a16z Podcast

Building Products for Power Users

David Ulevitch and Rahul Vohra

Posted June 23, 2020

As more digital natives have entered the workplace, they have brought with them the expectation that their software should both be a joy to use and allow them to be power users. That is, users who configure and control it to better serves their needs. And often, these digital natives aren’t just aspiring power users, they are also prosumers, who can and will pay for a premium experience. But first generation SaaS products have often struggled to deliver the experience these users crave.

For today’s founders and builders, how do you get the user experience right when a product has to delight your power users, while being something a less savvy user can pick up and learn?

In this episode, a16z general partner David Ulevitch and Superhuman founder Rahul Vohra discuss how to build products that can turn any user into a power user. The conversation touches on themes from David’s recent talk on products that adopt developer tools, like the command palette and keyboard shortcuts, to improve usability, and Rahul’s talk on how to apply game design principles to product design. They cover how to onboard users to drive virality, when to expand to a second product, and how to use pricing to position a premium product.

Below we’ve put links to the articles and books that Rahul and David mention on the podcast: 


David: As somebody who thinks of himself as a power user, I always felt like SaaS tools just really never allowed me to climb up that learning curve to being a power user. There was a whole sort of first generation of SaaS tools that really would hallmark their benefits of it’s easy to use, it’s really simple.

I always think back to Microsoft Excel versus Google Sheets, Excel is easy to use, but it allows you to be a power user if you want to be.

Rahul: I would argue that, actually, companies like Salesforce and Atlassian with Jira were catering to power users. It’s just that the complexity and the configurability were so strong that the companies almost lost sight of the fact that it should be enjoyable to use.

I think what’s changed recently is the ability to then also package it in a consumer-ish fashion.

The challenge with tools like Salesforce and JIRA as well as really any other piece of enterprise software, or more generally, any piece of power user software, is that they often end up looking inelegant and becoming visually cluttered.

Many prosumers don't necessarily have the time to become power users, but if you can get them there, they then represent a very powerful market. -Rahul Vohra

David: You’re totally right that Salesforce and JIRA are very configurable. But in fact, most companies that deploy Salesforce end up hiring a third party, or hiring a specialized person inside the company just to administer it just because it’s so configurable. And I don’t think anybody that ever uses Salesforce would call it a joy to use.

That lends itself to a conversation around design. And for those that don’t know, Superhuman is often driven through two mechanisms. One is called a command palette, which is sort of like a command line that pops up wherever you are, and then through lots of keyboard shortcuts.

And so, what gave you the confidence to build a prosumer application that really encouraged or forced people through their onboarding, to become power users? How did you know that was gonna work, and how do you think about usability in light of that?

Rahul: Those who know me well know that I started a PhD in computer science. I spent a lot of time thinking about human computer interaction and this idea of ease of use. And I came to the conclusion that as a whole industry, we had played way too hard into learnability and into ease of use, and that we were really doing our customers a disservice. We are assuming they’re not as intelligent as they actually are.

So, we aggressively push in the other direction. What we worry about is can it do all the things that you want it to do? When you’re a power user are you really, really fast? And we’ll solve the problem of you becoming a power user through other techniques.

I had this belief, having grown up as a developer that keyboard shortcuts were the way to go. But I knew that people generally don’t use them. Less than two percent of the entire Gmail user base have even used a keyboard shortcut once.

So, how do we solve the problem of teaching users? Funnily enough, I realized the answer had been staring me in the face for my entire career as a programmer. It’s this idea of a command palette. Every text editor for programmers from VI through Emacs through Vim to things like TextMate and Sublime more recently, and every big IDE has a command palette.

We’re able to add arbitrary amounts of functionality without ever making it more cluttered, because we found an entirely new way to interact and to operate the software.

You simply hit one master keyboard shortcut, in Superhuman that happens to be command K, and then you just type in the thing you’re looking for. At that point, you can hit “Enter,” and it will do it. But not only that, it will also teach you the keyboard shortcut for next time. So, you start up this mastery curve.

