The Secret to Building Great Subscriptions with Leading Silicon Valley Product Builder, Tom Willerer

Before you start building your subscription product, make sure you really, really understand the ongoing problem you are solving. So says product leader Tom Willerer. Tom, my guest today, has worked with some of Silicon Valley’s most renowned companies, like Netflix, Opendoor and Coursera. He’s also an Entrepreneur in Residence at Reforge and a Venture Advisor at VC firm NEA. Today we’re talking about how to define your forever problem so you can build a forever transaction. In our discussion, we share the secrets to building subscription products, how to build conviction you’re on the right path and when to commit to scale.

—

Listen to the podcast here

 

The Secret to Building Great Subscriptions with Leading Silicon Valley Product Builder, Tom Willerer

Before you start building your subscription product, make sure you understand the ongoing problem you’re solving, so says product leader Tom Willerer. Tom, my guest, has worked with some of Silicon Valley’s most renowned companies like Netflix, Opendoor, and Coursera. He’s also an entrepreneur in residence at Reforge and a venture advisor at VC firm NEA. We’re talking about how to define your forever problem so you can build a forever transaction. In our discussion, we share the secrets to building subscription products, how to build conviction you’re on the right path, and when to commit to scale. Welcome to the show, Tom.

Thank you for having me. I’m excited to talk with you.

I want to talk about your experience designing products, particularly products that are designed for recurring use and even recurring revenue. You told me once that the place where you like to start is focusing on understanding the problem, which is not necessarily what everybody does. What do you mean by that? Why is understanding the problem the most important thing to do in product development?

You’re touching on one of my favorite topics right out of the get-go. I love it. I would give a few responses to this question of why focus on understanding the problem first. One, it’s where everything starts. If you don’t have a deep understanding of the problem you’re solving for your customer, whether that’s a B2B customer or a consumer customer, you’re not going to be able to design solutions or a value proposition that is long-lasting and is something they’re going to want to pay for, subscribe to, and keep going on. A foundational step is one easy answer.

The other answer I would give to this is it’s often a step that we forget. We assume everyone agrees with the problem that we’re trying to solve. We instead go to what is the fun stuff, which is coming up with solutions. That’s where we all end up starting. We skipped the first two, which is to find your narrow set of users and deeply understand their problem. Instead, we go straight to the solution piece and go, “I got an idea. My idea for how to do this is XYZ. Let’s go and try it and test it.”

What can end up happening is no one went and did the hard work to understand your customers, their problems, and the jobs that they’re trying to solve irrespective of your solution. Therefore, you’re now throwing spaghetti at the wall to see what sticks because you’re in solution mode. What happens is you end up not going super deep in any one solution space because it’s not attached to any problem. You’re like, “Did this work? No. Throw it away. Try another thing. Throw that away.”

Instead, what I would advocate for is deeply understanding qualitatively and quantitatively your customers’ problems, their needs, and the jobs that they’re trying to solve, and because of that, you can have the conviction to keep going after that problem even if solution one that you try doesn’t work, “That didn’t work but we believe this is the right problem to solve. Let’s go try solution two.”

Your trajectory of ideas becomes more long-lasting and feels less “random” because you’ve nailed the problem. That’s where you’ve developed your deep conviction. That’s where you’ve developed alignment with the organization on where you need to go spend time. That then allows you to have a little more flexibility on the solution space but even if you get a couple of them wrong, you’re still going to keep whacking at that problem and try to find the right solution for it.

That’s so interesting. I had a little epiphany, which is that your point about focusing on understanding the problem before you start throwing out solutions is probably also a good management technique. I find that so often, people almost don’t let you explain the problem before they have a solution for you.

It’s 100% true. It’s a life lesson in some ways. There’s some psychology that I won’t be able to describe but I know I’ve done it, which is I zoomed straight to solution space first. I get enamored with my ideas, I fall in love with them, and then I realize, “It didn’t work. Move on to the next thing.” I never went back and did the hard work and the foundational work to understand the problem. It is the first step.

Sit with the problem wherever you are, especially if you’re trying to build a product solution. Can you give us an example? You’ve worked at some fantastic companies leading product teams at Coursera, Opendoor, and Netflix. Can you share an example of a time either that you sat with the problem longer than people wanted to and it led to a great outcome, or maybe the reverse where you jumped to the wrong conclusion and had to go back and re-examine the problem?

