First On: Product with Benji Portwin
My guest today is Benji Portwin, a Product leader who is no stranger to hypergrowth. Benji’s held Product leadership positions for some of Europe’s fastest growing startups including Spotify, Monzo and now accuRx. Benji is a highly empathetic leader who takes a people-centric approach to running an often opaque function. In our conversation Benji demystifies the Product function, challenges some frequently cited startup wisdom and shares what he thinks makes the great leaders truly great.
I’d like to start with understanding how you define product. It feels to me like it is universally appreciated and almost as universally misunderstood.
I think there are a few points of confusion to Product. The job means different things in different organizations. A Product Manager (PM) in one company can do a vastly different job to a PM in another company, even in the same space. There can also be confusion in how a job is advertised. There's a rhetoric around PMs being “mini-CEOs”, which I don’t think is true at all. If you're the first PM in a 10 person startup, there's a real CEO and it’s not you. If you are at a big organization and you have lots of autonomy, you’re still not the mini-CEO and actually behaving that way won't lead to good things.
I think what it really comes down to is responsibility and accountability. What do you want you Product people to own? If you can agree on that and know what that means before hiring them then you'll be okay. The main friction point comes from someone walking into a PM role and finds a different experience to what they were sold. There is a lot of crossover between the Product, Engineering, Commercial and Design functions so you need to have a plan for how all that works together.
We've been very explicit with what a PM means for accuRx and if you want that kind of role then come join. If you want a different PM role in the way that Facebook offers, that's fine as well. Neither’s wrong or right but it does impact if it works well for you. To give a concrete example, one of the big challenges in Product is who owns delivery. In most organizations Product teams will not own delivery and in fact are very averse to it. They want to own the vision and some metrics or KPIs, but actually making sure a product gets built is given to an engineering manager or technical lead. At accuRx we ask that PMs be accountable for their team, to design and to ensure it gets built. That's quite a lot to ask of but we're very explicit about this upfront and hire for a certain set of skills and motivation. Before I joined we had some real friction on this.
How would you advise founders on figuring out what Product should mean for their particular company and culture?
I’ll tell you the story of when I joined accuRx. In the beginning of any startup there is really one Product team; less than 10 people doing one thing, trying to find product-market fit. Once you find that fit you have to scale. Jacob is incredibly good at setting vision. I haven't met many people in my life who are better than him at thinking where we might be in two years time. He thinks about it all the time. He is not, however, good at delivery or thinking about what the next month looks like. Laurence early on was probably that guy. Laurence is very much a jump head first, no job is too small, bias for action type. Early on in a small team things worked but I imagine it was a frantic way of getting stuff done and that doesn't really scale.
We had a gap in the org. Jacob sets a really strong vision of our part to play within healthcare. Laurence is the CTO and owns all technical decisions like hiring and tech stack. Who owns the next six months? Who's gonna make sure we have really clear priorities, really clear goals and really clear org structure?
When I joined they had a very founder-driven, feature-led approach; “What is our two year vision? What are the next 10 things we need to ship? Team One you do X, Team Two you own Y. Let’s come back together in six weeks.” That can work, for a time, but it puts loads of pressure on Jacob and without him the whole company falls apart. So when I came in I kind of wanted to break up that system and give more autonomy to the PMs. Autonomy isn’t the whole job, but I wouldn't think that in the long run constrained PMs would be happy or stay in the organization. Churn, in my opinion, is the number one metric we take into account. It's what kills companies
If you're telling Team One to build X and they don’t get to choose what they build, they don't really care how it performs. It may work really well but it won't necessarily solve the users problem because you haven't asked them to do that. You just asked them to build a particular thing. If you want a group of people to be accountable, they have to have a say in what they build. That doesn’t mean they need full autonomy or will refuse to work on anything they’re given. But there has to be a collaborative process with those involved.
So within that version of Product Management how do you measure yourself? Or what are you measured on?
Number one is always people. Day to day if the people are happy and healthy and doing great work, I'm doing okay. Because as the organization scales, more and more you have to rely on those people doing really great things. PMs must make 20 decisions a day minimum and I’m not going to be available for most of them. So making sure your people are really great is number one.
Secondly is around delivery. I’m ultimately responsible for delivery and it’s important that Jacob and Laurence can come to me if teams aren’t performing. To do that I need collaboration between Engineering and Product. I explicitly say in our job descriptions that relationships are really important. At every level of Product and Engineering we measure collaboration skills.
You’ve mentioned org structure a few times and I've always been very interested in that topic. It feels like where Product sits in an organization is almost the defining element of what makes an Apple vs. a Microsoft vs. a Netflix. In your experience are there models that work extremely well? Are there models that don’t work at all?
One thing I’ve learned is that the team of Leads you have at <30 people will dictate a lot about what your structure will end up being. One of our close to founding members is the Head of User Research. Recently we decided to have the Design report into User Research. Now you could make the argument that Design should sit with Data and Data reports into Product, but that’s the way it shook out. So our new designers will have a much different experience as a result of this org structure. When you come into a company, whether a designer or a researcher or whatever, your position in the organization will be dictated by who those first 10 people in the company were. That’s not quite fair but it is worth noting. When you join a company it’s worth asking for that story, “Who was your first Product person that you hired? Where do they sit?” I would definitely ask this if I ever went back to a PM role.