David: I had an interesting reminder of exactly what you’re talking about during this quarantine. I decided that since I was stuck at home, I was gonna go buy an Xbox. And when I have been a gamer in other parts of my life, it’s always been on a PC, where I’ve used a keyboard and a mouse. For the first time, I had this controller, and all I could think about was, “Man, this is terrible compared to gaming on a PC, where you have so many buttons at your fingertips and you can drive shortcuts and remap your keys the way you want.”

And I actually found that the Xbox controller was nowhere near as fun as playing a game using a keyboard and a mouse.

Rahul: So, in game design, we have this notion of a lean-back or a lean-forward experience. And TV is a lean-back experience. Now, with console gaming, it really does depend on the game. You’re probably not going to play “Call of Duty” leaning back. But you might play “Animal Crossing” that way.

And so as a product designer, I think there’s just a set of questions you have to answer.

Are you building a lean-back versus a lean-forward experience? How do you want people to feel? What is the primary thing you’re optimizing for? Is it the feeling of power? Well, then, you’re going to want to let customers remap their keys and run things their own way.

David: You take a very consumer-like view to how you present, essentially, an enterprise software product, So, how do you think about the lines between consumer and enterprise software? And does that relate to that learning curve of letting something be powerful but also be a joy to use?

No feature or product has ever really had a viral loop of greater than one for any meaningful period of time. -Rahul Vohra

Rahul: In practice, we’ve ended up with a very interesting hybrid of both enterprise and consumer. Around 60% of our subscriptions are enterprise. They are expensed or paid for directly by companies. But the other 40% is what you might call consumer. They are paid for directly by end-users, albeit for their work email.

All of our customers come to us as prosumers, but not all of them come to us as power users. Historically, many power users simply have not been able to pay. That’s why they became power users in the first place, because often they were writing their own software. And many prosumers don’t necessarily have the time to become power users, but if you can get them there, they then represent a very powerful market.

And we see a big part of our responsibility, when they buy our software, to help them become the power user that they aspire to be. So, we put a lot of effort into helping folks level up. And believe it or not, we get many customers who come to us who have never used a keyboard shortcut in their life.

And this is one of the main reasons why we onboard users. For each new user, we do a 30-minute one-to-one video call. We teach faster workflows, powerful shortcuts, And using innovations like the command palette, we teach you how to teach yourself so you can quickly become a power user.

David: Do you think that the way that you’re doing your onboarding would be another reason why you couldn’t have gone down the mass adoption consumer route?

Rahul: The constraints around whether or not you can onboard fundamentally come down to unit economics. You can do it so long as your churn and retention are good enough.

There is an argument to be made that if something has more of a consumer price point, it becomes economically untenable. But in practice, it really does depend on your churn. Let’s say you had a $10 per month price point. If you could show me that your churn was zero or perhaps even negative, then by all means, I would say yes, you should onboard.

Zooming out all the way, the question is, are you making net promoters or are you making net detractors? And I don’t like to leave this to chance. Personally, I would rather invest every dollar we have in helping every single customer become a net promoter, because that is the flywheel behind any company that has reached massive scale.

And this was a lesson that I learned relatively early in my career. So, when I sold Rapportive to LinkedIn, I had the very fortunate privilege of working for Elliott Shmukler, who was our Head of Growth, and took the company from 25 million members to 250 million members.

In our first one-to-one I sat down and I said, “Hey, Elliot, please teach me everything that you know about virality.” And he said, “Well, lesson number one is that there isn’t such a thing as true virality. No feature or product has ever really had a viral loop of greater than one for any meaningful period of time. Even at LinkedIn the address book import has had a viral factor of 0.4. And even at Facebook, the most viral company of all time, they were not able to sustain 0.7 for more than about seven months.”

At the end of the day, it always comes down to word-of-mouth organic spread, it comes down to brand, it comes down to building a product that people love.

At the end of the day, it always comes down to word-of-mouth organic spread, it comes down to brand, it comes down to building a product that people love. -Rahul Vohra

David: I wrote a blog post last year about this idea of product-user fit, and you’ve talked about how do you satisfy that high expectation customer. One of the things I’ve noticed over the last 18 months or so, as I spent some time with, seed-stage companies that are building a product that they want to be widely available, is that, when they look at their numbers, the averages are terrible. So, if they look at retention, it looks terrible. If they look at, number of minutes per day in the application, it’s terrible.