I can give you an example of a failure mode because these are very interesting ones to learn from. It was from my time at Netflix. I’ll rewind the clock to probably 2010, 2011, or somewhere in that ballpark. The context here is we were starting to grow internationally. We launched our streaming service. It was very early. As everyone knows, or maybe they don’t know, Netflix transitioned from DVD to streaming. We launched the streaming service. We were trying to grow internationally. We were nascent in that.

We had seen Facebook grow and get very large. We have seen companies like Spotify, which started in Sweden and then moved internationally. A lot of how they grew was by getting their audience of subscribers and listeners to share everything that they’re listening to on Facebook with their friends. That then created this viral loop of those people wanting to listen to that song, seeing that, and then going and signing up for Spotify. We at Netflix saw that. We fell in love with the business problem, “How do we grow faster?”

It’s a good nuance on the problem. When I’m talking about a problem, I go deep into the consumer problem because the other thing that happens is you go deep into your business problems. This is what we did at Netflix. We went deep on the business problem, “How do we grow from an acquisitions perspective without spending exorbitant amounts of dollars? If we could figure that out, we would create an awesome business.” This was solving the business problem and maybe a go-to-marketing challenge that we had. We were using examples that we have seen in the market.

Go deep on the business problem. Ask yourself, How do we grow from an acquisitions perspective without spending exorbitant amounts of dollars? If you could figure that out, you would create an awesome business. - @twilly Share on X

We did this. We went out. We tried it in the US and went more aggressive in some smaller countries where we hadn’t grown as big yet because there was some risk to doing this, and we wanted to experiment pretty liberally. With that, we had people sign up. They connected their Facebook account and Netflix account. That became their account creation tool. When they started watching anything on Netflix, it is shared on Facebook. That was interesting. We did see some growth advantages to this.

We saw some member advantages to this because they were able to see, “What are my friends’ favorite on Netflix? That gives me good suggestions of things to watch,” but we saw a lot of complaints from our members because when we went and did the research to understand the problem that this could be solving. We realized there was no customer problem that this was solving. This created problems for them because people didn’t expect everything that they watched to either be shared automatically or at all. They didn’t want it or expect it. When it happened, they were like, “I don’t want my Facebook feed to be clogged with all the episodes I watched of Say Yes to the Dress, The Bachelor, or whatever it is.”

When you proactively share stuff, you curate much more. My view of what I watch is different than the reality of what I watch. What I want other people to think I watch is different than the reality of what I watch. I’m going to tell people about that obscure indie film that is going to win awards because it makes me look a certain way or a documentary that makes me look a certain way but the reality is I love watching Jeopardy or Say Yes to the Dress because it’s fun and entertaining content, Cake Boss, or whatever these shows are. That’s not what I want to project out to the world.

That became a real mismatch there. It illustrates for me a few things. One is it’s easy to fall in love with the wrong problem first, which is the business problem. The reality is you have to start with what is the customer’s problem and then figure out what is the solution, how that solves the problem, and how we as a business create something of value for them and then extract some of that value, not all of it. Two, we went straight to solution mode.

 

It was like, “Look at these other companies that we want to emulate. They’re doing it. We should do the same thing too. Let’s go and try that. It’s got to work for us too.” We didn’t go and live with the research and honestly live with the qualitative feedback because after we saw some negative sentiment come, we probably did 30 qualitative interviews. It was universal. People didn’t want it, nor did they expect it. It was obvious here. This wasn’t like it needed a multi-month ethnographic study to figure it out. It was like, “Go talk to 30 people about this, and you’re going to get a pretty good response.”

There are a couple of interesting things about that story. First of all, most people think of Netflix as phenomenal at understanding the problem. I appreciate you sharing one misstep. I think about understanding the problem, “I don’t want to run out and get a movie. I can’t remember what movies I want to see when I get there.” Solving those two problems of convenience and choice were such great examples of things that they have done.

The other thing that came up for me when you’re telling the story that I also get, having consulted at Netflix for a couple of years even prior to that, is people often will start the conversations or prospective clients or existing clients of mine that are focused on subscriptions, “We want the Netflix playbook.” To me, that’s also dangerous as your story, including the part about, “Spotify was doing this, and it worked. Facebook was doing it, and it worked,” but Netflix is neither Spotify nor Facebook. You have to also understand what’s unique about your product, your audience, the use case, and all of those things, which sometimes organizations miss the boat on.

