Idea
Disrupt the karaoke machine and ugly home tools
Harley points at categories ripe for a beautifully-designed 10x version: karaoke machines (so big and clunky that companies rent them out) and home tools (clunky Home Depot stuff begging for a Touchland-style redesign).
“someone should go and, like, disrupt the karaoke machine business. Like, that to me is, is such a brilliant idea. Or, you know, uh, tools. Like, you move into a new house and you kind of need, like, you know, random tools. Most of these tool companies you buy, like, at, you know, Home Depot, they're clunky, they don't look good. There's actually a couple, like, beautifully designed, almost, almost like the Touchland version of tools selling on Shopify.”
Steal thisPick a clunky, ugly home-category product and ship a beautifully designed premium version.
Framework
Great founders grow the pie, not just take a bigger slice
Harley says people treat TAM as zero-sum, but the best founders grow the market itself. Sequoia invested in Shopify by sizing existing retail SMBs, but what they missed was that Shopify (like Touchland) grew the pie by bringing in people who weren't even entrepreneurs before.
“What they missed was that, that we're actually growing the pie itself, right? There's like this positive sum that someone who's starting, you know, I don't know the person, um, behind Touchland, uh, Sean, but, but presumably like this person didn't necessarily like what she wasn't in or he wasn't in this industry previously. Maybe it wasn't even an entrepreneur previously.”
Steal thisDon't size your market by today's incumbents; build for buyers who aren't in the category yet.
Number
AG1 does $600M a year on a single SKU
Harley cites AG1 (founded by Chris) doing $600 million a year with one single SKU as proof that great entrepreneurs grow the pie infinitely rather than fighting for a slice.
$600M
Annual revenue · USD/year
“Kat Cole, who's the CEO of AG1, they're doing $600 million a year with AG1 with one single SKU.”
Number
Prose does $100M in personalized shampoo, 80% recurring
Prose, which makes personalized shampoo based on a hair-follicle analysis, does about $100 million in annual revenue with 80% of it from recurring customers.
$100M
Annual revenue · USD/year
“They're called Prose, P-R-O-S-E. They do personalized shampoo based on hair type. They're doing about $100 million. This is public information. They put it out there. $100 million in annual revenue. 80% of it is recurring, uh, customers.”
Tactic
A nanny is a marriage saver for new-parent founders
Harley's mentor Dave Siegel told him that if he could afford a nanny when his first child arrived, it would be a 'marriage saver' — luxuries don't remove the stresses of new parenthood but lessen them, freeing you to do better work.
“he said, it's going to be a marriage saver. And I didn't really understand what it meant. And what he meant was like, just everything, everything is easier as a new parent at home.”
Steal thisIf you can afford help as a new-parent founder, buy it back — protect the marriage and the work.
Framework
Company phases: Swiss Army knife, then triple-threat, then spiky objects
Harley's model of company stages: in phase one everyone is a Swiss Army knife doing everything; the second phase is a 'triple threat' where people do a few things well; the mature phase is finding each leader's single highest-value 'spiky object'.
“you think of the phases of companies, like at the beginning of the first phase of a company, everyone needs to be a Swiss Army knife. Like everyone do everything. And then the second phase, it's like more like what I call like triple threat phase”
Steal thisAs you scale, stop assigning by title and reassign people to their single highest-leverage 'spiky object'.
Tactic
Don't return corp dev calls — they plant the 'what if'
Shopify deliberately did not return phone calls from corp dev teams early on, because even taking the call starts a 'what if' process in your mind about the number and terms — a distraction for a company that wanted to build a 100-year independent business.
“Once you take that phone call from a corp dev person at a big company, you begin to inadvertently or advertently start a process in your mind of beginning to think about what if. What if this number is right? What if the terms are good?”
Steal thisIf you want to stay independent, don't even take the acquisition call — it poisons your focus.
Take
There is no magic number — every stage is just different problems
Harley says the 'magic number' founders fixate on (the passive income that covers all expenses) is bullshit once you hit it; just like there's no stage where you're cruising, there are only different problems at 7,000 people and $100B than at 70 people.
“We all have this number in our mind. Like, when I get that, I can think of— I can basically extrapolate my passive income from it, and that's going to pay all my expenses and I can do— and then you hit the number, you're like, it's like, it's bullshit. It's complete bullshit. There is no magic number, just like there is no stage of a company whereby you are cruising.”
Framework
Before hiring, prove AI can't do the job better
Shopify added a friction step: before hiring anyone, the manager must substantiate to the talent team why AI cannot do the job better. The byproduct is that while justifying the hire, people realize AI can actually cover much of the work.
“before you can go and hire someone on your team, you first have to substantiate to Shopify, to, to the hiring manager, to the talent team, why AI cannot do it better. And one of the things we're noticing there is by actually creating this like friction of saying like, before you go and hire, you must first do it.”
Steal thisRequire every new headcount request to first justify why AI can't do the job.
Framework
Before hiring, prove AI can't do the job better
Shopify added a friction step: before hiring anyone, the manager must substantiate to the talent team why AI cannot do the job better. The byproduct is that while justifying the hire, people realize AI can actually cover much of the work.
