Chris Grouchy:

Hey, everyone, and welcome to the Legends of Retail Podcast presented by Convictional. We talk to leaders in retail and e-commerce so you can learn from them and take their insights back to your company. I'm your host, Chris Grouchy, and we're about to get into an episode with Roger Kirkness. Roger Kirkness is none other than my business partner, and we have an interesting conversation about his unconventional upbringing, the hard engineering problems in B2B trade, and decisions around using email only instead of Slack at his company, our company Convictional. I hope you enjoy. Roger, what's your favorite food?

‍

Roger Kirkness:

Chana masala.

‍

Chris Grouchy:

Chana masala. How do you make that?

‍

Roger Kirkness:

So I actually do know how to make it. You take onions, canned tomatoes, I think it's like jalapeno peppers, and then a bunch of different spices like garum and chili, and you basically soup them all together and then you put two cans of chickpeas in. And then once you're done that, you have to cook it down. So I learned what reducing means, which just means reducing the water. Yeah. That's like the only thing I can make.

‍

Chris Grouchy:

That's a pretty tasty thing to make. Not your typical first meal, that one, which is probably for most people, I don't know, spaghetti, or Kraft dinner, or something. Thanks for joining the podcast. We're going to have a conversation about your life experiences and Convictional. What we do and more specifically how we're approaching building the business and the culture. So hopefully that sounds good and we'll jump into your background. Roger, we started Convictional together four years ago now. I don't know where the time went, but it's four years. But before that, way early back, you were this outlier in, I guess, school in your childhood because you didn't go to school and then you went to high school for three years. So can you talk a little bit more about what unschooling is and why your parents felt that was the best path for you at the time?

‍

Roger Kirkness:

Yeah. My parents felt that in their careers as entrepreneurs that really the things that they learned that were most valuable to starting and running companies were not things that they were able to learn at school. Both of them have MBAs that took a pretty traditional path to understanding business. And as kids, they basically told us that school was going to be value neutral if we wanted to do things that were new and we had to have basically belief that were different. And that the way to form those beliefs and think critically was to not expose yourself too early to another way of thinking beyond the one that you naturally held. And so until we were 12, specifically until I was 12, we did no formal instruction at all. And the team will probably resent this a bit, but one of the touchstones of Convictional culture is weekly updates. And so essentially from the age of six, I had a laptop that I had to write weekly updates on. And so no matter what I did, whether it was play with rocks, or dinosaurs, or run around in the forest, I had to write on a weekly basis about what I learned from that experience. And that was what it was pretty much until I was 12.

‍

And around when I was 12, I decided that I had more to learn from people than I had to learn from my own individual endeavors and decided to go to school. I spent one year in a private school kind of going hard academically and found that it took me a few months to basically catch up to where the other students were. And I really took it seriously and cared, and I found that that did not lead to what I expected in terms of social life. And so I kind of reset and went to the local public school and looked at it rather than competing on marks and outcomes, just basically trying to talk to people and meet them and get to know them and develop myself socially. And that's pretty much what I did from that age, kind of grade eight here until grade 10.

‍

I think at a certain point I realized that the things that I wanted were kind of gated by adulthood. And when I looked at it, there's like no law saying you can't work productively past a certain age. And so I basically just said, "I'm going to get high school done as fast as I possibly can. And then in doing so, hopefully access all life sooner."

‍

Chris Grouchy:

That's awesome. It sounds like your learning throughout childhood and your early adolescence was driven by purely what you wanted to learn about. It was just based on interests. And at a certain point it shifted to just the trappings of adulthood. And so you got there as fast as possible, but if you think about your early childhood experiences, how did you decide that early on in life what to learn about because your parents basically gave you freedom to pick and choose?

‍

Roger Kirkness:

It's a good question. It's more stressful than people think to have to answer that question for yourself because you do have to consider whether you just do the short-term thing or long-term. And I remember one thing that my parents would do is kind of challenge our beliefs by giving us too much of what we perceive to be a good thing and allowing us to learn from why that might not be the best thing for our life. So video games for example, that a certain point we could play it as much as we wanted.

‍

I remember how unsatisfying it was to be able to play video games all the time. You'd think in theory it was great, but after a few weeks of it was just soul crushingly boring. And so I think that showed me not only short term kind of maybe video games all the time is not the best way to lead your life, but also maybe the things that ... the kinds of freedoms that people seek when they're adults, like retirement, maybe that's not actually going to be a source of satisfaction. And that is when I really started reading books. Is to understand if leisure and doing things that I think are fun all the time is not going to lead to a satisfying life, what is basically? Starting to ask that existential question. That was around when I was 10. So I think from then on I spent more time reading and still mostly read as opposed to game. I don't know. It's a really formative lesson.

‍

Chris Grouchy:

I remember in conversations with friends, you sharing stories about how you actually learned to read, which I thought I've always felt is fascinating, but it sounds like the desire to not be bored by video games because they were widely available led to just a force of will to learn how to read. What was that like for you? How did you learn how to read without formal instruction?