One example of this, which may relate to Coursera and may bring up another story for you, is a lot of professional development and educational companies want to take a page from the Netflix playbook but there is a huge difference between watching streaming content that’s entertaining and watching it for work because your boss told you to. Few people binge LinkedIn learning courses unless they’re trying to get a job but it’s not the same as, “It’s already 4:00. I’ve spent the whole day watching Cake Boss.”

It’s almost like smart people don’t always eat their veggies even though smart people know that’s the right thing to do. You’re right. Netflix is seen as a company that had smart people running it but didn’t always do things in the way that we should have done them. There are several examples of it. These are worth highlighting because they get glossed over when you see up into the right chart of all the moments when you made mistakes. It’s good for me to go and remember those because I can remind people that think, “If I make a mistake, maybe I’m not good enough.” The best companies do this type of stuff too.

When you were talking, it did remind me of another example. It’s a little more nuanced. It’s from Coursera. I’ll go into that one. When I joined Coursera, we had a business problem and a customer problem. It was the way in which we ran our courses. We took the old model that existed in universities. Coursera is taking university content and putting it available online for people to access all over the world. We were digitizing a lot of what had been locked up behind the ivory towers of some of the world’s best universities and then making that accessible to everyone around the world.

In the first incarnation of this, we took the old model, which was, “Classes are available in the spring and the fall.” We said, “Put it on Coursera, and we will run them in the spring and the fall.” That does make sense in a world where the professor and TA time is scarce but in a world where all of this is prerecorded, it doesn’t make sense. We also solved the customer problem, “I don’t always have time to learn in the spring and the fall but I do want to learn this content. Sometimes summer or winter break is better for me because I’ve got that gap in my schedule and my time.” It was like, “There’s a real customer problem here.”

We also saw a business problem, which was we weren’t able to grow as fast if our content was gated on a spring and fall release schedule. It needs to be smoother than that. It should be more available than that for growth to take off. I had come from Netflix. It was like, “That’s an obvious thing here. The content should be available on demand.” We rebuilt the platform and redesigned things so that the content was available on demand. What we missed we were hearing from our university partners, which was, “It’s not just about the growth piece or the top-of-the-funnel part. It’s also about getting people to complete a course.”

When we launched this, we saw phenomenal numbers at the top of the funnel. We got lots of people engaged. It validated our hypothesis that people wanted to learn no matter what time of the year. This content should be available but we also made it completely flexible for them to move through at their pace. It was like Netflix. Click and go whenever you want to go. You move at your pace. It’s on-demand and available to you whenever you want. We thought that would be the right benefit to both get growth and get people to give the flexibility to finish the courses.

People want to learn no matter what time of the year. Content should be available, but you also want to make it completely flexible for them to move through at their pace. It’s like Netflix. They can click and go whenever they want to… Share on X

It turns out it did get growth but didn’t solve the problem around getting people to complete the courses. We had heard this critique from our university partners because they are in the job of educating people. They know what it takes to get people through these courses. What they said was, “You’re losing all of the cohort elements or the elements around keeping people together through something, and because of that, no one is going to finish.” That was one.

Two, it was like, “Learning is a little different than entertainment, to your point. You need a little more structure to do the learning than you need to sit down, zone out, and watch Netflix maybe unfortunately for the world but that’s how it works.” We were pretty bullheaded and thought, “We came from Netflix. We will figure this out.” We did it and learned that part 1 of the hypothesis was right and part 2 was dead wrong. We got many more people at the top of the funnel and fewer people at the bottom of the funnel completing.

It led to an iteration because we felt confident in the customer problem that we were solving, and we got part of it right. It led to an iteration where we kept the availability but made frequent train schedules. On popular courses, the start date was every week, or you could do it twice a week. With that, you had a cohort effect on people that you were moving through this with. Even more importantly, you had deadlines and due dates that you were implicitly signing up for by starting it on a certain date. We could say, “Week one, this date. Week two, this date. Week three, this date.”

It went from the important but non-urgent quadrant to the important and urgent quadrant by giving it a structure around due dates and deadlines. We learned that you need the availability but you also need the structure around it to get people to invest in the learning to move through this at the right pace. With that change, it’s still the way Coursera works to this day. We captured the best of both worlds when we drastically increased completion rates through courses. We were able to hit our growth goals and make the content available to customers whenever they wanted it.

