Link to podcast. Link to transcript, which I organised as 20VC didn’t publish one.

8th March 2023.

Sajith: really good podcast that is relevant for anyone in the tech world, but PMs and anyone with a strategic bent of mind will really dig it. Glen comes across as some one who can do both blue-sky thinking / strategy and building / coding really well. He is thus able to talk nitty-gritties and tradeoffs while building, even as he speaks about PM-ing and org strategy from a 30k view.

What I really liked from the podcast

  • The Outcomes-Principles-Assumptions framework that you should align your team and peers in the org around before you start building. While it isn’t strictly a PM framework, I thought this was particularly relevant for product development.
  • When you build products that have extremely wide appeal and user bases, ensure progressive disclosure of features to keep life simple for your users, and ensure that third party apps can build high quality products for your users
  • The closer you are to the bottom of the stack, the less likely that you will be able to tie a particular feature development to revenue generation, and hence the more you should ensure you build for the future, not the next 2-3 years
  • Storytelling matters – if someone cant see your product as relevant to them, or cant empathise with it, no amount of numbers will convince them. And, repeat repeat, repeat. People forget all the time. Per Glen Coates, There is barely three words you can lodge in people’s minds, and you need to chose what they are.

Highlights or my public notes from the podcast below.


Outcomes, Principles, Assumptions

Glen: “Outcomes is how do we know this is gonna go well? When this is all done, you’re telling me you wanna run this project and it’s gonna take 10 engineers for six months at the end of this, how are we gonna know that it was worth it? It’s a really easy question and you’d be surprised how often teams do not have an answer to this question.

Sometimes you can say, Hey I want to take a conversion rate from X to Y, I wanna take a performance metric from Z down to P. Sometimes the answer is a metric. Not everything is a metric though. Sometimes the answer can, when Toby looks at this page, he’s not gonna vomit anymore. It can be something subjective because sometimes the work we do is really aesthetic work. It’s make this page beautiful, it’s godawful right now. But just being clear upfront on really, hey, when you do this project, when we come back six months later and I ask you did it go well? How are you gonna answer that question? Let’s just agree upfront what the format of your answer is gonna be so we all know what we’re actually working towards. And you’d be surprised the number of times people don’t even know what the answer to that is or people have different answers to what that’s gonna be.

Principles are probably the most difficult and counterintuitive, but principles are basically how are we gonna make the very hard decisions, the 51-49 decisions that reasonable people could take either side of. In this project we can anticipate that there are gonna be some forks in the road in here and reasonable people might take either side. And as a team we all want to go into one side. We don’t want to do this thing where half of us are pulling in one direction, half of us are pulling in. 

So obviously at Shopify we serve primarily three constituents. The merchants who run their businesses on Shopify, developers who build apps and themes on Shopify, and then the buyers who are buying from Shopify stores. We have these three constituents that are always circling. So a principle you might put into a project would be something in the cases where the interests of merchants and developers conflict, we are going to break for merchants every time. Right now different people might take reasonably either side of that argument, but it’s really bad if half the team is taking one side and half the team is taking the other. When you can see that a project is gonna have these conflicts in it, it’s good upfront to say, hey, when we hit this fork in the road, we’re all going left so that the team doesn’t just split in half and start destroying itself.

And then we get into the bottom of the conversation and what we realize is that we have some foundational assumption about reality that’s different, which explains why we are talking past one another.

When mobile is on the rise, you could see someone being like a mobile product for Shopify or maybe like a mobile web mode for Shopify where in the back of their head they’re like, okay, mobile’s on the rise. I assume that all web traffic is gonna eventually become mobile. The way they’re building the product becomes very extreme because they’re like basically all on online shopping is eventually gonna be mobile and another person might have an assumption that’s, hey I agree mobile’s on the rise, but I think it’s gonna asymptote to somewhere around 60% of all web traffic, which means that 40% is still gonna be desktop, which means desktop’s still very, very important. If two people on the same team have those two different assumptions, the product that they’re gonna end up building is gonna be wildly different and they’re not gonna understand why.

The one that is done the least is probably weirdly, the outcomes one.”

Marketing is red, sales is green

Glen: “This is the marketing is red, sales is green problem that I used to have at Handshake, which is one of the things, sorry we’re getting off product now…one of the things that like you often see in go-to-market organizations is that because sales and marketing people are very goal-oriented and often compensated on the numbers, they really want numbers that they’re in control of. They really want the number that they’re gonna be paid on to be a number that they actually control. This leads marketing teams to want to declare goals that are about the marketing funnel but not the bottom of the sales funnel. Which is exactly what you just said. Oh I met 20 people, but 19 of them are homeless.

And this is when you get into the weird parts in businesses where I guide. The marketing’s doing great but sales is doing terrible, what’s going on? Or like marketing and sales are doing amazing but customer success seems to be really struggling. What’s up with that?”