‍

Roger Kirkness:

That was another thing. There's this kind of movement towards paleo-learning in kids that essentially if you look at countries that we call underdeveloped, that at a certain point, kids in those countries do adopt reading, writing, and math and it takes about a 100 hours of deliberate practice. And so I think a lot of the reason it takes so much time for people to learn certain topics where we're from is that we learn them at the earliest possible age you can rather than when you choose to for reasons of utility. So I just had no interest in reading, and so I never learned to and I was never asked to. So it didn't occur to me that it would be necessary or that my peers were learning. And so I didn't learn to read until I was almost 10, which is ridiculously late.

‍

I think if I was in school that would be not okay. But at the same time, I had this motivation to learn about things that I cared about that led me to learn to read. And really it didn't take anyone's formal effort beyond a couple days of exposing me to the right things that I would need to bridge into long form books. And I remember when I did ultimately go to school in grade seven, I was essentially the fastest reader. And there's some arguments that learning to read naturally, you kind of speed read by looking at the shapes of words. If you learn phonics, you never learn to read that way. I know a lot of adults use YouTube videos to teach themselves to speed read. So I think there's arguments on both sides, but for me it was very motivated by I wanted the content of the books to be in my brain and to understand, and that didn't occur for me until much later than most people.

‍

Chris Grouchy:

One of the patterns that's becoming clear to me in this conversation is that to basically achieve a goal right out of kind of self-interest, this element of push through whatever roadblock stand in the way to get what you want. So if it's reading or if it's just the content of the book that I want to have to entertain myself, I have to learn how to read and so I got to solve that problem. Do you see this pattern emerge in business?

‍

Roger Kirkness:

I do see this pattern emerge in business. I guess the word that I've been using a lot to describe our style in particular mind, but also just the way that we've thought about focusing is that we're not superstitious. So a lot of companies create problems for themselves because they have these sort of arbitrary beliefs that may not ... that have side effects that they're not willing to recognize. And one of the ones I was thinking about was how people give promotions. So a lot of companies have a 12-month cycle for giving promotions. And I don't think that's invalid because it allows you to establish some amount of fairness across the team. So what you don't want is your extroverts pushing really hard to get bigger comp and you got some introverted people and just people who don't push as hard who end up getting paid less. And so it's this constant cycle of making sure those people aren't missed. But at the same time, if someone is clearly doing a good job, you should consider what it would cost to reprice that person and try and essentially reprice them. No one really appreciates being repriced down, so you can't give people pay cuts, but if they're doing a good job, you should push their comp to the level where you think reasonably that's reflects their seniority. And if they're learning faster, you shouldn't be superstitious and stand in the way of that.

‍

I think that the way that companies harm themselves is that they hire people who are really smart, who may learn faster than average, but then they don't recognize them for that faster learning. And so what happens is two years later, they're vastly ahead of what the company considers to be their peers. They've kind of waited out the same promotion cycle everybody else has and they end up leaving because if you look back at the market and the distance over time that they've traveled in terms of learning, it's much further and faster than the people around them and someone outside that company is going to recognize it. So I think that's just an example of where if you weren't so superstitious about it, you could probably get better outcomes and your people would perceive that they're being treated more fairly. And it does feel like one of those things where the default MBA answer to that question would be pretty different than maybe the one you'd come to if you just thought, how should this work?

‍

Chris Grouchy:

Do we also see a similar pattern emerge in customers? One thing that I've thought about is the customers that who ultimately buy Convictional products and services and are successful recognize that there's going to be this initial period of struggle and it's going to require change management, and implementation, and corralling resources together. But then there's this sort of pot of gold at the end of the rainbow that if they just are able to do hard things properly and not expect an easy path, they're actually more likely to be successful. And I'm wondering if that's been your observation too or maybe how the superstition point dovetails to customers.

‍

Roger Kirkness:

Yeah. Another way of looking at it is that some things ... the things that compound tend to take a long time, and they tend to start with small numbers. And the problem is people quit right in that J-curve moment at the bottom of the curve right before it goes up again. And that really happens because you're measuring some kind of short-term valuation of how well something's going. And it's essentially gone from some optimized state of an amount of energy in versus an output and then it gets worse temporarily. So you're putting a ton of energy into something and it's not really loving you back. And that's kind of like the story of anything long term, is that if you're doing something that compounds, you have to make the initial investment. I think especially this is why companies get less differentiated. If you're doing something and it seems like it's not going well and it's different than what everyone else is doing, it's very hard to justify continuing. But arguably, all the good things that humanity needs to unlock are down paths that emotionally feel like that. Yeah. I feel like you could apply it to so many things.

‍

Chris Grouchy:

I think I heard this quote from Sam Altman, former president of Y Combinator and founder of OpenAI once that all of the benefits of compounding happen at the very end.

‍

Roger Kirkness:

Yeah. Pretty much. Warren Buffet is a good example of this. Where it's as much about age as it is anything else. So you can make beta or beta plus for a number of decades, but if you diligently pursue it consistently, you can compound it into very, very large numbers at the end, which at the start maybe confusing to people who might optimize for something different.

