've been in

[00:07:33] [SPEAKER_00]: a cybersecurity role for eight or nine years. I did some time around product management partnered

[00:07:40] [SPEAKER_00]: with Pivotal and started to learn the product management ways and the practices around

[00:07:47] [SPEAKER_00]: experimenting and trying things out and failing and trying again and learning and

[00:07:52] [SPEAKER_00]: contributing there. So ultimately ending up in what was at the time Ford X and is now Ford next

[00:08:01] [SPEAKER_00]: that spawned the canopy organization that I'm in. So along the journey never knew that I wanted

[00:08:07] [SPEAKER_00]: to be a CTO but actually really the first time that I've actually thought and reflected on the

[00:08:12] [SPEAKER_00]: CTO role is in canopy and it's not because just because of the tech but it's because of

[00:08:18] [SPEAKER_00]: the people that I've had a chance to work with throughout the journey and learning from them

[00:08:24] [SPEAKER_00]: along the way and supporting them and developing some technology that solves a real problem.

[00:08:31] [SPEAKER_01]: So a few things I want to dig into there. First something prompted you to do that research

[00:08:36] [SPEAKER_01]: about the company and ask me a little bit of a broader question than the role that

[00:08:41] [SPEAKER_01]: you were interviewing for. Can you remember what inspired that were you coached to do that

[00:08:45] [SPEAKER_00]: or did it come naturally? No, it just came naturally. Being a Metro Detroit native my

[00:08:53] [SPEAKER_00]: entire life, my dad was in the automotive industry for a competitor and I knew a lot about the

[00:08:58] [SPEAKER_00]: competition. I didn't know a lot about Ford and so I was just intrigued by it and I was interested

[00:09:05] [SPEAKER_00]: like why they were so profitable in comparison to the automaker that my dad had worked for at

[00:09:10] [SPEAKER_00]: the time. And so I just always wanting to learn more. I searched the web and came up with some

[00:09:16] [SPEAKER_00]: ideas that I was able to bounce off of those interviewers. Just the drive for knowledge

[00:09:21] [SPEAKER_00]: for me it wasn't any sort of prompting that I was given it's just kind of something at the

[00:09:26] [SPEAKER_01]: core of who I am. And so now that you've been at Ford for a number of years, what would

[00:09:32] [SPEAKER_01]: your answer be to an interview candidate if they asked you that about why is Ford so profitable?

[00:09:38] [SPEAKER_00]: Yeah, so I think my answer would be that Ford focuses on the needs of their customers and deeply

[00:09:46] [SPEAKER_00]: understanding what the problems are and the jobs that they're doing and what needs to be solved.

[00:09:52] [SPEAKER_00]: Instead of just rolling something out with the hope that it actually solves a need, there's

[00:09:57] [SPEAKER_00]: actually a level of validation that happens. And it's because of that validation

[00:10:01] [SPEAKER_00]: that the customer see the value in what's delivered in their products and services.

[00:10:06] [SPEAKER_00]: Which then ultimately leads to a profitable company.

[00:10:11] [SPEAKER_01]: So Canopy was spun out of Ford and talk about that journey. I know a lot of companies try to

[00:10:19] [SPEAKER_01]: spin out startups. There's a debate about should we do it in-house? Should we spin it out?

[00:10:24] [SPEAKER_01]: Should we just acquire a company that has nothing to do with the parent company? There's

[00:10:28] [SPEAKER_01]: a lot of options to go down with respect to corporate innovation and keeping your edge.

[00:10:33] [SPEAKER_01]: And I think Canopy's got a unique journey there that I'd love to hear about.

[00:10:37] [SPEAKER_01]: Actually, you know what? For the audience, talk about what Canopy is. Of course,

[00:10:40] [SPEAKER_00]: I know what it is pretty well. Yeah, so Canopy's actually core audience is

[00:10:44] [SPEAKER_00]: pickup truck owners. So we have a product called the Pickup Cam. It's actually developed to

[00:10:49] [SPEAKER_00]: provide security around the things that customers care about in their truck bed.

[00:10:53] [SPEAKER_00]: It provides peace of mind and enables the detection and turns of theft of items that

[00:10:59] [SPEAKER_00]: are within the truck bed. It's actually interesting you're talking about kind of the genesis of Canopy.

[00:11:06] [SPEAKER_00]: I remember very early on four or so years ago, talking with some folks about this opportunity,

[00:11:13] [SPEAKER_00]: this idea and sitting there thinking like, is this really a problem for customers? Is this

[00:11:19] [SPEAKER_00]: really something people worry about? There's insurance that exists for this. This certainly

[00:11:23] [SPEAKER_00]: doesn't happen all that often. I don't really hear about it much when I talk to friends and

[00:11:28] [SPEAKER_00]: coworkers. And so I really question like, is this really a path that makes a lot of sense that

[00:11:35] [SPEAKER_00]: provides value to people and serve a need? I remember very clearly that that changed for me

[00:11:41] [SPEAKER_00]: when I sat down and I listened to an interview with a tradesperson. And that tradesperson was

[00:11:48] [SPEAKER_00]: the sole owner of a company that they purchased all the tools to do the work that they did

[00:11:55] [SPEAKER_00]: day in and day out. And we were talking to them about the issues around theft that they had

[00:12:04] [SPEAKER_00]: experienced. This individual had actually lost all of their tools a few weeks prior,

[00:12:10] [SPEAKER_00]: completely stolen. And it didn't impact them just because the tools were stolen because

[00:12:15] [SPEAKER_00]: it also impacted their livelihood because they didn't have the tool. They weren't able to

[00:12:20] [SPEAKER_00]: actually go out and do the jobs that they had bid and signed up for and complete those.

[00:12:26] [SPEAKER_00]: He had gone and pawned his fiance's ring at the time to go and purchase new tools because he

[00:12:34] [SPEAKER_00]: wasn't insured. He didn't have the cash to go out and buy it. And he actually ended up having the

[00:12:40] [SPEAKER_00]: tool stolen a second time and lost there. And here I am questioning that this was a problem

[00:12:46] [SPEAKER_00]: and this guy was in tears sitting there in an interview about how deeply this had impacted him.