This leads directly to the next thing I wanted to ask you. How does this problem solution balance play out with subscription? One of the things that I heard you say is there are acquisition challenges. How do we get people to sign up? There are engagement and retention challenges. How do I get them to stay? I’m thinking about companies that sell a big product one time and they’re done as opposed to the ones that depend on many small payments.

If you can get them to drive the Lamborghini off the showroom floor, that’s their problem. You don’t have to worry about how engaged they are with the car once they get it home but if you’re relying on them to continue consuming your content or using your service, both pieces start to matter. I believe it makes it even more important to understand the full problem and its ongoing nature.

I don’t mean this in a slight. I do this all the time. It’s easy to fall back on one-dimensional thinking when in many ways, these problems are very multi-dimensional. In the Coursera example, we were more one-dimensional in thinking growth was what we needed to optimize for but it wasn’t just growth. It was sustained growth. You get people starting but because they’re not finishing, that was a failure. It would have been easy to say, “Ditch the experiment. It didn’t work. Go back to the old model,” because the old model was objectively better than the new model. We hadn’t thought through the multidimensional problem, “How do we keep people going?”

We thought too simplistically about that part of it and dismissed some of the qualitative signals we were hearing from otherwise who were experts in this where we were not and fell back on our experiences in a model. It’s like the Spotify and Netflix examples. Netflix didn’t map completely to Coursera. We took too many lessons from it and said, “They all map perfectly.”

We should have said, “There’s inspiration from this but we’ve got to go deep and listen to some of the experts here.” That would have probably would have saved us 6 to 8 months when we churned through the next iteration of this. In the grand scheme of things, it didn’t diminish Coursera’s ability to find success but it was slower and required more resources because we didn’t go about it in the multidimensional way of thinking about both growth and retention at the same time.

The thing that you did well is to get clear on what the problem was for the customer. September and March are not necessarily the best times for me to start a new course but what you didn’t do was come up with the right solution the first time, which you pointed out early on is okay because that happens. I would love to get your thought tactically on how you do these two things. Do you feel like you could have spent more time on the problem? Do you think that you were too quick to jump to a solution that had been successful in another organization, in this case, Netflix?

When you’re launching a new thing, there are many ways to validate it. One is through direct-to-consumer interviews. Another might be through expert interviews. You could do surveys to get data on this too but if you take those first two of going deep with your customers, we did that. Customers liked the idea of it being available whenever they want it. They also liked the idea of it being on demand. Interestingly, they were like, “That’s right.”

We didn’t set out to do expert interviews but because of how we were a platform company in a marketplace, we needed to get our university partners, most notably professors, to buy into this, and because of that, they gave us their expert opinion. We heard them say, “People aren’t going to complete this. They need these other elements.”

Customers aren’t going to say that to us. You need an expert to tell you that type of thing because they can tell you parts of it but they can’t always give you the full picture of it. The experts were telling us. We didn’t listen to it. We thought we were smarter than the experts. If we would have coupled those two pieces the first time, we could have nailed it but it took us one time of being wrong to realize, “We should have listened to people that know how to get people to complete educational content and coupled of those two pieces together to get to the right solution.”

There’s more than one source of information. The customer doesn’t always know best. Maybe talk to some people who have seen the customer in action in a similar environment.

These expert interviews can save you a ton of time because they have seen a lot. I would recommend this when you’re launching a new thing or launching a new company. This is what a venture capitalist would do. You’re going to talk to twelve experts. It’s a hack to bootstrap your knowledge quickly.

 

It’s about being patient and spending more time with the problem. You’re listening to a lot of experts. You don’t have to take their advice. Sometimes the experts are wrong. Your professors might have said, “Online content isn’t as good as real-life content. This is never going to work.”

One example of exactly that where they were wrong was they said, “It needs my involvement in the course. I’m the professor. I’m not the watchmaker putting my watch out in the world for anyone else to tell time with. I have to be actively involved or else people aren’t going to understand the concepts well enough.” There’s no doubt in my mind that if we were doing a one-on-one with a professor who’s an expert in their field, I would learn well from them but it’s unfeasible. It won’t work. You can’t scale that.

It wasn’t true that in the Coursera model where you’re getting thousands or tens of thousands of learners at a time in these courses, very few were having any live interaction either through videos, in a forum, or anything with a professor or any faculty member associated with that course. It was not impactful at all but they thought it was an impactful piece. We were pretty confident that was not the case, and that turned out right.