“before you can go and hire someone on your team, you first have to substantiate to Shopify, to, to the hiring manager, to the talent team, why AI cannot do it better. And one of the things we're noticing there is by actually creating this like friction of saying like, before you go and hire, you must first do it.”
Steal thisRequire every new headcount request to first justify why AI can't do the job.
Story
Shopify was born because Tobi couldn't get a job in Canada
Harley explains Shopify's origin: German immigrant Tobi Lutke couldn't get hired without working papers, so he started a snowboard shop in 2004. He built the store software in Ruby on Rails, and by 2006 it was clear the software was a far better business than the snowboards.
“But someone had told him that, okay, you can't get a job, but you can totally start your own business. And so being in Canada, he decided he would start a snowboard shop. And this is 2004.”
Story
Competing with Walmart from tax law class
Harley's 'holy shit' moment: as a 21-year-old law student he built a Shopify t-shirt store ('Smoover') during tax law class, and within weeks his biggest competitor was Walmart Canada. That a kid in class could out-compete a giant retailer revealed Shopify was far more than a software company.
“And my biggest competitor after a couple of weeks was Walmart Canada. It was great. I made some money. I was able to pay tuition and help the family. But the fact that I, as a 21-year-old law student, was able to build a company sitting in class that competed against the largest or one of the largest retailers on the planet at that time, that felt incredibly different.”
Number
COVID compressed a decade of e-commerce into six months
Harley says the pandemic delivered 8-10 years of acceleration in digital commerce in roughly six months, with e-commerce jumping from about 15% of total retail to 25%.
$25
E-commerce share of total retail (up from ~15%) · percent
“which is that we've had like 8 or 10 years worth of acceleration in digital commerce as a percentage of total commerce, um, in the last like 6 months. So we've gone from like 15% of, of e-commerce to total retail to like 25%.”
Fact
Influencer brands aren't promo products, they're real businesses
Harley argues that creator-owned brands (Kylie Cosmetics, Jeffree Star, Drake's OVO, Yeezy) are deeply misunderstood as mere brand extensions. They're built by people who understand their category cold, and unlike Michael Jordan licensing his name to Nike in the '80s, today's stars can own 100% via Shopify.
“If Michael Jordan started the Jordan brand today, he would own 100% of it, just like Yeezy does, just like Drake owns OVO, just like Kylie owns Kylie Cosmetics. But because the Jordan brand started in the mid-'80s, he was a licensor of his brand to Nike. He received a royalty. The royalty was substantial, but he had no choice but to work with Nike on this deal because fundamentally Nike had add the means of manufacturing and the means of distribution.”
Billy
Fashion Nova's Richard backed Cardi B before anyone knew her
Harley marvels at Fashion Nova founder Richard Saghian, who discovered Cardi B and planted small investments in her before she was famous. Because he was part of her story when no one else would back her, Cardi gives Fashion Nova attention and loyalty no money from Pepsi can buy.
“But Richard, the genius of Richard is that he discovered Cardi B before anyone even talked about it. And he planted these seeds, these small investments with people that he thought had the potential to be the next biggest pop star in the world. And because of that, what you end up with is a relationship between Fashion Nova and Cardi B that is unlike any other relationship from any other brand because Richard was part of Cardi's story.”
Take
Dropshipping is a process improvement, not a get-rich scheme
Harley reframes dropshipping against its course-hype reputation: it's simply a process improvement for retail that lets entrepreneurs access the market in a capital-light way. He says it's misunderstood, and as a business model it's very effective.
“I think dropshipping is being completely misunderstood. Completely misunderstood. Dropshipping is a process improvement for retail. That is all it is.”
Idea
Sell a course on the skill you already have
Harley's pick for an underrated business: take a skill you're already good at and sell a course around it. His example is an Italian grandmother who taught tourists to make pasta, then pivoted during COVID to selling pasta-making lessons online via Shopify.
“I think something that's underrated is people taking a skill set they currently have and that they're really good at, something that's unique to them, and actually selling a course around that skill set or selling something that helps people acquire that skill set.”
Steal thisPackage the skill you already have into a course instead of chasing a new product to sell.
Idea
Sell a course on the skill you already have
Harley's pick for an underrated business: take a skill you're already good at and sell a course around it. His example is an Italian grandmother who taught tourists to make pasta, then pivoted during COVID to selling pasta-making lessons online via Shopify.
“I think something that's underrated is people taking a skill set they currently have and that they're really good at, something that's unique to them, and actually selling a course around that skill set or selling something that helps people acquire that skill set.”
Steal thisPackage the skill you already have into a course instead of chasing a new product to sell.
Idea
How a content brand should sell product: ask the audience, sell the AV kit
Harley's playbook for a media brand monetizing via commerce: don't slap your logo on a t-shirt. Ask your audience what they need, then sell something genuinely useful. For The Hustle's desk-bound entrepreneur audience he suggests a $250 camera/mic/setup package or an Ember mug collaboration.
“And if there is a way for you to put together a package for $250 that comes with a better camera, a better microphone, and a better setup, that may be something that works really, really well.”
Steal thisDon't sell logo merch; ask your audience what they need and bundle a genuinely useful product around it.