[00:12:52] [SPEAKER_00]: And so it was at that moment that I understood that, yes, this is actually a real problem

[00:12:57] [SPEAKER_00]: and it impacts people beyond just the loss of the tools. It impacts their livelihood and the

[00:13:02] [SPEAKER_00]: livelihoods of certainly the people that they support. That was when I knew this was the

[00:13:06] [SPEAKER_00]: mission that I wanted to be on and be a part of with the team to bring a technology

[00:13:11] [SPEAKER_00]: solution to this problem that went beyond the traditional locks that folks had on their cargo.

[00:13:21] [SPEAKER_01]: There's such an important point there with respect to having the technology team,

[00:13:26] [SPEAKER_01]: not just the product team and the marketing teams really interface with the customers.

[00:13:31] [SPEAKER_01]: And it's not always that as an engineering team we can go out and sit in the interviews,

[00:13:36] [SPEAKER_01]: but today with recording technology and everything else that's around there's nothing stopping

[00:13:42] [SPEAKER_01]: everyone from listening to the customer pain point and multiples of these interviews.

[00:13:47] [SPEAKER_01]: And then no doubt it gets you engaged, gets your team all fired up and then I'm sure

[00:13:53] [SPEAKER_01]: helps to better architecture and other decisions downstream as well once you have a better sense

[00:13:59] [SPEAKER_01]: of what's certain and what's uncertain and our understanding of the problem.

[00:14:03] [SPEAKER_01]: So what was the process like there with respect to that customer research?

[00:14:09] [SPEAKER_01]: How did that get initiated? Is that something typical that happened there?

[00:14:13] [SPEAKER_00]: Yeah, so we had a research arm within the organization that actually set up them and

[00:14:19] [SPEAKER_00]: there are multiple ways that we conducted interviews. We had surveys that we conducted

[00:14:23] [SPEAKER_00]: that we sent out. We actually reached out to customers and spent time with them in the

[00:14:27] [SPEAKER_00]: field. We had them come into our offices, we had conversations with them and what I've found over

[00:14:33] [SPEAKER_00]: the course and what I like most about understanding the customer more deeply is that it's not just

[00:14:39] [SPEAKER_00]: what they say, it's what you see. So I found the most value in actually going out and sitting

[00:14:45] [SPEAKER_00]: side by side with customers and seeing the problem because they'll tell you one thing

[00:14:48] [SPEAKER_00]: around what they believe that the problem is but you'll see something vastly different

[00:14:53] [SPEAKER_00]: when you're sitting by them in the field. An example is they might say that

[00:14:58] [SPEAKER_00]: they want you to disable a notification, a simple toggle to turn off a notification because it's

[00:15:05] [SPEAKER_00]: disruptive to them but what you see is that their workflow doesn't necessitate them to constantly

[00:15:10] [SPEAKER_00]: go in and turn it on and off throughout the day and a solution, technology solution could be something

[00:15:15] [SPEAKER_00]: like automated notification disabling or smart notification disabling so that it doesn't interrupt

[00:15:22] [SPEAKER_00]: the flow of their day and only notifies them when the notification is important to them.

[00:15:25] [SPEAKER_00]: So I think it's critically important for the full balance team of product managers,

[00:15:30] [SPEAKER_00]: designers and engineers to all sit alongside with the customer and deeply understand how

[00:15:35] [SPEAKER_00]: they operate and the problems that they see on a daily basis. And you started talking about

[00:15:40] [SPEAKER_01]: the team there. I imagine there's a mobile app, there's fun stuff going on in the cloud,

[00:15:46] [SPEAKER_01]: probably machine learning, there's hardware. This is a lot of components to a team.

[00:15:54] [SPEAKER_01]: How is the team organized today and was there some iterations that you took to get there?

[00:16:00] [SPEAKER_00]: Yeah, so definitely iterations. There continues to be iterations of the team.

[00:16:05] [SPEAKER_00]: What's important to me is that the work dictates the team structure, not the other way around

[00:16:12] [SPEAKER_00]: and it's something that I've learned time and time again. It's very similar to the tools

[00:16:16] [SPEAKER_00]: that we employ making sure that we use the right tools for what we're trying to achieve

[00:16:20] [SPEAKER_00]: and not try and pigeonhole things into the tools that we're kind of provided.

[00:16:26] [SPEAKER_00]: The team started out pretty small at the beginning and we've grown over time to roughly

[00:16:33] [SPEAKER_00]: 70, 80 or so people completely. And we have managed to stay relatively flat

[00:16:41] [SPEAKER_00]: in our organizational structure whereby individuals have access to

[00:16:45] [SPEAKER_00]: pretty much anyone including myself at any given time. I'm a big advocate of having

[00:16:51] [SPEAKER_00]: open office hours so that folks can ask me questions at any time and I make myself.

[00:17:00] [SPEAKER_01]: Fostering an engaged product organization and aligning them with the principles around lean,

[00:17:06] [SPEAKER_01]: human-centered design and agile will more than likely lead to successful business

[00:17:11] [SPEAKER_01]: outcomes for your organization. But getting started or getting unblocked can be hard.

[00:17:16] [SPEAKER_01]: This podcast is brought to you by the player coaches over at Integral. They help ambitious

[00:17:22] [SPEAKER_01]: companies like you build amazing product teams and ship products in artificial intelligence,

[00:17:28] [SPEAKER_01]: cloud, web and mobile. Listeners to the podcast can head on over to integral.io

[00:17:35] [SPEAKER_01]: slash convergence and get a free product success lab. During this session, the integral team will

[00:17:42] [SPEAKER_01]: facilitate a problem-solving exercise that gives you clarity and confidence to solve a product,

[00:17:49] [SPEAKER_01]: design or engineering problem. That's integral dot io slash convergence. Now back to the show.