‍

Chris Grouchy:

We've had many fun conversations with customers or potential customers and have maybe even disqualified them in some cases because they're not willing to go the distance to recognize the benefits of compounding.

‍

Roger Kirkness:

Change is hard.

‍

Chris Grouchy:

Yeah. Indeed. Let's jump ahead from high school and talk about your early career case. So Roger wanted the trappings of adulthood, so plowed through high schools as quickly as possible and all of a sudden you're in this stage of your life where you're perhaps graduating earlier than your peers who were in high school. And instead of going to the social experiment that is college or university, you did something different. And so what did you end up doing next?

‍

Roger Kirkness:

The feeling I had in grade 10 was just one that ... was as motivated by not liking high school as it was by things that I wanted. But I think at a certain point I decided that I wanted to work in a business and I just wanted to kind of understand it. And I felt like I kind of talked to the people who were in my high school business class and they're working on some really weird stuff. And I didn't really understand just from watching my parents run a small business, how those things connected. And I don't think if my parents weren't small business people that I would've understood that what they were working on maybe wouldn't end up being that useful in the long term. And I just wanted something real. And so the way that I thought about it was basically that one lever I had was whether or not I get paid and that ultimately I can do something useful. I just don't know if it's useful enough to justify a job relative to someone else.

‍

So I kind of cold emailed a bunch of people and networked a little bit and found my way into this company, GNC. And the proposition was basically that I'll work for free for an indefinite period of time, but I need you to commit for long enough that I get my co-op credits for high school. So it's kind of double dipping in terms of learning about business and practice as well as getting credits to finish high school. This was possible because I lived at home, I had savings and I was very determined to become independent from summer jobs. And so I drove an hour and a half down the highway every day super early to go to my unpaid job. Essentially the strategy I figured out was that people are happy to give you their work, especially their least favorite work. And that one way to progress is to just keep doing that work while you ask for more. And so essentially what I did was I would ask for a project, do that project well, keep doing it on an ongoing basis, but then ask for more stuff from other people. And eventually, I think a lot of the actual work that was happening, the kind of judge work of filling out and completing spreadsheets and reviewing and checking things and even literally filling mail slots, those things were all on me. But over time I got more interesting learning oriented projects.

‍

So when it came time to say, "Hey, the options are I go and find a paying job or I can get paid here." I was so dependent upon that I would've left this fairly large hole in the way that things operated if I had left. So they decided to pay me, and I just kept going down that cycle. And so when I left, I was still doing all of those same kind of unpleasant projects for other people on the team, but I'd also absorbed all of the new things that the company was doing, including at a pretty high level. So I started their social media, but because it was a well-known brand that actually kind of grew quickly. And then they realized that there was this opportunity to sell online. So I started the e-commerce, which again, compared to the field which was making large sums of money, online seemed like this sort of side thing, but on reflection, it's where things are heading. And so that presented this really interesting opportunity to learn at a level that I probably would not have accessed if not for this kind of strategy I used.

‍

Chris Grouchy:

GNC is this big retailer I think that a lot of people are familiar with based out of Pittsburgh. And you were basically working in a Canadian office and basically exploring all of these digital experiments at a time when e-commerce perhaps there was fairly nascent. And so what were you trying to do with their e-commerce? It seems like all of your learning and asking for projects just continued to level you up into harder problems and more ambiguous territories, but you're at this point now where you're like, "Okay. I'm going to go figure out e-commerce for you." And what happened next?

‍

Roger Kirkness:

Yeah. It's a good question. I think, again, one way of looking at it is that in companies you kind of have these respect hierarchies. And the things that get the sea levels and the most central people to their roles are kind of winning the last war. And so these were people who knew obviously much more than I ever will about retail who held those roles and people respected them for that reason, but they tended to undervalue new stuff. And new stuff is more about learning rate than it is about experience. So no one had run a supplement company online for 10 years before because this was 10 years ago and it just wasn't a thing. And therefore it was much more important that you move quickly and learn fast than it was that you had done this before because no one had, And I think on reflection, that's also a good way to optimize problems for when you start a company, but I just basically threw myself into it. Again, learning rate very high, failed at a lot of things, but because of the brand interest, it grew fast in spite of the tactical mistakes that we made.

‍

The situation by the end was we have a fulfillment center and then a marketing function, and both basically reported to me. I was one of two people that had a P&L, but I was still an 18-year-old. I wore suits every day, but it was still this kind of ridiculous situation. There's even funnier aspects to this. So they included the Canadian units P&L in the same email that I got sent with the financials for e-commerce. Again, public company, I was one of two basically officers. Just a lot of crazy stuff, but extremely valuable for business learning.

‍

Chris Grouchy:

I'm sure the gold tie that you wore to work every day also earned you a lot of respect in addition to your ability to take on really hard projects as an 18-year-old.

‍

Roger Kirkness:

It's just ridiculous on reflection. Again, this was not some cool thing. I just worked all the time. I don't think I was self-aware of how I was perceived, but I was productive, so people had to respect me.