The closer you are to the bottom of the stack, the more long-term oriented you have to be

Harry: “Can I ask you, you mentioned there about being close to the finances in the products that we build. How do you, and this is a really tough one I find with product, how do you balance in terms of product decision making between generating revenue today and innovation for the future?”

Glen: “Yeah, I think the closer you are to the bottom of the stack, the more responsibility you have to build for the future. “

The two elements for building a universal product

Harry: “Can I ask you speak about the effect on the world and the product magnitude? You mentioned I think it was like the 14 different teams earlier, yeah. How do you think about retaining simplicity in product with great depth of customer base and great needs? How do you retain that simplicity? “

Glen: “Obviously look in the product itself, there’s really only two solutions to, if you wanna do the stretch up thing, which is something we say at Shopify, we serve people that’s like literally a 12 year old kid running a lemonade stand up to like Supreme and JB HiFi and like all these insane companies. If you really wanna stretch like that, there’s very few pieces of software that have ever adequately served both ends of that. And the only two pieces of leverage you really have to allow that stretch is, one is extremely good execution of progressive disclosure in the core products. So features that appear at the right time only when you need them. The canonical example of this is like users and permissions management in business software, right? When you only got one user, you literally do not need user permissions management cuz there’s nothing to do.

So the user and permissions management thing being there is literally making the product worse because there’s nothing to do and it’s just adding complexity. Part A is progressive disclosure of what I would describe as antifeatures, like a feature that makes the product worse by being there because it’s not needed yet. And then the second thing, which is probably the bigger piece of leverage is an app platform and an app ecosystem apps can be built on top, they extend the platform with functionality that doesn’t have out of the box. But the really critical thing here is you need two things and they’re actually basically conjoined at the hip. You need an app platform that allows apps to be built that are of core quality to the core features of the platform. That is absolutely critical. Look at iOS. iOS is like a great example of this.

Like when I open up a first party app like Messages and I open up a third party app like Spotify or Lyft or something. The third party apps are as good quality as the first party apps because they’re all built on the same toolkit and Apple made the tools available such that third party apps could be as good as the first party apps. This strategy does not work if the third party apps are all trash and it’s only the first party apps that are actually good. But if you have an app platform that basically gives the same tools to third party developers that the first party ones have, then you enable this rich ecosystem of actually high quality apps that allow the large enterprises to scale up because they add apps as they need them, but the core products stay simple for the people who don’t need that yet. Which is again just part of that stretching. And I guess apps in a way are a form of progressive disclosure. “

Storytelling, and, repetition

Glen: “This storytelling theme was beaten into me almost as an existential threat. When I was trying to raise money at Handshake, I was a member of a CEO peer group… But one of the guys in the group was the managing partner of a VC fund and his name was Phil. And on the first one of these meetings, each founder got up and did the little pitch for their business and I got up and I was like, yeah, it’s Handshake and it’s wholesale and there’s these sales reps and there’s these businesses and they need to trade and it’s a huge problem and we’ve got this mobile app and blah blah blah blah blah.

…at the dinner that evening, Phil sat down next to me and he was like, Hey man, tell me the story. And I was like, I did tell you the story. And he’s like, man, tell me the story. Tell me something that I can feel. And I was like, oh yeah, it’s this wholesale thing and there’s the truck lifts and there’s the FedEx trucks and there’s the iPads and and there’s the order books and blah blah. Like no, no dude, tell me a story. And he beat me into the ground over, I don’t know, I must have sat there with him for an hour with him just being like, just tell me the story over and over and over again.

And I eventually got to a story, I literally ended up making it up on the spot. I was like, you know when you go into a store and you walk in there looking for a particular thing, you want a white t-shirt in a particular size and you go to the shelf where that shirt should be and it just isn’t there and it should be there. That is literally what we are doing. We help the companies with the shirts make sure that they get to the right stores in the right time so that when you get to the store, the shirt that you expect to be on that shelf is actually there. And just multiply that by all the people in the world and all the shelves in the world and all the number of times that you’ve gone there and it’s not been there. That’s what we’re solving.

That’s not even a very good story. But finally Phil was like, I get it. I could explain that to my mom and she actually understand what you did. Since then, I’ve just understood that if you do not tell someone a story that they can see themselves in and they can empathize with, you can say all the numbers in the world, you can describe all the processes and analysis in the world. No one gives a, no one’s got time for your shtick. I really try to encourage people to find a way to tell the story. Humans are emotional creatures. If you can’t get me to feel something, I’m probably not gonna care no matter how many numbers he showed.”

Harry: “Yeah, I totally get that. In terms of the communication, I think a lot of founders struggle with the repetition. How important is repetition in storytelling?”

Glen: “You’d be shocked how fast people forget things. This is why compelling storylines with just memorable sound bites are so important cuz you only get to lodge like a few words in someone’s brain before every word you add pops one out their other ear and you only get to choose maybe three words you can lodge in someone’s brain at any given point in time and you really need to choose what they are.”