[00:18:01] [SPEAKER_00]: Is available as possible. We're structured in terms of the different components and it is

[00:18:06] [SPEAKER_00]: pretty complex that we have end to end with this product. So we've got the mobile application on

[00:18:12] [SPEAKER_00]: the front-end side which is React Native that the team is developing in our back-end is AWS for

[00:18:19] [SPEAKER_00]: our hosting services. And then certainly on the hardware side and hardware is definitely

[00:18:25] [SPEAKER_00]: hard. It's the toughest part that I've learned throughout all of this because of the constraints

[00:18:29] [SPEAKER_00]: there. But we've actually sourced and worked with a manufacturer to develop our hardware that is

[00:18:36] [SPEAKER_00]: you know comprised on the software side of AI components that actually do the computer vision

[00:18:41] [SPEAKER_00]: detection to identify and notify threats. And then the software kind of components are C++,

[00:18:48] [SPEAKER_00]: Kotlin, et cetera, the hardware side. So pretty complex to stitch together end to

[00:18:56] [SPEAKER_00]: end in a cohesive manner in the time with which we were provided to actually delivery

[00:19:02] [SPEAKER_01]: the canopy product. And then given that there's so many elements of or so many different areas

[00:19:09] [SPEAKER_01]: of the stack, did you have to change your interview process to incorporate the AI and

[00:19:16] [SPEAKER_00]: hardware? How did that go? At the core we try to employ a pairing process at the

[00:19:24] [SPEAKER_00]: tail end to understand how people function, how they operate in the software development scenario.

[00:19:33] [SPEAKER_00]: Most of my experiences had been historically on the software side and not much with the hardware

[00:19:37] [SPEAKER_00]: and AI bits. So there were slight modifications on like for instance the AI side because you're

[00:19:43] [SPEAKER_00]: not going to have somebody necessarily pairing on AI and hardware quite the way that you would

[00:19:49] [SPEAKER_00]: with software. Through it all what's important is that the individuals that we select are one

[00:19:57] [SPEAKER_00]: certainly competent but also culturally a good cultural fit for the organization because to me

[00:20:07] [SPEAKER_00]: if somebody is a good fit in the way that they operate work and fit with the team you can

[00:20:13] [SPEAKER_00]: teach the other stuff as long as they have kind of the starting points of those.

[00:20:17] [SPEAKER_01]: And you mentioned pairing in. Folks are not familiar with what that is. How would you describe it?

[00:20:23] [SPEAKER_00]: So we employ where it makes sense a pair of software developers that work together side by

[00:20:28] [SPEAKER_00]: side one drives one kind of watches and overseas the end result is that you have better quality

[00:20:34] [SPEAKER_00]: code that comes out of it as an effect so anytime that we have the opportunity to do paired

[00:20:40] [SPEAKER_01]: programming we take advantage of that. And you find now with tools like co-pilot

[00:20:47] [SPEAKER_01]: that maybe your teams are more familiar with that model of having someone else and in this

[00:20:55] [SPEAKER_01]: case a computer or AI talk to you about the code and adapt it faster. You notice that with

[00:21:00] [SPEAKER_01]: new folks who haven't paired before coming in and see a correlation with around co-pilot.

[00:21:05] [SPEAKER_00]: It's interesting that you mentioned that because there has been gravitation towards

[00:21:09] [SPEAKER_00]: the AI side of having a paired programmer versus a human. One of the things that we also kind of

[00:21:17] [SPEAKER_00]: grapple with with canopy is we were at the tail end of COVID when canopy came to being so everyone

[00:21:24] [SPEAKER_00]: was remote and remote has some of its challenges while we strove to be a remote first organization.

[00:21:31] [SPEAKER_00]: It did frankly it was difficult for us so things like co-pilot did provide opportunities for

[00:21:39] [SPEAKER_00]: folks to be able to quote pair but not necessarily with the human. It also allows people to pair at

[00:21:46] [SPEAKER_00]: different times that fit their working models and needs when maybe humans aren't available to partner

[00:21:54] [SPEAKER_01]: and pair in. You mentioned earlier that the work dictates the organization and I'm curious from

[00:22:01] [SPEAKER_01]: a multi-stack perspective did you have front end and back end folks working on similar pods

[00:22:08] [SPEAKER_01]: or did you have that kind of separated? How's that all organized especially with hardware and AI all in

[00:22:14] [SPEAKER_00]: play? Yeah so we actually started out in like single domain where we had everybody who worked on

[00:22:22] [SPEAKER_00]: platform together in one group and that worked for a while and then we pivoted to

[00:22:29] [SPEAKER_00]: pods and squads organizationally. That worked for a little while here and we're actually

[00:22:35] [SPEAKER_00]: in what I'll call like our third iteration of working together. What we found with the latest is that

[00:22:40] [SPEAKER_00]: while it did make sense to have for instance front end and you know platform folks working together

[00:22:47] [SPEAKER_00]: we didn't always have the appropriate level of coverage and backup so if folks took you know

[00:22:53] [SPEAKER_00]: vacations it becomes a little more difficult and as backlog started to you know prioritization

[00:23:01] [SPEAKER_00]: of backlog started to fill out like it just didn't make quite well and so the organization is actually

[00:23:07] [SPEAKER_00]: this past week started to look at what makes sense to move us forward against the objectives that we

[00:23:13] [SPEAKER_00]: have in front of us and I expect in the next couple weeks we'll you know come to a decision

[00:23:18] [SPEAKER_00]: of do we stay as we are or do we pivot to some new model that makes more sense? I look forward

[00:23:24] [SPEAKER_01]: to hearing how that goes. It's complex problem and the lead time for knowing whether it was

[00:23:29] [SPEAKER_01]: right or wrong decision can be a little bit longer than some other experiments that we get to do

[00:23:35] [SPEAKER_01]: where you can make decisions after observing for half a day or a couple of days kind of thing.

[00:23:40] [SPEAKER_01]: So not easy work here that y'all are doing and for a few different reasons and I think

[00:23:51] [SPEAKER_01]: one of the things you mentioned earlier that were that the partnerships that you had on play

[00:23:55] [SPEAKER_01]: really helped you get things out the door. Talk to me about that. Yeah so what's important

[00:24:02] [SPEAKER_00]: first and foremost is that we whomever you're partnering with has a level of appropriate

[00:24:08] [SPEAKER_00]: competency in the area. You can quickly teach them about the context of what you know you're

[00:24:15] [SPEAKER_00]: working in so I think where this actually worked well a good example for us is you know partnering

[00:24:20] [SPEAKER_00]: with IntraGul and we did some work particularly on our front end application brought in IntraGul team.