Another thing is matrix management. It’s quite common to have tension between Engineering and Product. Again, we have Product be accountable for delivery but actually we found that there was equally strong sentiment for Engineering to be accountable. So we actually have double coverage. Our PMs are tasked on three things and our Technical Leads are tasked on three things and two of them are the same. You might ask what happens if things go wrong, who do you blame? The answer is both of them. We don't care, you're both gonna get blamed so work it out. We believe in the principle of “leadership in pairs”.
What makes for a great Head of Product? If you had to stack rank the skills that really matter, what would they be?
I think this is super specific to the organization. It really depends on who the founders are, what they do, who your bosses are, etc. Within Product Management the number one thing that I care about most is, do you get people? There's no degree in Product Management, which is a travesty by the way. But if there was a degree I would base it in psychology. The stuff you learn from behavioral economics and psychology are the most valuable things for you as a PM, in my opinion. They’re also the things that are hardest to teach. It’s so important to understand your user, which is really just getting people. You’ve got to want to go out of your way to learn what that person thinks - what they’re struggling with, what they like, what they don’t, etc. - and to really enjoy the conversation.
Secondly, you've got to want to make your team work well together. If your team is healthy and everyone's enjoying the work and collaborating, you can basically take the week off. But when a team isn’t working you won't sleep and it's going to be fucking exhausting because you should be wearing that responsibility. The skill set required for your team to work well together is the same I mentioned before, do you care about them as people? Do you want to get to know them? Do you know about their kids? What are their interests outside of work?
The last thing is that you absolutely need to be introspective. A lot of new managers that I see struggle with the new reality of how on display they are. When you move from an engineer to an engineering manager, as an example, everyone looks at you differently. Everything you do now has 10x more impact than ever before. When you do something it will get mirrored and what you won’t do won’t happen. You have to hold yourself to a higher standard than everybody else to make sure you can get everyone to work the way you want them to.
When we last spoke you mentioned the idea of a gap between “the values on the floor vs. the wall”. In other words the explicit values listed from HR and the informal behaviors you see everyday. To some extent it seems almost inevitable that you will have some deviation on that so I’m curious how you’ve tried to correct this when it happens.
I think it's one of the most dangerous things that can happen in an organization, because it reveals that what you're saying isn't what’s happening on the ground and that shows that you’re disconnected. That’s how you lose power and influence.
I’ll give you one example around goals. Several months ago we made a decision to try and help the National Vaccine program. We are now the main way you get a vaccine booking slot in the UK. We’ve done over 20 million bookings through a platform we only started building six months ago. This required shifting a huge number of people and loads of time, well over half the company for months. This wasn't part of our vision or annual goals, we just saw a huge problem that needed to be fixed. We had some real tough moments and a lot of long hours but we ended up with a product that's going great. What we didn’t realize was that over half the company wasn’t working towards any of our strategic goals. Again, we can’t become disconnected. So in the short term we just call a spade a spade. “Hey, we realize this isn’t what we said before. We're not idiots. We see what you see, we're not out of touch. These are still our long term goals. But right now, of course, there's a short term goal around vaccines.” And actually we went further to change our company goals because we felt it’s important for our team. We want everyone to be able to say, “I'm actually helping work towards company goals.” so everyone feels motivated and connected.
One of our values is balance and it’s easily the value we're living up to least right now. Building this vaccine delivery product really pushed us to our limits. We built the first version at a breakneck speed and it worked but the workflows weren't intuitive and the design needed rework. It wasn't a good product and we recognized that. So what do you do? Everyone works all hours of every day and every night to make it better. Our team just worked themselves into the ground for two months solid, including both founders, our full support team, our operations team and everyone that surrounds Product. At some point we had to take a look at ourselves and ask, is this the new normal? Is this us now? Balance to us literally reads as “we make sure you leave work at a normal time, have a healthy lifestyle and don't work weekends”. But this was categorically not the case.
This all led to a very interesting Leads meeting where we saw two options. One was deciding what needs to change and stop certain behaviors. You know, “we’re no longer doing X”. The other was just to remove balance as a value and acknowledge that if you come to work here we will expect a lot of you. In the end we changed some behaviors and I'm glad we did. We adjusted how we communicate internally, we set the norm of notifications being turned off on weekends, things like that. We also tried to establish a clear view of stage gates. For example, we need to get the vaccine tool to a point where support queries are below X. We stated that once we reach that gate there would be a significant change in the workload. And if you, our employees, don’t see that change you can hold us to account.
It strikes me that the understanding of values is very difficult to assess in a short period of time. How do you assess those types of skills in an interview process? How do you build that system to ensure that's properly vetted?
