On the web:
Follow Brandon at:
To join in on future Q&A, check out our Product Manager HQ community!
Select questions and answers from the AMA:
In a platform like Shopify, how do (or did, or would) you draw borders around different parts of the application to segment out products? Would you mind giving specific examples?
Great question. Right to the meat! At the highest level we break up the platform as the ‘Commerce OS’ portion (the APIs that drive basic Shop management) and everything that will consume that. Examples are Channels (places people can sell like their own online store), or Services (drop-shipping like Shipwire).
As a PM in a massive shopping platform, I’m sure you all do very little/big improvements to increase conversion. Can you name some changes that you made that you ‘thought’ were minor but ended up being a major differentiator? Or simple things that ended up having a massive impact? Any other tips on funnel conversion?
The quickest I can think about is we tested 3 field signup vs. 10-field to see if less friction would increase conversion. It had the opposite effect because the ‘extra work’ increased the investment into the product early so that it motivated people to continue to check out the rest of its capabilities. The biggest lesson is that it wasn’t just friction to consider, but at the heart, it’s the motivation a user has to continue and there are lots of ways to increase that.
Can you speak a bit about the product management frameworks you use?
The single most constant one I use is, for every decision, optimize an equation that is saying “for every unit of time the team has, how fast can we progress the vision”. This lets you balance things like technical debt vs. shipping quickly, and defining the MVP.
How are the growth teams structured/organized at Shopify and how do they work with the other teams like marketing/engineering/etc.?
Growth teams are structured into top of the funnel teams that work on content (written and video), SEO/SEM, tools, and then product growth teams that work closely with PMs to communicate the product best at launch and then work to iterate and grow the product after launch (the often ignored but critical part of product development)
What does the product macro process look like? How does an idea get shipped?
At Shopify, it’s very organic. Literally 100 slack channels with everyone constantly spewing ideas across the product. The ideas that win are the ones that can’t be stopped because they’re too obvious or compelling. We have a simple heuristic for individuals trying to add to the product roadmap. “Do Things, or Tell People”. Either make a prototype to show what you mean, or write a project brief and start to socialize it around the company.
We do have the idea of product themes, however, which are very long-term goals, almost “things that should be true in x years” that help guide our focus. Like “We want to be 50% of commerce in India by 2021”. But no more granular than that.
I’ve just started my journey to becoming a PM, and I’m trying to get my head around (1) what it means to be a PM, (2) what qualifies an exceptional PM, and (3) what areas to deep-dive into before I make my first pitch for a PM role. Any advice for someone just starting off?
- Help your team ship the right product in the context of your users, your company, the market.
- A good PM I’ve written about (https://.com/swlh/mvpm-minimum-viable-product-manager-e1aeb8dd421). An exceptional PM I think matches the hard skills in that article with high levels of Leadership and Empathy. They can inspire.
- Skills-wise, same article as above. Other than that, try to come to the table with a specific problem/domain that you have strength in from experience and pitch that as the best place to start.
Hi Brandon, how do you take future feature decisions for an e-commerce platform? Are they more data-driven or design eccentric with data to guide? How do you prioritize these?
For product iteration on existing products, it’s mostly data-driven and based on qualitative feedback. For brand new stuff, it’s qualitative from customers but also strongly based on our own opinion of how the world/technology/market is progressing. We feel it’s our jobs to anticipate how the world will evolve on behalf of our customers.
Does being located in Toronto have any advantages/disadvantages in terms of talent available for the product and engineering teams?
Advantages – Far cheaper labor costs (driven by lower cost of living relative to valley) while still an incredibly deep pool of talent from Waterloo, UofT, and more recently, the growing tech ecosystem. On the negative side: Fundraising. Here, Canadian investors are typically far more risk-averse relative to US counterparts and more short-term thinking. Also, the serendipitous nature of biz dev in the valley is not nearly as strong, and the access to a customer base that’s always willing to try new things is a net disadvantage too
With the high-level product breakup, how do you manage backlog that bleeds across multiple products (Channels, Products, commerce OS)? For example, assuming each team has a dedicated engineering team, does each team sprint on it at the same time, or do some lag behind?
Shopify is interesting in that it’s purposely architected to exist in a state of chaos. The reason simply that tech moves too fast and we should always be willing to change our focus rather than over-plan. Currently, we have about 80 teams running nearly autonomously, all with the power to deploy at their will. There is no master backlog, there is not even a standardized team-level product development process like Scrum. The idea is simply to start treating everyone like adults, hammer home the vision and the goals, then trust each other to use judgement to know whom to align with when building.