[00:24:28] [SPEAKER_00]: They had the right level of competency to hit the ground running and so it was very quick to ramp

[00:24:33] [SPEAKER_00]: them up you know I worked with you a show on that to try and you know bring that to

[00:24:40] [SPEAKER_00]: fruition. We were up against some very tight timings to deliver things we ran a couple of

[00:24:46] [SPEAKER_00]: different pilots where we had quick feedback loops and things we had to iterate on and so having

[00:24:50] [SPEAKER_00]: that partnership where we could you know staff up quickly bring folks in and then deliver and get

[00:24:56] [SPEAKER_00]: it out the door to meet some of the beta testing that we were doing really helped out

[00:25:01] [SPEAKER_00]: and is something that I think you know everyone should think about is having the relationship

[00:25:06] [SPEAKER_00]: built with some really strong partners within your space so that you can call on them

[00:25:11] [SPEAKER_01]: if and when the time is needed. Yeah for sure I mean I know that our team really at IntraGul

[00:25:19] [SPEAKER_01]: enjoyed working with your folks and just love the culture and also the meaningful product that

[00:25:24] [SPEAKER_01]: y'all were working on and speed is one of the things that tends to get our team pretty

[00:25:30] [SPEAKER_01]: excited over at IntraGul and I understand that y'all went from concept to launch in 18 months

[00:25:38] [SPEAKER_01]: which is for a software product maybe seems long for an automotive product seems very very fast

[00:25:46] [SPEAKER_01]: and I think for something that has software hardware and AI pretty fast in itself especially

[00:25:52] [SPEAKER_01]: when you do it with some folks on your team that I imagine are more familiar with three to five

[00:25:58] [SPEAKER_01]: year cycle times for launching your vehicles so how did you shake that all out and push through

[00:26:04] [SPEAKER_00]: to get something out in 18 months? Yeah I think the biggest challenge for us

[00:26:11] [SPEAKER_00]: what or at least for me was on the hardware side and understanding you know spoke to the

[00:26:15] [SPEAKER_00]: cycle times that the vehicles have and the levels of rigor that need to be put out put into an

[00:26:20] [SPEAKER_00]: automotive product that is delivered so the product that we you know put together on the

[00:26:26] [SPEAKER_00]: hardware side is consumer based and so it has a we look at as like a three year cycle time for

[00:26:34] [SPEAKER_00]: the hardware before we refresh and replace it so that allowed us to be a little more flexible in

[00:26:41] [SPEAKER_00]: terms of the components that we selected for it still very challenging I mentioned you know COVID

[00:26:49] [SPEAKER_00]: was tough for us as we were coming out we were saddled with trying to figure out

[00:26:52] [SPEAKER_00]: how to source some of the key components when we were looking at lead times for components of

[00:26:57] [SPEAKER_00]: 40 weeks in some cases sometimes even more and trying to figure out are these the right

[00:27:03] [SPEAKER_00]: components to put into the device so in some cases you know I might not have 100% confidence

[00:27:09] [SPEAKER_00]: that this is the right component I might be 70% there and trying to de-risk as much as possible

[00:27:14] [SPEAKER_00]: but I had to make a commitment call on sourcing the the components to go into the final device

[00:27:20] [SPEAKER_00]: very early on and a change to that could be you know catastrophic and so that was very stressful

[00:27:29] [SPEAKER_00]: the the way that we worked around that was just getting the you know the team

[00:27:33] [SPEAKER_00]: collectively working in the field to de-risk it as much as possible test things out experiment

[00:27:38] [SPEAKER_00]: there were many at nights I remember as an example you know in December of 2022

[00:27:46] [SPEAKER_00]: at the tail end when we were entering our one of our phase gates to actually lock down the

[00:27:52] [SPEAKER_00]: bill of materials I was in the field with some of the team out in California actually doing

[00:27:57] [SPEAKER_00]: testing in the evenings and weekends some nighttime testing to make sure that we had

[00:28:03] [SPEAKER_00]: the right illumination for the device to function properly but we were we were going

[00:28:08] [SPEAKER_00]: through just kind of crazy types of experimentation where we were moving things and taping them in

[00:28:16] [SPEAKER_00]: position to validate the appropriate angles and then developing 3d prints and testing it out

[00:28:23] [SPEAKER_00]: and placing components in those 3d pins at one point we were even someone on the team was using

[00:28:28] [SPEAKER_00]: Legos and pulling those together to kind of figure out what the positioning would be

[00:28:33] [SPEAKER_00]: because it's very tight timeline and the stakes were pretty high and so you know 18 months as

[00:28:37] [SPEAKER_00]: you mentioned for software it's not a big deal like folks tend to do that pretty regularly when

[00:28:44] [SPEAKER_00]: you talk about hardware and the level of commitment that you have to make very difficult not to mention

[00:28:49] [SPEAKER_00]: AI AI trying to do detection and setting up models and training them in a space that

[00:28:56] [SPEAKER_00]: the data set there wasn't a data set that was on the shelf that we could draw from we had to

[00:29:01] [SPEAKER_00]: out and actually collect samples employing actors to do you know specific scenarios and testing to

[00:29:08] [SPEAKER_00]: build up that test set so that we knew that we would reduce the number of false positives and

[00:29:13] [SPEAKER_00]: deliver a product that was catching things that it needed to but not distracting individuals with

[00:29:18] [SPEAKER_00]: false positives when it shouldn't so really really challenging and tight timeline but ultimately

[00:29:24] [SPEAKER_00]: happy with where it landed and we continue to iterate on it. That's awesome and I was curious

[00:29:31] [SPEAKER_01]: to know what experimentation looks like on the hardware side and I think you alluded to some

[00:29:36] [SPEAKER_01]: of that you're just taping stuff together you're using Legos 3d printing did you what was it like

[00:29:44] [SPEAKER_01]: having like a really big notable parent company when it came to that did that get you access to

[00:29:50] [SPEAKER_01]: folks that you wouldn't otherwise maybe had had if you were if you were a standalone startup

[00:29:58] [SPEAKER_01]: and then there are also some tradeoffs there that made things not as fast for you what's all that like

[00:30:04] [SPEAKER_01]: when when you have a big parent company while also being a standalone startup in a way.