‍

Chris Grouchy:

There you go. Okay. Well, I don't want to jump too far ahead in the GNC story because there's actually some pretty informative learning here that then leads to us wanting to start Convictional together, but can you talk a little bit more about the tactics that you use to grow Canadian e-commerce at GNC? My understanding of GNC is it's a multi-brand retailer. They have stores, they have online, they have their own products, but mostly it's third party vendors. And so if they're in this world now where they're launching a direct consumer e-commerce website, what did that entail and what were you basically trying to figure out to basically get them set up to start selling online?

‍

Roger Kirkness:

Yeah. So there was really two sides. They had their own brand and the other brands were kind of used as leadgen to get people in the door so that they could sell that brand. And they curated but not that aggressively. So there was thousands essentially of third party sellers. And so when we did online, it was kind of this question of should it just be the highest brand kind of products or should it also include these third party brands where there's more competition? And so two different stories there. Getting the first party brands like the GNC brand live online was quite easy because we had all the assets we needed. So we had images, descriptions, all the things that we needed to train stores on how to sell the product, basically translate well to a customer trying to evaluate if it's for them on the internet. So it was easy to get our own brand online. And there's more to that.

‍

So the e-commerce software was pretty good, the marketing software was pretty good, but in terms of getting the third party brands online, we experienced a huge amount of pain. And it seemed like the reason was basically because of the software we chose, but also the process that we had for getting them online. And despite our best efforts, a huge amount of time and money got sink into that effort. And most of the third party brands that we had in stores ... So we were already carrying their physical products and shipping them around the country successfully. We were not able to get them online successfully because they had to go through this kind of path and process using old school tech in the form of EDI in order to get there. But then EDI doesn't support a lot of the stuff. You need to provide a web experience, you need images and content and all of that separately. So organizing that effort was what I spent most of my time doing and still it didn't really reach its potential.

‍

Chris Grouchy:

I want to double click on EDI, pretty important concept in our world and our customers world. I'm going to pull a Y Combinator interview question here, which is explain EDI to me like I'm a 10-year-old. And not a Roger level intelligent 10-year-old, but maybe the average 10-year-old.

‍

Roger Kirkness:

So when you walk into a store and you want to buy a product, somehow that store has to get that product there for you to buy. And historically, before the internet and phones and all of these easy to use communication technologies, something was needed in order for the store to communicate to their supplier that they needed that product, and that's basically EDI. So you can imagine if you have a little store with a small number of products that it's pretty easy to call someone on the phone and say, "Hey, I need more of these." But if you have a store as big as Walmart or a grocery store, it's absolutely massive stores with tons of products, in those cases, you need something that's a bit more automated. And so historically and unfortunately, still that's primarily done using this technology called EDI.

‍

Chris Grouchy:

So most EDI is pre-internet today. Correct?

‍

Roger Kirkness:

That's right.

‍

Chris Grouchy:

Makes sense. And so why do businesses continue to use EDI then if it's pre-internet and largely inflexible?

‍

Roger Kirkness:

I feel like at the time that I was at GNC, it made more sense because companies were running their business on software that was primarily run on their own hardware systems. And so it was a way that you could standardize the communication between your systems and someone that you do business with. So it's kind of like the language. And because it had become the standard language, it was very hard to get off of. So if we just all unilaterally decided we wanted to speak a language that wasn't English, we would have to change over so many aspects of how we live. Labels, signs, the way we talk, how we educate people. And so it's kind of similar with EDI where the fact that it's so universal. It kind of traps people in using it. And in particular, if all of your core systems speak that language, then I don't think beyond a certain point you have any incentive not to use it. Just maybe some foreshadowing, but it seems like as time goes on, people still use it even though their core systems don't. So we talk to companies that are still using EDI to enable these connections between them and the companies that they buy from, but their core systems are using APIs and cloud software. And that's where it stops, in my opinion making sense.

‍

Chris Grouchy:

So EDI, just a fact of life in the world of B2B trade because of how entrenched it is. And yet the promise of EDI was that you could implement it once and be able to talk to any buyer or seller globally, and that was definitely not the case. And it seems like parties with market power have essentially corroborated EDI into making its own sort of format, they've made their own flavor, but that has some pain associated with it if you're a trading partner who perhaps has some experience with EDI or may not. And so if you're a large retailer and you have your own EDI feed, it works great for you, but it might be really difficult for another supplier that you're trying to onboard to actually get set up and integrate with it and start transacting with you. Is that kind of a fair summary?

‍

Roger Kirkness:

Totally accurate. And further to that, it's something where if you don't ... people thought that the standard would be able to continue to be used, but it's much internet standards. So while it's nice to have everyone standardized on exactly the same definition of how something works, if anything about the world changes, it kind of loses relevance quickly. And that's the problem with the internet is that it's created new ways of transacting, which we're not accounted for when the standard was made, But the standard by definition is so explicit about what it expects and what it can't handle. That the only way to introduce those changes is to deviate from the standard somehow.

‍

Chris Grouchy:

And at GNC, were you aware just how much transaction volume was happening through EDI or were you basically like, "This technology is really painful. My vendors hate it. And that's basically the extent of my insight here."