Judgment is a massive component of this because there are those times when you have to be non-consensus. That’s a very important thing because if you only go for consensus ideas, you’re going to water it down to the point where it’s not super interesting. Is what you’re doing differentiated enough in the world that everyone else hasn’t already thought of it and exploited all the benefits from it?

I imagine at some point a long time ago, you were an individual contributor but a lot of your work over the last several years has been either as a manager, as a product leader, or at Reforge as an instructor, an educator, a coach, and an advisor. In those roles, you do sometimes have to take a strong point of view and make the call. I would guess that one of the things that you’re going to say about that is that if you can come up with a small experiment to see if you’re right rather than betting the farm on it, you have to make the call but if the call has smaller consequences, or you can figure out a way to shrink those consequences, that’s also going to help you.

I can give an example. It’s a non-subscription-related example but it is a consumer-oriented example of this. It will illustrate it well. At Opendoor, we launched a buyer-oriented product.

Can you explain briefly what Opendoor is?

Opendoor is an online real estate company. We started by helping homeowners sell their homes more quickly. The way that we did that was they would come to us and type in their address, and we would give them an instant offer. It wasn’t a Zestimate, which was an estimate of what their home was worth. It was an offer on their home. We would close in as little as ten days. It was trying to give the certainty of an eCommerce type of experience to a very manual process of selling your home. It’s an uncertain thing to sell your home. You got to play the market. You don’t know how long it’s going to take.

Having the certainty of an Opendoor offer became very popular. That’s how Opendoor scaled to some success. With that, once you have a seller, the majority of sellers are selling because they want to buy another home. It was natural for us to get into helping them buy their next home. We use that one piece of helping sellers to move into an adjacent part of the job to be done for them, which is moving and getting into their next home. In that, we acquired a company called Open Listings which had a nicely done product that helped them browse homes. It had a Redfin model. It was a discount brokerage.

They would help you buy a home and take a lower commission than what a normal agent would. They tried to give you a little bit more DIY tools to go through it on your own, which appealed to a subset of users but not mass scale. We had a foundation. When I joined the company, we had that foundation on the buyer side. We were trying to figure out what is the right problem for us to solve for buyers.

In this one, we identified a number of problems. I can enumerate a few of these. One is finding a home as a buyer is hard and complex. That is true but it’s already solved. Zillow, Redfin, and others have created amazing apps that everyone should use to find a home. It would take us hundreds of engineers over many years to recreate that experience fully.

That’s a third thing to do with understanding the problems. Is this a big problem? Are there any solutions already?

It was finding a home but that problem was already solved. An additional element is how underserved is this problem in the markets. Let’s pin that. We can come back to that in a minute. The second problem that we identified was you need some expert guidance. If I’m a buyer in the market, I might not be buying in a market that I know. I need someone who can say, “The school districts here are good. This is a nice neighborhood. I’ve sold homes here to people. They’re happy.”

A real problem that people have is a giant purchase. The average home price is hundreds of thousands of dollars. You don’t want to spend that willy-nilly but there are three million real estate agents in the world. They probably are your neighbor, cousin, and relatives. In some way, you know them. They’re going to be way better positioned to give you that expert advice around this large-scale purchase than a faceless company would. That was like, “It’s probably not the right one for us.”

The third problem on this was winning the home. Once you find the home and you know it’s the right one, you have to then put in an offer that is going to win that home. Regardless of whether you’re in a competitive market or not, it’s nerve-racking for a buyer when they’re putting in an offer that they think they might not win the home because even if it doesn’t have 10 offers, it might have 2. You want to be the one that they select.

We thought, “There is a problem that we think we can solve here because we already do this on the seller side. We use our money and our certainty around buying a home to get sellers to sell to us. What if we adapted that idea and backed buyers’ offers with our cash, effectively turning them into cash buyers?” We have seen a ton of research that existed out there that showed cash buyers can be up to two times as likely to win a home. We got excited about this but that was one solution.

 

The problem though was, “How do you help them win a home?” We had another solution that we were piloting and thinking through the details of as well, which was around, “What if we split the equity of the home with them?” In other words, they put $20,000 down. We put $20,000 down. We were part owner of that home with them. That either lowers their monthly bill or helps their money go a little further.