[00:30:10] [SPEAKER_00]: Yeah I think we've actually been really fortunate in that regard I look at is taking the best of

[00:30:17] [SPEAKER_00]: both worlds that make sense for us so we definitely had the access needed to some of the best

[00:30:24] [SPEAKER_00]: and brightest folks in an automotive to understand how the vehicle functions some of the uniqueness

[00:30:29] [SPEAKER_00]: of vehicle designs how the different materials in the vehicle you know we might need to consider them

[00:30:36] [SPEAKER_00]: for things like electromagnetic interference and environmental concerns so it was great there

[00:30:41] [SPEAKER_00]: but then we also had the autonomy and flexibility to go and do things in a in a very agile and

[00:30:49] [SPEAKER_00]: fast fashion where we could you know put the Lego blocks together do things where we would

[00:30:55] [SPEAKER_00]: rapid prototype and have 3d prints in house with in in the canopy offices to actually

[00:31:00] [SPEAKER_00]: you know test things out and have different housings within hours you know versus weeks or

[00:31:08] [SPEAKER_00]: days or months so that was great the other benefit that we had with Fortis they have

[00:31:14] [SPEAKER_00]: they have a vast number of test you know settings and labs and tools there that we had access to so

[00:31:23] [SPEAKER_00]: when we needed to go and do some environmental testing that needed to be done in a controlled

[00:31:28] [SPEAKER_00]: environment we didn't have to go out and necessarily source it from a supplier and set

[00:31:33] [SPEAKER_00]: up a purchase order and go through all of that process we could call on for to say hey

[00:31:40] [SPEAKER_00]: is there a way that we can get some time in one of your labs to do this testing we need to

[00:31:44] [SPEAKER_00]: go through that to meet you know FCC requirements as an example so having that the ability to do

[00:31:51] [SPEAKER_00]: that at our fingertips was definitely beneficial and helped us move in the right way very quickly

[00:31:59] [SPEAKER_01]: certainly part of testing is running betas and and recruiting users that aren't from within

[00:32:07] [SPEAKER_01]: the building I know you all did a pretty big beta program and I'm curious to know how that came about

[00:32:14] [SPEAKER_01]: and also any big nuggets any things that you were like it's a really good thing we did the

[00:32:19] [SPEAKER_00]: beta sort of thing that happened yeah we actually we did a couple rounds of betas the pickup

[00:32:26] [SPEAKER_00]: cam was developed to be multi-make from the onset and so the beta customers that were recruited

[00:32:32] [SPEAKER_00]: to participate actually had all types of vehicles everything from you know the f-150s that you know

[00:32:40] [SPEAKER_00]: we expected and had experience with to Dodge Rams to Toyota pickup trucks to I mean we even had people

[00:32:47] [SPEAKER_00]: that wanted to join that were like I have a a Mustang or a Taurus that I want to put this device

[00:32:53] [SPEAKER_00]: in really it's targeted right now for for pickups but it gave us some insights to the desire

[00:32:59] [SPEAKER_00]: that folks had even beyond pickup trucks so that was an interesting aspect of it definitely was great

[00:33:05] [SPEAKER_00]: that we did we learned a lot of things around the way that in along the way that influenced the

[00:33:10] [SPEAKER_00]: product in a positive way one of the things I think that that stood out for me was we felt like

[00:33:16] [SPEAKER_00]: we knew that the install process was designed in the right way and it was very simple we were

[00:33:23] [SPEAKER_00]: close to it and in how we installed it every day in our experience one of the things that stood out

[00:33:31] [SPEAKER_00]: was it seemed very apparent to us our devices in two pieces there's a hub there's a sensing unit

[00:33:36] [SPEAKER_00]: the sensing unit goes in the back window of the truck the orientation of the sensing unit

[00:33:41] [SPEAKER_00]: it was clear to us that the the connection was to be at the top and routed along the

[00:33:46] [SPEAKER_00]: headliner though we had customers who actually flipped it the other way with the LED opposite

[00:33:52] [SPEAKER_00]: and so that allowed us to improve our onboarding experience through the app to make sure that it

[00:33:58] [SPEAKER_00]: was identified and to be able to improve the installation instructions for the customer that

[00:34:02] [SPEAKER_00]: it would be guiding them in a more appropriate way so that they were successful from from the

[00:34:08] [SPEAKER_00]: onset there so that was just one of the nuggets that we got out of it cannot understate or

[00:34:14] [SPEAKER_00]: overstate rather how important it is to go out and do this level of testing before you

[00:34:19] [SPEAKER_00]: actually take a product to production because you think you know everything you're definitely

[00:34:24] [SPEAKER_00]: learn a lot when you go to testing with real users and so what are you excited about next for

[00:34:31] [SPEAKER_00]: the product that you can share yeah so I'm excited about the advances in AI so I mentioned you

[00:34:39] [SPEAKER_00]: know all the data collection that we've been doing to understand you know how it needs to

[00:34:43] [SPEAKER_00]: operate in the field so that it doesn't provide false positives we've done a ton

[00:34:48] [SPEAKER_00]: in this regard and we have made considerable advances I won't tip our hand to everything

[00:34:52] [SPEAKER_00]: that we have there but I've been encouraged by the great work that the team's been doing and

[00:34:57] [SPEAKER_00]: the research and innovation there around ensuring that the right level of detection is provided

[00:35:03] [SPEAKER_00]: by the product and that we provide utmost assurance to the protection of customers

[00:35:09] [SPEAKER_00]: belongings within their truck so the AI space is still there's still a lot to do there

[00:35:14] [SPEAKER_00]: and that's the I think the bit that I'm most excited about at this time yeah I'm not going to

[00:35:20] [SPEAKER_01]: try and get you to divulge too much off your roadmap however I'm curious to know

[00:35:25] [SPEAKER_01]: with so you've got hardcore machine learning engineers on your team that can do computer

[00:35:32] [SPEAKER_01]: vision to detect things like threats or bad actors stealing things from pickup trucks right