What will e-commerce look like in 10 years?
10 years is reallllly far away it’s hard to even imagine. I’d say in 5 years, VR shopping is the fastest growing, and anticipatory shopping (AI knowing you need toilet paper and auto-shipping it to your place) starts to take over staple purchases. And drones, lots of drones.
Any thoughts on returning to entrepreneurship after Shopify?
Absolutely. Tobi, Shopify’s CEO says the best way to leave Shopify is to start your own company. Every-time I think about Tunezy (my first ), having now gone through half a decade of Product Management afterwards, I think we could have reached the same point in about 1/3 of the time.
What’s one thing you do at work that you think makes the biggest difference in terms of impact on the team or product?
Biggest impact is probably helping to nurture developing PMs and building a subculture in the product team to challenge each other constantly that we’re working on the most important thing and being ruthless about where we focus… we have significant influence as a team on what 1500 people work on and we shouldn’t ever forget that and get lazy in our decision making
After a prototype or project brief breaks in, what are the next steps taken by the product team?
Start to flesh out the MVP, build the smallest thing possible and get it out to users. Once you get some product/market fit and traction, double/triple down on the team to accelerate the impact of the product. If it doesn’t work out, the harder but equally important thing is to stop working on it and move onto better opportunities
As a company scales getting wide alignment seems to be a challenge, curious given the growth at Shopify how you manage to keep everyone aligned?
It’s really really hard… ultimately I don’t think everyone can know what’s going on at all times, even the CEO. It’s just too complex. We tend to focus on promoting heuristics, strategy, and patterns internally in order to scale out alignment.
As an individual team member, you should be able to answer why what I’m working on matters, how we’ll measure success, and what tools/resources I have to get my job done. We try to ensure everyone knows that.
When you’re growing really fast, the other vector to attack is on-boarding. Make sure it’s impactful and is giving new employees the most important context so that they can autonomously act but in the of the vision.
Can you offer an example of heuristics, strategy, and patterns?
Sure, heuristic would be like – “design the UX for the 80% use case, but ensure the 20% is enabled through APIs so that developers can build for the niches.”
Strategy would be like – “we need to become the best messaging platform for commerce.”
Patterns would be the design and technical patterns that teams can use to move fast while maintaining consistency in the back and front-end.
Being an e-commerce platform, what are some features that created the most engagement in your product?
Great question Rickard, selling stuff is really hard for merchants starting out. So we do our best to constantly inspire and motivate them to continue and not give up. We have this part of the merchant-journey call the AHA moment, which is the first time a merchant gets a sale from a complete stranger.
To them, it’s validation that their product resonates with the world and their engagement goes up 10x immediately after. Knowing that, we try to do everything we can to get them their first sale through data insights to content about how to market products to integrations with ad platforms – and then make sure we celebrate it when they do.
One of the funniest in recent memory is adding a cha-ching sound whenever a merchant received an order. Engagement is off the charts when they get that sound for the first time.
“It’s purposely architected to exist in a state of chaos” — that sounds like something we could benefit from understanding better. Can you add any more detail to how this was achieved, or point me in a general direction (ie, Netflix/Chaos Monkey)?
The chaos as an operating principle stems from our realization that the world has operated under the assembly line model since the revolution. This was a good model when product development was complicated (needs 10,000 steps) like a car. But today, we characterize product development much more as a complex system vs a complicated one.
The core difference is that in complex systems, you cannot derive causation (doing A will result in B) because there are just too many inputs going on at the same time. Users of your product are constantly evolving themselves, being given new versions of your product and about 100 different alternatives every week. They have so much choice and low switching costs.
The logic goes that in this world, it’s better to just act fast and focus on creating really quick feedback loops (data) to identify if something you’re doing is at least correlating to the behavior you want to grow, and then doubling down on it, as opposed to over-planning based on a world that is changing too fast.
What’s your favorite interview question to ask a mid-level PM? Or a new PM?
Mid-level -> What’s been the project that’s defined your career as a PM to date? How did it change you as a PM and what you think about your role?
New PM -> What do you do in your spare time? What side projects have you done? (looking for people passionate about building and have entrepreneurial spirit)