‍

Roger Kirkness:

Mostly the latter. I think I overestimated how many of our suppliers would be able to use it. I just said at least the big companies, you would think they would be able to use it, but the big companies that sold to us were not big enough to use EDI. And I was like, "Who is big enough to use this?" It seemed like something that was limited exclusively to companies as big as GNC. Yeah. I didn't really think about the global implications. Mostly it just made my life much worse.

‍

Chris Grouchy:

Once you start to dig into the mechanics of EDI and just how much GMV flows between buyers and sellers using EDI, it's kind of like the matrix, but it's very difficult to participate in the matrix unless you're some sort of insider who has basically taken on the cost and complexity to implement it. Let's jump ahead to a few years later when you got recruited to join Shopify Plus, a modern e-commerce platform. A different experience, I'm sure, than working at a large retailer implementing old school B2B trade technologies. Why were you recruited to Shopify Plus?

‍

Roger Kirkness:

Yeah. That's a good question. I met a mutual friend of ours at an event who presented a really compelling picture about what was happening. And it occurred to me that there was ... it was parallel universes. So this exposure to B2B trade and how it worked and Shopify and how easy it was. And it even had a Shopify store at the time trying to sell this salt water plant that you can turn into biomass, kind of like an unrelated story, but it seemed so much easier to set up than the technology that we were using. And it didn't make sense to me that there was no bridge. So it essentially joined saying to myself, there's probably an opportunity for this company, which I respected a lot and still do, successful Canadian tech company to make this process actually usable for small companies, which it seems like they had done successfully for e-commerce software. That to me was just too compelling not to consider it. And there was logistical things. I had to move cities, but that was the assumption I joined on.

‍

Chris Grouchy:

So you were recruited to join Shopify, you are a very early member of Shopify Plus, and there are sort of some interest in exploring B2B because why wouldn't you take this B2C e-commerce platform and see if you could sell it and generate money doing B2B transactions? What was that experience like? What did they have you actually do, and what were the conclusions from that customer development process?

‍

Roger Kirkness:

So I think the experiment started with product, and that was one interesting thing. Lesson I learned was whether your product led or sales led and the effect that it has on what you build and how sellable that product is. So in that case, I was joining an already existing effort. The product was created by others and conceived of by others, and my role was to help figure out how to commercialize it. And there's kind of a small number of companies on this beta, a few of them successfully using it and a ton of demand. And we tried to figure out basically how to sequence that demand and especially how to adopt a new kind of customer that would not have otherwise worked with Shopify in selling it.

‍

So that was the intent. And I spent a lot of time just talking to companies and mostly disqualifying them. And what we found was ... or I guess what I personally started to notice was that depending on who your customer was in a B2B context, the technology that would be optimal for enabling that customer to purchase from you differed quite a bit. And so if you sold to an offline company that was small, kind of in the analogy story that I used, it's very similar where it's feasible to either enter an order in online or call someone. And in both of those cases, what Shopify had built was well suited to it. And so if that's who your customer was, like a clothing brand, especially a high end clothing brand, then it made sense, the product that we had made sense. We observed though that if you sold to big companies or if you sold to online companies ironically obviously because many of those companies were using Shopify, the technology that would be needed had to be more software enabled, which kind of brought me back to the CDI thing. And so I think the conclusion I had after a few months of talking to customers was basically it wouldn't be enough to just build a great portal experience in online ordering. It would also have to enable some amount of system to system purchasing.

‍

Chris Grouchy:

I would want to summarize the insight, which is that a portal-based approach would work great if you are selling to small offline buyers as a brand and perhaps you have some degree of market power over them, but Walmart as a buyer is never going to log into your portal to buy your shoes. And why is that?

‍

Roger Kirkness:

Nike can tell a sneaker shop, "Hey, this is how you order. Do you want your Nikes or not?" I think in many contexts that's not the power dynamic that exists in a trade relationship. And if you sell the Walmart, especially if you sell something that's either under their brand or just a brand that doesn't stand on its own, then you kind of have to go forward with the assumption that you have to do things that suit their purchasing process and it's more about how they buy rather than how you want to sell. And in those cases, there's a lot of standardization. And so one of the things EDI did well in theory was standardized how you buy things as a large retailer, and that's I think why a lot of them adopted it. And so those worlds were kind of colliding where you'd have tons of brands on platforms like Shopify trying to expand and go B2B and then you have these massive companies, Walmart, but not to pick on them and many others who had used CDI historically to purchase things. And those two worlds didn't talk to each other at all.

‍

Chris Grouchy:

We're using Walmart as an analogy of a party clearly with market power.

‍

Roger Kirkness:

Pretty good execution though.

‍

Chris Grouchy:

Yeah. Oh, yeah. Absolutely. Okay. So at a certain point we basically decide we're going to go and start a company around this, and there wasn't much interest from other players in the market to do this properly. You left to start and started basically figuring out, okay, what are we even going to build? And those very early days of Convictional, what mistakes were made? I know some of the mistakes, but I want you to tell listeners what some of the early mistakes were made once we both started getting working on Convictional?