[00:35:37] [SPEAKER_01]: there is also this new wave of generative AI now

[00:35:42] [SPEAKER_01]: LLMs diffusion models and things like that that also can maybe help the product certainly help

[00:35:48] [SPEAKER_01]: your development team we already talked about co-pilot so what are some things you're seeing

[00:35:52] [SPEAKER_01]: there what are some things that are surprising around around that or things that you're really

[00:35:58] [SPEAKER_00]: excited about one of the areas not just the product specific AI but we're getting a ton of

[00:36:05] [SPEAKER_00]: data to analyze on product side so customer feedback and research there's only so much that

[00:36:11] [SPEAKER_00]: human see and you know we tend to I believe we tend to miss things so I'm excited about leveraging

[00:36:19] [SPEAKER_00]: AI from the perspective of analyzing what the customer feedback is how we can actually

[00:36:25] [SPEAKER_00]: use that information to improve the product itself not just use it within the product

[00:36:31] [SPEAKER_00]: but actually the product improvements to head us in the right direction and focus

[00:36:35] [SPEAKER_00]: the team on the things that make the most sense right now so that we're prioritizing

[00:36:40] [SPEAKER_00]: that we're rigorously ruthlessly prioritizing the things that we need to get out to customers

[00:36:45] [SPEAKER_01]: to increase the value cool and then are there any tools or services that come to mind

[00:36:50] [SPEAKER_01]: that your team is using that you've been impressed by or the encouraged other folks to use

[00:36:55] [SPEAKER_00]: like copilot my team has actually been spending a fair bit of their time with chat gbt believe it or

[00:37:02] [SPEAKER_00]: not I know it's kind of like one of the the basic ones that everyone knows about but just

[00:37:07] [SPEAKER_00]: leveraging that for some day-to-day things has been been helpful for the team so they can focus

[00:37:11] [SPEAKER_00]: on some of the things that require more of their expert focus and takes away some of the

[00:37:17] [SPEAKER_00]: the repetitiveness of what they're doing as we're kind of coming to the close I'm curious

[00:37:23] [SPEAKER_01]: to know any final words of advice for folks who are working on spin out startups within big companies

[00:37:30] [SPEAKER_01]: or hardware software combination products other things that you may have learned along the way

[00:37:35] [SPEAKER_00]: that you may want to share as as wisdom probably the one thing that stuck with me most on this journey

[00:37:42] [SPEAKER_00]: is this notion you know everybody that that or many of the people that you talk to along

[00:37:49] [SPEAKER_00]: the way will be risk averse what's the risk of doing this the risk of doing this and always have

[00:37:54] [SPEAKER_00]: this worry about the risk of doing I've long thought about it in terms of the risk of not doing

[00:38:02] [SPEAKER_00]: and so when the advice I would give to folks is you know there's always the what if what if I did

[00:38:08] [SPEAKER_00]: this what if I you know took this this leap and made this decision do it there's always going

[00:38:15] [SPEAKER_00]: to be the risk of not doing and there's no better opportunity than now to make that leap you know it can

[00:38:20] [SPEAKER_00]: be daunting there can be some scary moments to it but I had those along the way but you know you

[00:38:27] [SPEAKER_00]: surround yourself with good folks good people build a great team just continue to learn and

[00:38:32] [SPEAKER_00]: that risk of of doing actually goes away in many cases you'll find and you'll actually realize

[00:38:38] [SPEAKER_00]: the risk of of of not doing is is you know the the more appropriate thing to focus on yeah I mean

[00:38:48] [SPEAKER_01]: it's such a subtle reframe because I've definitely tried to run into that brick wall with uh I don't

[00:38:54] [SPEAKER_01]: want to call them naysayers but the folks who are familiar with a different risk profile

[00:38:59] [SPEAKER_01]: and try to kind of make the argument using the opportunity and the risk of not doing

[00:39:06] [SPEAKER_01]: is very related to the opportunity but you're talking risk versus risk that way so you're speaking

[00:39:11] [SPEAKER_01]: their language and then where they are uh that is something that I had not thought of before and I'm

[00:39:15] [SPEAKER_01]: gonna I'm gonna steal that today this is something I like asking all the guests um and this doesn't

[00:39:21] [SPEAKER_01]: have to be at home uh it could be or this doesn't have to be at work it could be at home too

[00:39:25] [SPEAKER_01]: rather um what's a product recently or a service that you your team or your family got to

[00:39:31] [SPEAKER_01]: experience that just blew your socks off you're like hey these folks really get it

[00:39:35] [SPEAKER_00]: one that stands out for me recently is you know being a techie uh I do a lot of home automation and

[00:39:41] [SPEAKER_00]: I'd love to see data so that I can have uh more data you know to make uh data driven decisions

[00:39:48] [SPEAKER_00]: recently you changed homes and I purchased the SENSE IoT device that actually manages home energy

[00:39:55] [SPEAKER_00]: and the reason that that I really like that is you know a couple things it was dead simple to

[00:40:01] [SPEAKER_00]: board with the install instructions were really smooth really well done I thought I was up and running

[00:40:08] [SPEAKER_00]: in 15 minutes you know a little bit scary tying it into your electrical box um but you know was

[00:40:14] [SPEAKER_00]: able to manage it and it plugged in and it just started working right um it's gone and identified

[00:40:20] [SPEAKER_00]: signatures for the different devices in my home the refrigerator the air conditioner I didn't

[00:40:25] [SPEAKER_00]: need to program any of that which would have taken time for me to go and figure out and do

[00:40:29] [SPEAKER_00]: and research um I've really come to appreciate the value of time so anything that's a time saver is

[00:40:37] [SPEAKER_00]: really important to me but it is allowed me to understand where uh and what devices are using

[00:40:43] [SPEAKER_00]: power without my home and it integrates with home assistant which I'm a huge fan of for home

[00:40:47] [SPEAKER_00]: automation so I can see a full picture of you know however things operating in my home and make

[00:40:51] [SPEAKER_00]: data driven decisions to improve cost and actually you know help the environment by

[00:40:56] [SPEAKER_00]: reducing the amount of electricity that that I'm using so super cool with that one it's just