I definitely agree that this is incredibly hard to assess. We do two things. Number one is that we've chosen to hire more junior people. We're trying to build a certain culture around Product Management here. A lot of our interview process is focused around mindset and coachability. Does a candidate realize they know fuck all? We're all making it up as we go every day and we look for those who can admit that. Sure they’ll have experience and perspective and ideas to improve, which will be valuable. But we want to see someone who wants to join accuRx by coming in, watch how things are done now and make themselves a better PM.
The other thing is to make the person do the job. In our final round we make candidates run a planning session with a real team. We realized that facilitation is a key skill so we give them a real team, access to their backlog and an outline of what they’re trying to achieve overall with their product. They go in and facilitate and we take notes. We need people who can hit the ground running so the bar for us is if we can give them a team on day one.
As you certainly know there is a lot of interest in Product careers from young talent, but anecdotally it seems like very few know where to begin. How would you advise the current crop of inexperienced but hungry junior talent around Europe that want to pursue a career that looks and feels like yours?
I’ll first tell you what I wouldn't do. I wouldn't I wouldn't go and do an MBA. If you want to be a PM at Amazon then get one. They’ll hire you everyday and they have, I think, 10,000 opportunities right now across the globe. And to be clear, that is a perfectly fair route into a startup Product role after a few years. Or maybe you just love Amazon, which is also fine. If you want to be at accuRx it's a lot harder right now because we have a minimum experience bar. But we have started an APM (Associate Product Manager) program for people who haven't got formal product management experience but they have skill and intellect. Google and Facebook do this in a big way. Many smaller companies are starting to run these programs. If you're currently in a role and you want a career change, find a way in your current role to get exposure to Product.
So as an example. We hired an APM who was previously working in a circuit board manufacturing company where the product development process is three years. It's not at all like how we do Product, but they did get exposure to a form of Product Management. So when we were interviewing this person we knew that they hadn't worked in a small org or run software development teams. But we also knew that they got people and we could tell they knew what they didn’t know. They could demonstrate success. They could tell a good story. This is all core to being a PM. I think there are hundreds of APM roles in London each year.
I really wish there was an undergraduate degree in this. I've actually looked into creating a course at a few universities as an elective.
For my last question I wanted to touch on the topic of hiring with diversity in mind, which is an area that accuRx has done some great work on. Diversity is very top of mind for so many people but everybody struggles with it. You have a deep calibration of culture and mobilizing groups of people so I'm curious, what approaches have you found useful in this area?
When I was at Spotify we had a lot of people working on diversity and inclusion efforts. A lot of discussion centered around, “we can remove names or gender or backgrounds from CVs.” Yes, you can do those things but it's a lot of work and doesn't necessarily buy you as much as you might think. At some point, you’ll need to know who that person is and at that point the bias can come back. It may help with CV screens but at some point an interview is going to meet the candidate and think, “Oh, you’re different to me.”
There was a woman in the D&I group at Spotify who I found very inspiring. She was an openly lesbian, black woman, product leader who made it very clear that she was fucking tired of this conversation. She said, “If we have two candidates that are exactly the same, then in that case, hire the woman or the black person or whatever. But that's never going to happen. We're never going to be in a room and have this situation. So let's stop fucking talking about it. And if that's what we're talking about, I don't want to be in the room. I'll put my energy elsewhere. If, however, we're actually going to say this is an element we actually care about in the same way we talk about product intuition or collaboration skills or communication skills, then I’m interested. Are we actually saying we are going to look for more candidates like this and hire these people? Because that’s what moves things.” From that day onwards we agreed, as a group, that we were going to change the diversity of our business unit. Now you can debate quotas and I do appreciate that there's leftover effects to it, but if you actually want to make significant changes to an organization that’s already in a bad [diversity] place, there isn’t really a more effective lever to pull. I know a lot of amazingly smart people with all the best intentions who have made lots of small changes to their hiring process with almost zero impact.
At accuRx the founders really care about this. They looked around the room and realized they didn’t have a diverse organization. And they knew that if we don't try and change it now, when it gets really bad with scale it's incredibly hard to undo. At some point you have to say, “it's going to take us longer to fill some of these roles, which will put us behind on our goals and puts the company at risk. But we’re trading that risk because if we don't change our diversity now we’ll never fix it later”. And over the last nine months we've made massive, massive improvement. On gender we're now better than 50%, female-to-male. On minority backgrounds, we've made double digit improvements. We still struggle with educational background and that's the hardest one.
We waited three months to fill an open engineering manager role because we really wanted to have someone with a different background to other engineering leaders. If you’re patient and you care about it, you can solve it. But it does feel fucking hopeless in the thick of it and you’ll think about giving up five or six times. But what we’ve found is that eventually those people you hire will refer other people and eventually get to a state where you can lean back to referrals and network, which is the natural and way to source for anyone.
First On is an attempt to uncover functional greatness by asking the experts themselves. I hope you enjoy the mini-series and with enough interest I may extend this into a more ongoing effort. If you have any functions / roles you’d like covered or have ideas for future guests please let me know.