‍

Roger Kirkness:

So one mistake I feel like we avoided making was that I recognized the need to specialize as a founder. And so one thing I would've regretted if I tried to do that on my own is not bringing you along because even if now we kind of have this team around us who arguably are better than we are in their area of focus at the time, it was essential that you could specialize, then I could specialize in order to progress our learning. In terms of mistakes, I feel like the thing that we learned was just there's like this fallacy that most software can be reproduced by an engineer in a weekend and like, "Oh, I could build that in a weekend. Like Twitter, it's got tweets and a timeline and stuff. How hard could that be?" And I think the fallacy is basically that you can write V1 in a weekend and then V2 will cost millions of dollars and require years of learning and a ton of customer development time and failure.

‍

So I think maybe we just underappreciated a little bit how true that is and that realistically customers don't get value into V3, V4. And so you somehow have to bridge this period of time where you're primarily learning how inadequate what you do is for the needs of customers. And it's also hard, I guess, to make a decision about which of those customers to focus on. So I think my observation is the product stayed pretty similar, but the go to market changed a bunch of times, and that's because it came closer to truth in terms of how people actually adopt and think about the software. But yeah, mistakes-wise, I feel like we made basically all the mistakes we could make. We had enough savings to paper over them.

‍

Chris Grouchy:

Right. Yeah. It makes sense. I think it's hard to say whether they were mistakes that ultimately set us back or that led to actually significant results in the future Once we wrapped our head around why we made the mistakes and how we could have actually corrected for them. One thing that comes to mind is revenue dating, the early product. So actually getting people to charge money. And it felt like a mistake to do that because we were asking for pretty large sums of money from customers very early on without having anything really. And we were trying to build up to their requirements. But on reflection, that may have resulted in a lot of rejection, but it probably instilled a value in us and the team that we are actually a business that's going to create products that large companies want to pay money for, and it will be a considered purchase for them. And as a result, that's going to force us to deliver outsized return to those customers. When we went to Y Combinator, it seemed like a lot of these B2B SaaS startups were just giving away their products for free or pennies on the dollar.

‍

Roger Kirkness:

I totally agree with that. I don't know if it's the insecurity that comes with Canadianness or what led to that for us, but one way to qualify how much someone values your product is to charge for it, which seems obvious, but I think startups almost delay the inevitable conversation about how valuable it really is as long as possible. I think the other thing I learned, I guess, would be that criticism is actually good. If people don't like your product, they'll just churn from it. And so if they have chosen not to churn and they ride it out with you and give you critical feedback to push it forward, it's an indication that the problem is important enough that they're kind of in it with you. And I feel like that's also been common the whole time, but also my observation of Y Combinator startups in general was that people would say they love the product, but if you attached a more reasonable and sustainable price tag, all of a sudden they would churn.

‍

Chris Grouchy:

It's a great way to validate whether you have something that people are willing to pay for if you just attach a price tag to it and see what happens. I want to go back to mistakes, but double click on the engineering mistakes that we made in the early days, specifically. I think some people listening to this are going to be technical and they're probably interested in understanding, "Okay. Am I going to agree with tech decisions that we've ultimately landed on since having built an actual engineering team?" What are some of those early mistakes that we made in terms of tech choices and tech decisions?

‍

Roger Kirkness:

Very good question. I guess the jobs to be done, we were trying to accomplish, we sort of validated quite early, but the approach we had to accomplishing them, it turned out needed to be quite novel. So bridging kind of the world of batch process stuff where you do a bunch of work and then at the end you see if it was successful and it's more modern world of like API based events where something happens, it's reflected in API, but it's real time. Bridging those two worlds together is actually quite architecturally complex. And so I would say the first versions that I wrote as an example, I wrote it in JavaScript with a node back end, and which is nice, you can iterate quickly. The one painful lesson I learned there was the necessity of basically static types. So saying this is a number and this is a string, and the difference has meaning in the code and not just when it's running. We basically had to start from scratch because those things, while not essential in all domains, various big companies use node back-ends in our domain of bringing in arbitrary data from other systems, especially data that doesn't explain itself.

‍

So like EDI data doesn't actually have a type. It's kind of all strings and you have to figure out whether it's a number or something else. Doing that in a type unsafe language is actually deeply flawed. So it's like you got to rewrite. The other thing we learned was around how to make the processing of those events durable and the meaning that those events have compared to some contexts. So some applications are sending a high volume of emails. If one email gets missed, you have to think about what's the business value of the failure of that email sending? And I would argue it's meaningful but not significant. In our case, we're literally dealing with money, so this would move fast and break things is turned out to be entirely inappropriate when it comes to areas of actual money changing hands. And there's so many kind of architectural trade offs that you have to make sure that those things are durable and happen reliably.

‍

I feel like every possible way that it could go wrong from the seller's data, the buyer's data, it did. And so over time, we've adopted this practice of ops. It's an active thing that we do. And every time something goes wrong or happens and it's unexpected, we escalate it to ourselves via email and we've continued to do that even as we've scaled usage a lot. And so in the early days, that meant that most orders required some kind of human intervention. Now almost all orders require no human intervention, but the way that we got there was just lots of sending ourselves emails and iterating how we did things and definitely a lot of mistakes made.