[00:41:00] [SPEAKER_00]: one of the ones that I've thought of as of late and it's called sense sense yeah cool well make

[00:41:08] [SPEAKER_01]: sure to look it up and put it in the show notes there it is uh quite amazing how unique the

[00:41:15] [SPEAKER_01]: signatures of each of our appliances and things like that are um I was at uh I was at a bootcamp

[00:41:23] [SPEAKER_01]: at MIT recently and got to see some data and discernment where folks are able to tell the

[00:41:29] [SPEAKER_01]: difference between one laptop and another we're looking at these signatures you can say yeah

[00:41:34] [SPEAKER_01]: that's my son whose laptop is on because he has a windows machine versus this is the mag signature

[00:41:40] [SPEAKER_01]: and um and yeah you didn't think that there was enough discernibility but I think the

[00:41:47] [SPEAKER_01]: impressive part is not just the sensing capability but also the AI models that are applied to the

[00:41:54] [SPEAKER_01]: data that are able to discern these things in a really smart way versus back in the day we would

[00:41:59] [SPEAKER_01]: be putting like a clamp meter or something on each of these and then figuring out how to route it

[00:42:03] [SPEAKER_01]: back to uh to that one central area in order to make the same decisions and they were only

[00:42:10] [SPEAKER_01]: viable for commercial use not for residential so now you know you've got a lot of other

[00:42:14] [SPEAKER_01]: options I'm curious though since you mentioned home automation um what are some other things that

[00:42:20] [SPEAKER_01]: you have nerded out with or got your hands dirty with at your house oh man a number of things so

[00:42:28] [SPEAKER_00]: I've actually um I've built a number of electronics I have an automated gate to access the driveway

[00:42:36] [SPEAKER_00]: that I actually uh built some circuit boards around and connected up to uh to home assistant

[00:42:43] [SPEAKER_00]: to be able to drive that and understand the state of where it's at um thermostats are connected

[00:42:49] [SPEAKER_00]: garage door openers are all connected smart plugs are all connected um so a number of things

[00:42:56] [SPEAKER_00]: all the appliances washers dryers refrigerators um it I find it very helpful because you know

[00:43:04] [SPEAKER_00]: if I leave the house and I'm traveling I know what's going on if I've left the washer running

[00:43:09] [SPEAKER_00]: I don't have to worry about is there a leak is there a problem I can easily check into it from

[00:43:15] [SPEAKER_00]: anywhere so it gives me a level of peace of mind um you actually said something interesting there too

[00:43:20] [SPEAKER_00]: about the signatures while it's completely awesome and understanding what device it is

[00:43:24] [SPEAKER_00]: there is a small part of me that is you know thinks about the privacy aspects of it and

[00:43:30] [SPEAKER_00]: that's probably a discussion a much bigger discussion for a different podcast but um for

[00:43:38] [SPEAKER_00]: right now I will say that that stuff is pretty cool and I do try to keep everything as local

[00:43:43] [SPEAKER_00]: as I can for devices where it's available so definitely nerding out and I continue to pick

[00:43:49] [SPEAKER_00]: things up the one that's uh on order right now is actually an automated lawn mower for the yard

[00:43:56] [SPEAKER_00]: so to be my second automated lawn mower to pick it up I'm looking at the husk of arna

[00:44:02] [SPEAKER_00]: automated lawn mower that is actually doesn't leverage boundary wire it does gps tracking and

[00:44:09] [SPEAKER_00]: you set it all up through the app super seamless super smooth and again just works and gives me

[00:44:14] [SPEAKER_00]: time back so another cool device that's on my list hey ben I know you're busy dude and appreciate

[00:44:20] [SPEAKER_01]: you making the time here um thanks a lot for coming on the show absolutely thank you for

[00:44:27] [SPEAKER_01]: having me on appreciate this has been awesome hope you enjoyed that recording with ben one thing

[00:44:40] [SPEAKER_01]: that I really loved was that anecdote he shared with us about his customer the tradesperson who

[00:44:47] [SPEAKER_01]: lost their tools and was in tears their anecdote on how they had to pawn their fiance's engagement

[00:44:54] [SPEAKER_01]: ring because of the problem that ben and his product teams were setting out to solve now

[00:45:00] [SPEAKER_01]: if you think this was really lucky I want to argue that there's a lot more that's within your control

[00:45:07] [SPEAKER_01]: to increase the likelihood of you two getting access to stories like this and all of the

[00:45:13] [SPEAKER_01]: downstream benefits that you can imagine right like you're gonna build better products for your

[00:45:18] [SPEAKER_01]: customers faster and with less you're going to have a much more delighted product team because

[00:45:27] [SPEAKER_01]: they are now very engaged with the customer and the problem they're solving you're gonna have

[00:45:33] [SPEAKER_01]: higher alignment higher engagement and likely higher employee retention because now you have a team of

[00:45:40] [SPEAKER_01]: missionaries versus mercenaries as they're called the other thing is that the architecture

[00:45:46] [SPEAKER_01]: of your products when they're informed by the customer obviously tend to be way more resilient

[00:45:51] [SPEAKER_01]: and less likely to need to be changed and the cost of change is way lower working back from

[00:45:57] [SPEAKER_01]: ultimately who you're solving this for and what you're solving it for the bone that I have to pick

[00:46:03] [SPEAKER_01]: though is that it's very often when we start engaging with a product team that we're working

[00:46:08] [SPEAKER_01]: with that we find that these customer interviews these talks aren't happening I think there

[00:46:14] [SPEAKER_01]: are usually two main reasons and I want you to ask yourself when was the last time

[00:46:19] [SPEAKER_01]: you got to hear that level of emotion from a customer you were serving about the problem

[00:46:25] [SPEAKER_01]: you were solving not about the solution that you're putting out there right was it last week

[00:46:29] [SPEAKER_01]: last month last quarter last year and what can we do about it now usually when we're helping

[00:46:35] [SPEAKER_01]: solve this problem and changing this team towards better product practices I've noticed two

[00:46:42] [SPEAKER_01]: broad reasons where there is a hesitation to do this more one of them is the culture of

[00:46:48] [SPEAKER_01]: the organization and the second one is the skills of the team the culture of the organization