It’s another flavor of helping them win a home that we felt we were well-positioned to go after. We parallel-tracked two of these concepts on that one problem. To the point of starting small, we built a PDF that described this the best that we could for customers. We didn’t even build a website. We put it in a PDF, tried to describe it as best as we could, and did 100 UserTesting videos of each of these concepts.

You used the company UserTesting.

We created the collateral and put it out there. People go through. The PDF looked like it was a webpage. They navigate their way through and tell us what they were thinking. We have prompts for them. Through that, you’re able to get pretty mass-scale qualitative research and get a good sense of stuff. That was one. The other is these are complicated financial products. We had to find partners and get this off the ground. As we went through both of those, us being an equity partner with them was hard for customers to understand. They were like, “What happens when I sell the home?”

It led to 5,000 questions. The back end of us pulling this off was not straightforward. There were a bunch of complications to it versus, “We will back you with our cash. If your equity financing falls through for any reason, we will step in and buy it for you.” It was like, “That makes sense. That gives me a superpower.” You could get the SNIP test quickly through the interviews.

That’s one step of starting small. The next step was maybe we put the equity co-investment piece on ice a little bit for now. We don’t throw it away completely but we say, “Pause there a little bit. We might keep things going but we’re going to put most of our effort into this project and this innovation around backing their offer with our cash. Let’s try to launch it. Let’s try to get 30 people to sign up for this.”

It was super manual. We would get someone to come in as a buyer. We would then call them, “It’s Opendoor here. We’ve got this new offer. Here’s what it looks like. Would you like to try it?” Yes, no, maybe. We would get ten of those. We pulled them off. They worked. We said, “Let’s go to 50.” It got to the point where I knew that in order for us to go big on this, we needed to flip the switch and say all of our offers because we were tracking the percentage of offers that people want to be backed by our cash versus not backed by our cash. It became almost a confusing customer experience, “Is this what you stand for? Is this not what you stand for? Is this an option? How many options do you have for it?”

At some point, it was like, “We have started small. We have built over time. My conviction in this whole thing has grown. I’m not 100% certain that this is going to work at scale but I’m pretty high certainty.” As an exec, I had to say, “We’re going all in. I want all of our offers backed by Opendoor cash. Team, reorient everything around this concept. The old concept is no longer the winning one. We’re not going with it anymore. This is it.”

That’s how I thought about going from an idea and inception to a little qualitative research project. You can honestly do lots of interviews with UserTesting. Launch it. Call people, “Would you use this?” and then go a little bit bigger. You’re never going to get enough data that says, “This is the right thing to do. You should do it with 100% certainty.” At some point, you have to make the call. I made the call, and we went all in on it.

That’s such a beautiful example. I also wanted to call something that you said at the very beginning of the story. You went down this path because you found an adjacent job to be done. You continued on that customer’s journey with them and said, “What’s the next thing I can do for them?” That’s important because you have this customer you already know. You’re trying to build a forever promise.

You’re trying to solve more of the problem for them over time and be with them on the journey. You spent so much to acquire them in the first place. Let’s do more for them. It’s the way that you talked about starting small, experimenting, answering some questions, and mitigating risk. At some point, as a leader, you say, “This is where I’m comfortable jumping in. Let’s pour the concrete. Let’s do this with rebar. Let’s make it permanent.” That’s a good story.

The frequency of use of the products is interesting. With Opendoor, it might be every decade you use it. We needed to maximize the adjacencies that we could capitalize on from a business perspective immediately. If we spent whatever dollars on the acquisition, we want to get that customer selling to us but then we can get them to buy with us, do title with us, and finance with us.

If we could make that a seamless experience for them and differentiate it for them, it’s both good for them and good for us because we don’t have the advantage of going and getting them to subscribe to us or buy again with us in a day, a month, or a week. It’s 1 shot 10 years later, whereas Coursera might be every 6 months to a year product that they’re engaging with. It’s more intense. They might subscribe for a year, and then they’re done, “My needs are good for now.” They might come back three years later and subscribe for a year, whereas on Netflix, they’re subscribing. If they don’t watch it on a daily basis, they’re going to turn. The business problems are very different there. That’s an excellent point.