‍

Chris Grouchy:

That resulted in learning ultimately.

‍

Roger Kirkness:

Yeah. Ultimately. Learning is hard though. I feel like that's the other thing. I feel like because of growth mindset, learning has become the sort of mean where people are sort of like, "Yeah. I love learning." This is kind of unrelated, but I feel like it's still related to the point you made. It's like if you look at people's resume and they've held the same role for seven years straight, it's like, "Are you learning learning or are you learning just enough to keep doing the same thing that you're doing?" Because there's a sort of rate of entropy of stuff that you know. Not 100% of what you learned last year will be relevant this year and you'll have to learn new stuff. But to me, learning is significantly moving the needle. You used to believe X and now you believe Y. And you had to go through some painful experience to bridge those things.

‍

Chris Grouchy:

I think that one thing that people often fail to incorporate into their learning process is reflection, and there's often just too much distraction and stimulation to justify the actual opportunity cost of slowing down and pausing. So this relates to my next question on culture. We've been very deliberate in terms of how we've designed culture at Convictional and communication, and one of those examples is that we don't use Slack. We're email only. That's a very important point that we try to communicate to candidates that are looking to join Convictional. Can you tell me more about the reasoning behind why we don't use Slack and prefer email and documentation?

‍

Roger Kirkness:

One observation I had, I guess, about Shopify was that you had a lot of really smart people. And that different people would depend on each other for different things. So no one person could solve these super hard problems on their own. It required this kind of mix of people who think differently, but that required a lot of communication to bridge if you wanted to go fast. And so the solution to that was basically a ton of Slack messages at all hours, confirming assumptions with the people you knew were the best in a certain area.

‍

I saw at one point in internal report that said that the average person was sending a message every 11 minutes and spending more than two and a half hours a day using Slack. And I kind of thought about what are the other uses of that time? Well, it could be talk to customers, it could be hang out with my friends, my family, people that I care about. It could be write something down so that every time someone has that question in the future ... I'm reducing the amount of Slack burden overall. I don't know how many people there were that were thinking that critically about it. And I think our culture initially was almost a criticism of that where we wanted to be able to spend uninterrupted time focused. And then we realized as we grew that it had to become more nuanced, that we wanted people to understand that it's okay to take an entire half day off the grid and focus on writing something that will live on for a long period of time because you don't lose knowledge learning that way.

‍

Chris Grouchy:

I've had very heated debates with friends over dinner and wine where I've been like, "We don't use Slack and we're able to function very well as a result." And they're mind blown and just completely disagree that that's even possible. How do you stay aligned in a fully remote work environment in the example that you're sharing where there's no synchronous communication happening?

‍

Roger Kirkness:

I would argue a lot of the best, most dominant technology products that exist now were created in a time where there is IM, but it was not nearly as company wide widespread. It was more team specific. So Apple, Microsoft, Google, Facebook, all of those existed before Slack did. Somehow they created trillion dollar companies. I think alignment is more like you shouldn't hire people who need to be aligned. You should hire people who kind of understand the philosophy and the values with which you approach things, but who have better ideas than you do in the area that they're taking responsibility for. And I would argue a lot of companies start to compromise on that at a certain point. And that's where you need a lot more aligning work. Is to give people the tactics that pair with the strategy of the company. And to me, strategy is just like how do you engage with customers? What is the definition of those customers? What don't you do? Who don't you serve? And once you define those things, I think our engineers know better than I do in many ways. What builds up, what things to do that meet that definition of the customer that we have. And so I think it's worked excellently to our current stage and the question is how you scale that.

‍

So I think you and I genuinely have these beliefs and have this kind of these learning experiences that have led us to have conviction in this way of communication. The question is how you grow the team large. And I'm sure we'll learn a lot doing that, but I'm quite convinced that it can be done.

‍

Chris Grouchy:

Well. If you look at a company like GitLab, which has not only everything written down, but it's all publicly available, all of their team handbooks, how they manage people, compensation bands and so on, it's all documented properly. And so I think a company of that size and scale where you're on a track to actually going public can function adequately by believing in this sort of asynchronous first approach.

‍

Roger Kirkness:

I think that burnout is probably the central problem in business right now. And that at any given time, you have to focus on the bottleneck to progress. So the team is smart, the mission we're working on is good, the product that we have I think is relevant to customers, but if we burn out, all of those other things will suffer significantly. And this whole Zoom meetings all day is clearly not sustainable. I think people recognize that, they mean it, they take time off. Companies are literally shortening their week. But when you shorten the week, it's just concentrating the Zoom meetings. It's not really solving the problem. Somehow we need to be able to communicate and share knowledge in a way that doesn't require one-to-one person contact for all forms of communication. Obviously it has its merits in some cases, but if you project this forward, people are going to burn out. The joke I make is I don't want anyone on the team on our team. If they want to leave and start a business, I'm obviously supportive, but no one on the team should leave and become a Van Life YouTuber because they're so burned out. This is really an important thing to humanity and probably a way to have significant impact, and burnout is the enemy of that.