[00:46:53] [SPEAKER_01]: especially a large companies can stem from the fact that everything number one is being optimized for

[00:46:59] [SPEAKER_01]: efficiency and scale and secondly there tend to be a lot of silos within the broader company

[00:47:06] [SPEAKER_01]: and as departments and sometimes have competing priorities now in order to solve that I can

[00:47:12] [SPEAKER_01]: think of two ways one of them is maybe less risky and takes a little bit longer

[00:47:16] [SPEAKER_01]: where you want to remember that the folks that are on the other silos within your org that know more

[00:47:23] [SPEAKER_01]: about your customer or more about the product are people too and building one-on-one relationships

[00:47:29] [SPEAKER_01]: requires time and effort but when you're there that trust and influence can be something that

[00:47:35] [SPEAKER_01]: you can harness to earn a favor with this person to go on a ride along with them and

[00:47:41] [SPEAKER_01]: do this customer interview if your organization historically doesn't allow that now if you're

[00:47:46] [SPEAKER_01]: less patient than that like maybe myself you might want to take a little bit more of a risky

[00:47:52] [SPEAKER_01]: approach and before I tell you the approach I want to remind you that as leaders it's really

[00:47:58] [SPEAKER_01]: important that we're constantly applying our judgment to do things where we don't know what's

[00:48:03] [SPEAKER_01]: going to happen and we may get our wrist slapped but that's why we're here and that's

[00:48:06] [SPEAKER_01]: why we're leading to help clarify that ambiguity and uncertainty for ourselves and our team right

[00:48:12] [SPEAKER_01]: and so I have on numerous times gotten to an automotive dealership unsanctioned conversations

[00:48:19] [SPEAKER_01]: with folks who work there folks who are visiting there I've done the same at grocery stores and

[00:48:25] [SPEAKER_01]: other places for B2C type applications and this has helped me personally gaining a lot more

[00:48:33] [SPEAKER_01]: engagement and attachment to the work that I'm doing but also in having much smarter conversations

[00:48:39] [SPEAKER_01]: where the secondary research the surveys the pretty decks from the consulting companies

[00:48:45] [SPEAKER_01]: compete with the insights that I've learned through primary research and can force a very

[00:48:50] [SPEAKER_01]: critical conversation early that avoids a lot of downstream risk so that's an other approach

[00:48:58] [SPEAKER_01]: that I think folks could really consider doing and risking your wrist being slapped

[00:49:04] [SPEAKER_01]: and in all fairness if you do it smartly the the implications are unlikely to be really high

[00:49:11] [SPEAKER_01]: compared to the fear of those ruffled feathers and and what can happen if something were to go

[00:49:19] [SPEAKER_01]: wrong. One book to check out is transformed by Marty Kagan where he's got a bunch of spicy takes

[00:49:27] [SPEAKER_01]: on feature factories and maybe companies that could do a lot better about initiating

[00:49:33] [SPEAKER_01]: this type of behavior at a cultural level the second thing I think is the skill set on the team

[00:49:40] [SPEAKER_01]: right and when you say go out and talk to your customers how good are you at doing the talk

[00:49:46] [SPEAKER_01]: and versus are you looking at survey data and now you know I don't want to hate on surveys

[00:49:54] [SPEAKER_01]: but I think there's a time and place and personally if a in-depth conversation hasn't

[00:50:00] [SPEAKER_01]: happened yet I would much rather go speak to five folks and make sure that they are within the persona

[00:50:08] [SPEAKER_01]: of the customer that we're serving spend some time understanding the various risks and assumptions

[00:50:14] [SPEAKER_01]: in our systems and ideas and prioritizing the really scary ones and then crafting an

[00:50:19] [SPEAKER_01]: interview working back from de-risking those assumptions I'd much rather have that conversation

[00:50:26] [SPEAKER_01]: with the right person with just five folks versus having data from surveys that has 10,000 data points

[00:50:34] [SPEAKER_01]: larger organizations tend to value volumes of data points versus granularity of data or efficiency

[00:50:40] [SPEAKER_01]: versus effectiveness so that might be what's happening at your org and see what you can do

[00:50:46] [SPEAKER_01]: to push boundaries on having those one-to-one discussions then if you want to further

[00:50:52] [SPEAKER_01]: de-risk things later and use a survey I promise you that those one-to-one interviews that you've

[00:50:58] [SPEAKER_01]: done are going to help you make much better surveys when you get to that point as well

[00:51:03] [SPEAKER_01]: in the case of a company like canopy that was building a new product in a new market segment

[00:51:08] [SPEAKER_01]: I think it's especially important to get in there and have those one-on-one conversations

[00:51:13] [SPEAKER_01]: versus relying purely on secondary research or on surveys if you're a little bit more curious

[00:51:20] [SPEAKER_01]: on how to craft those discussions check out a book called mom test by rob Fitzpatrick it's a

[00:51:27] [SPEAKER_01]: classic that I go back to even now maybe 10-12 years after I first read it in order to go back

[00:51:33] [SPEAKER_01]: and get some more insights or refreshers on how to have really great interviews

[00:51:38] [SPEAKER_01]: the third thing that I recommend you check out is an episode from our archives with Jim

[00:51:44] [SPEAKER_01]: Simpson from Blue Myra where he talks about different tactics that he puts in place to

[00:51:49] [SPEAKER_01]: drive discipline amongst his product folks and technology leaders to continuously be having

[00:51:55] [SPEAKER_01]: that conversation with the customer that leads to much better outcomes for their business

[00:52:02] [SPEAKER_01]: hope you enjoyed the episode and let me know what you think as always we will be back

[00:52:07] [SPEAKER_01]: on Tuesday with the next one and I will see you there

[00:52:16] [SPEAKER_01]: thank you for joining me on the convergence podcast today

[00:52:20] [SPEAKER_01]: subscribe to the convergence podcast on apple podcast spotify youtube or wherever you get your

[00:52:27] [SPEAKER_01]: content if you're listening and found this helpful please give us a five-star review

[00:52:32] [SPEAKER_01]: and if you're watching on youtube hit that like button and tell me what you think about

[00:52:36] [SPEAKER_01]: what you heard today