If we could make a seamless experience for the customers and differentiate it for them, it's both good for them and good for us because we don't have the advantage of getting them to subscribe to us or buy again with us in a day, a… Share on X

It’s interesting. I love that you have three examples. Thank you for working at three good and different places. It makes for a much more interesting and fun interview. I want to circle back to something that you said when you were sharing that excellent Opendoor example. Not every problem is the right problem to solve. How do you know if it’s not the right problem to solve? What are the key categories of a bad problem to solve?

This is an excellent question. It’s hard. It’s an easy one to screw up for sure. One framework that I’ve been introduced to that I found very helpful in this is after you’ve identified the problems, the next step to do for that is to determine whether that problem is over-served in the market or underserved. For instance, when you go back to the Opendoor example if you look at finding a home, that is a problem. Seeing all of the homes that are available in my neighborhood at my price range with how many bathrooms I want in the school district that I want and all that is a foundational problem.

If it didn’t exist already, it would be a big problem but it is very much over-served. Realtor.com has a great app. Redfin has a great app. Zillow is a great app. Everyone has a great app for this. There are some table stakes elements that we need to do if we’re trying to build a buyer experience but it is not the area where we should double down and innovate on.

Making that clear to the team is hard because we as builders want to make everything great. It should be top-notch, and as an exec to say, “Make that good enough,” it’s a weird but important message to send to a team because we also have to realize we have scarce resources, and we can’t make everything as good as it possibly can. We have to be good enough so we don’t get fired for that thing. We can’t have people firing us because we’re not sufficient enough on the job to be done of helping them find it but that can’t be where we’re innovating and trying to get people to hire us.

That over-served or under-served is important to think through. What we found was that this notion of backing your offer with our cash and making you a cash buyer was completely underserved. If you weren’t rich or if you didn’t have a rich family member, then you did not have access to the capital that you needed to be able to remove the contingencies of financing. We also knew that you could though because there are rich people that have rich family members or have money themselves. They win at double the rate. It was like, “That’s the goal.” What sellers want is a cash offer.

That’s why Opendoor is so popular because we give them a cash offer. There are no contingencies on it. It’s not going to fall through. We knew all those things. When we looked at it, we were like, “There are some startups doing this but there isn’t an incumbent who already exists in the space that makes this irrelevant for us to try to go double down on.” That’s how we use that concept. It was powerful for us.

That’s great. When you’re thinking of a feature, there are features that drive acquisition, “We’re going to make it possible for you to give a cash offer.” There are probably people who said, “We weren’t interested in Opendoor, and now we are.” There are features. If you don’t have them, people will either not engage with you or they will leave you. There are features that deepen the relationship with you. If it’s not one of those things, then maybe you don’t do it but it’s hard.

I thought where you were going to go with this with your team is, “We can make it better.” They put so much effort into making something better but the other product is just fine. It has a market share. It’s not a very considered category like laundry detergent. Once you choose your laundry detergent, it doesn’t matter how good the one next to it is. You’re like, “This one is good enough. It works. I know how to use it in my washer. I know how it works on my clothes.”

Understanding what to say no to as a product leader is as important. In the world of subscriptions, in particular, people have a tendency to keep layering in more stuff because they can. A lot of the stuff isn’t that great, it’s not needed, it’s confusing, or it becomes somebody’s job to keep that maintained, which means that they can’t work on the next feature that’s going to drive acquisition or keep people from canceling. It’s very hard to kill your darlings.

We used to talk of this and use a concept at Netflix called scraping barnacles. It’s almost like you think of it as you’re building the ship, and the ship is trying to move as quickly as possible. Some of the things that you’re doing are going to become barnacles. It’s hard to scrape that barnacle because our natural inertia is to add new things, not to remove old things. Removing old things can be important because it helps you go faster.

 

To your point, you’ve got scarce resources. You’ve got to keep a subscriber. You’ve got to be zeroed in on what their problems are and solve them on an ongoing basis. If you don’t because you’re instead focused on redesigning a feature that doesn’t matter, then you’re going to fall behind because someone else is going to figure that out.

I love scraping barnacles. There’s always some customer who loves the barnacles, “These are beautiful. They add charm to the boat. I love boats with barnacles.”

Typically, that customer is also a board member. It becomes hard to scrape that barnacle. We had that issue at Netflix for the queue or the list of movies. That came back at some point but there are profiles. In the DVD world, having these profiles became too much but then it came back. Some of these concepts are interesting in that they work for a period of time, and then they don’t but then there’s a reason to revisit them. None of it is forever. You might end up adding it in a different way and then reimagining it. That makes more sense.