‍

Chris Grouchy:

I think a lack of self-awareness gets in the way of the goals of reducing burnout. And it's not to say that, "Oh, one is self-aware, achieve this state of enlightenment and therefore they will never be get burned out." Life happens. But it certainly helps to be able to have a sense of awareness about am I redlining and am I doing this for way longer than I should be and actually take steps and measures to prevent that from going any further to a point of self destruction. So one piece of this is self-awareness. And we really like to work with people who are self-aware and if they're self-aware and they're vulnerable in sharing that, we can actually work with them better. And one tool we have internally for doing this is user manuals. Can you talk a little bit about user manuals and why we do them and ultimately what they are?

‍

Roger Kirkness:

I think user manuals is something that ... There was someone, a Facebook executive, who had a user manual that was like, "Here's how to get what you want from me." Kind of thing. And then people started to view user manuals as this sort ... It's a way of asserting power. If only managers have them. It's like, "Here's how you can get what you want from me. Here's how you will engage with me." Kind of a one way thing. I think we just tweaked that and said, that should probably apply to everyone in the company. And this notion of flexibly working the way that you want to and rules of engagement, it's useful to have that understanding of everyone that you work with regardless of where they occupy in the org chart. We all, I think really early, started writing user manuals for the initial people joining our teams and then we asked them to do the same. And then as time has gone on, we just have this giant repository of a kind of a one pager life story. Here's how I like to work, here's how I get your needs met from me, here's my strengths, weaknesses of everyone in the company, which is extremely useful because without that, it does make it harder to get your needs met and likewise harder for people to understand the ways that you like to engage and also what your boundaries are.

‍

Chris Grouchy:

Now, I want to wrap up with maybe one or two more questions. And the question I want to end on is just where we're going. We've assembled this great team, we have amazing customers, we have a great product, a huge market opportunity that seems to be infinite. And one thing that really stood out to me was recently in a company-wide town hall, you had mentioned that we had completed our five year roadmap or we were on track to complete our five year roadmap in less than a year. Okay. Team is insanely productive. What do you envision the next five years of Convictional will look like? What will we pursue over the course of that five-year period?

‍

Roger Kirkness:

I think when we built the roadmap that it was a lot of ... or I guess when I did, and obviously with a lot of feedback from the team and customers, it was primarily oriented around table stakes functionality. How do you take what exists now and make it better? And I've realized that the bar in terms of what exists now is just so low that the difference between something that's marginally better than what exists and the theoretical maximum of what exists is very large. And so a lot of that roadmap was focused on things we could build that were single sign on. And we found a really short path to building really high quality single sign on in a lot of areas. But when I think about what we can build in the future, it's more based on what is the theoretical maximum of this thing.

‍

So if it's supplier onboarding, well, ideally there's no effort at all. Ideally both start, "I have some software, you have some software." Those systems pretty much come together magically. They map everything that needs to be mapped, they sync everything that needs to be synced and we're set up. It's just about consent basically. I'm consenting to giving access to this data and then we're done. So because that's the theoretical maximum, what I'm trying to do in the future and what I think we should aim for is to think about, "Okay. There's this no effort, consent handshake process that has to happen to share data with each other and then all these initiatives underneath that." So whether it's like augmenting product data with AI in order to avoid a person having to go in and add attributes to a product, those are investments worth making, which again, like we talked about, they compound. So you need a lot of data and you need a lot of time in order to reach that state, but because it's possible to reach that state, I think that's what we should be aiming for. And that to me, like lesson learned, that's probably more likely to take a while.

‍

Chris Grouchy:

Although maybe we'll do a part two to this in a year and we'll be eating our words that.

‍

Roger Kirkness:

We'll be done again. Yeah.

‍

Chris Grouchy:

Exactly.

‍

Roger Kirkness:

We'll see.

‍

Chris Grouchy:

This has been a really fun conversation with you, Roger. I just am so grateful I get to be in this journey with you and learn from you every single day. I know the team is super appreciative of your leadership and what you bring from the table too. So I hope our listeners enjoyed this conversation. It was a lot of fun and we'll do it again soon and talk a little bit more about maybe what some of those specific hard problems are that we're going to solve for the next five years.

‍

Hey, everyone. Thanks for checking out this episode of the Legends of Retail Podcast. We covered a lot of ground today around company building, communication, and of course some gold nuggets that you won't want to miss about B2B trade. We talked about how you can proactively promote your employees before your competitors get to them. We talked about the flaws of electronic data interchange in a world that is post-internet. We talked about email instead of Slack to prevent employee burnout, while encouraging the use of written documentation to get information across company networks. We talked about why writing employee user manuals is a fantastic tool for helping people get to know each other in remote organizations. That's it for today. Thanks for listening. And be sure to subscribe on Spotify, Apple Podcasts, or wherever you listen to podcasts. My email is chris@convictional.com. That's chris@convictional.com. In case you have any feedback from me, I'd love to hear from you. Thank you and see you next time.