This is great. I have 1 million gems. I’ve learned a lot. I want to close out with a speed round if you’re up for it.

Let’s do it.

What’s the first subscription you ever had?

Probably Netflix DVD.

What’s your favorite subscription?

YouTube TV. I’ll give you a different one. It’s a terrific product. I was not a big cable bundle TV watcher but they have reimagined the user interface in ways that are delightful. If you watch sports, they give you stats and summaries. You can watch the recap of the game in three minutes. They have done a terrific job on some of their core user experience innovations. They have made that invaluable.

What’s something you have learned in your new roles either as a teacher or as an investor that you did not know back when you were running products?

I didn’t value the power of frameworks as much when I was running products or as an individual product manager. I thought, “I’m going to reinvent everything from whole cloth,” but I’ve learned that there is real power in frameworks because it’s a tool, not a rule. It gives you a starting point that you can adapt. I’ve learned how important that is to have a helpful starting point to adapt from. As long as you can remember that it’s a tool, not a rule, it’s a thing that’s helping you accomplish something, not a rule that you need to blindly follow. They’re so valuable.

Besides you and me, who’s one LinkedIn thought leader you follow?

Elena Verna does a wonderful job on the growth topic. She’s done a ton of B2B SaaS scaling. She’s helped a bunch of companies. She’s funny. That’s an added benefit.

She’s great. She’s a past guest. Your choice, what’s the best or worst product experience you had this week?

I’ll go with the worst. I’m coaching Little League for my son. He’s in Little League Baseball. There are a bunch of logistics you have to coordinate. There are all these apps to communicate practice times with parents. My daughter does soccer. They use a whole different app. There’s Byga, TeamSnap, and GameChanger. All these different apps are confusing and hard to use but as an administrator, I went in thinking I was super organized. I created the team, put everyone in it, and put our first five practices in there so people could start using it.

The lead came through. They created their version of my team and put everyone and all the games and practices in there. I couldn’t merge these two. I couldn’t delete the one I created. Amazingly, I couldn’t delete it. I had to do some hack where I put void and then removed everyone manually from one. I know I probably got an edge case but it was a very difficult edge case to be in. It was a bad experience.

Especially as a proactive team administrator.

I was trying to do the right thing, and it screwed me over.

The lesson there is to do less. Tom Willerer, thank you so much for being a guest. It was a pleasure to have you.

This was a delight. Thanks so much. I appreciate it.

—

That was product leader Tom Willerer who has led product teams at Netflix, Opendoor, and Coursera. He’s currently Entrepreneur in Residence at Reforge. For more about Tom, go to Reforge.com. Also, if you like what you read, please go over to Apple Podcasts or Apple iTunes and leave a review. Mention Tom and this episode if you especially enjoyed it. Reviews are how audiences find our show, and we appreciate each one. Thanks for your support. Thanks for reading.

 

Important Links

• Tom Willerer, EIR, Reforge
• Reforge
• Opendoor
• Coursera
• NEA
• UserTesting
• Subscription Stories Episode 25: Freemium, Free Trial and Free Surprises with Elena Verna of Reforge. When Does It Make Sense to Give It Away, and When Doesn’t It?

About Tom Willerer

Tom Willlerer is currently EIR @ Reforge, Venture Advisor @ NEA, and an advisor to several tech companies, helping them with product strategy and growth. Prior to this, he was Chief Product Officer at Opendoor and Coursera. Tom joined Coursera after having served for six years at Netflix, Inc., where he became one of the company’s most highly regarded leaders in his role as Vice President of Product Innovation.

Tom co-led the product management team at Netflix and was responsible for setting the vision and managing the teams responsible for Product Marketing, Messaging, Social and Individual Profiles. Among his accomplishments during his time at the company, Tom pioneered ways for members to exchange recommendations with friends, and helped Netflix grow from 8 million U.S.-only subscribers to 40 million subscribers in 41 countries.

Tom holds a business degree from Indiana University, where he was a pitcher on the varsity baseball team, and received his Masters in New Media Studies from DePaul University.

 

Love the show? Subscribe, rate, review, and share!

• Here’s How »

Join the Subscription Stories Community today:

• robbiekellmanbaxter.com
• Twitter
• Facebook
• Robbie Kellman Baxter on Instagram
• Subscription Stories: True Tales from the Trenches on Apple Podcasts