But then if they focus on the users who are using the application every single day, those users are using it constantly, they’re spending hours a day in it. And so they’ve started to try to figure out how do you focus on what’s making those users really happy.

Rahul: I see this all the time in founders. They’ll come to me lamenting, “Hey, my month-to-month churn rate is north of seven percent. Surely, we are screwed.” And I’m like, “Wait, hold up, hold up. Let’s segment these users out. First of all, let’s discard all users who didn’t activate, because we need to separate activation problems from ongoing value proposition problems. And then next let’s segment out users who have different personas, different job titles, those who are paying personally versus those who are expensing it.”

Because if you’ve built something even half interesting, there will be a set of users who find it compelling. And the whole purpose of the product-market fit engine is to identify that segment, and then to systematically expand it over time.

David: When users love a product, you’re constantly getting suggestions and requests for features or changes for what would make it better, how do you balance those priorities for new users while you’re still trying to serve that core base of power users?

Rahul: In my experience, startups tend towards one of two extremes. Some teams, especially those which are driven by vision, tend to constantly double down on what users love, whereas some teams, especially those which are driven by data, tend to systematically address user issues.

In practice, you need to keep vision and data in very careful balance. Now, doubling down on vision is truly an art. This is where we get to exercise the so-called product instinct. And this is all about talking to your users, and figuring out the most impactful ways to give people more of the feelings that they crave.

Now, addressing user issues is much more straightforward, and for that, we use a very simple cost-impact analysis. We then pick the projects, which are both low-cost as well as high-impact.

David: But then there must be things that are hard and take a long time. How do you deal with things that are high-cost, high-impact?

Rahul: So the answer is yes. I have a metaphor that I wrote down the other day that I’m looking for.

David: Did you use a command palette to find it?

Rahul: I used Command E to open the document, so actually, yes. So, what do we do about the big things that are high-impact?

If you've built something even half interesting, there will be a set of users who find it compelling. -Rahul Vohra

We get asked this question a lot by customers who write in, and they say, “Hey, why haven’t you built my pet feature?” that incidentally, would take us a year or two to build. And so I think of our team capacity as a glass jar. Launch projects are big rocks, medium and small projects are pebbles of various sizes.

The most efficient way to fill the jar is a combination of big rocks and small pebbles. First, add in the big rocks, and then the pebbles can fit in and around them to fill the jar. If you were to add the pebbles first, fewer big rocks would fit in. But if you added big rocks only, there would be unused space.

So, it’s sort of classic Silicon Valley tradition that when a new engineer joins a team, they should ship on the first day or two. That’s a really good opportunity to work on a small pebble, to fix a bug, to get something out into production.

Every quarter, we also identify one massive rock and a few medium sized stones and a large number of small pebbles. We know roughly what order we’re going to do them, but what the small pebbles give us is the opportunity to be flexible with the plan in the face of unknown unknowns.

Let’s say we run into an API issue, or there’s a scaling problem, that throws a wrench in our otherwise beautifully designed plan. Well, then we can move the small pebbles around and still ship with high cadence, despite the changes.

David: So, how do you know when your product is mature and you can sustain the sort of innovation on the core product, while expanding into a new product or a new surface area? When do you make that decision, and how do you think about those things?

Rahul: It’s not: “is the product mature?” It’s: “is the company ready?”

I think it’s a fairly common mistake for founders to see that a product is working well, and then they move on to starting the next one. But they often forget just how hard it was to get to product-market fit in the first place. Now, imagine searching for product-market fit again, whilst maintaining product-market fit on the first product, presumably in the face of competition, and on top of that, balancing resources and attention between the two.

For us, it was absolutely the right strategy at the start to sell to prosumers, the right thin edge of a wedge, to paraphrase C. Dixon.

As for whether you lower the bar to appeal to customers, the question I would ask is this: do you have to? Because it comes with significant cost, attention, time, money, brand.

Another way of asking the same question is: can you be a billion dollar company without lowering the bar? Let’s assume that at the point, we’re a billion-dollar company, our valuation is 10 times our run rate, and so our ARR is $100 million. That would then imply we have 300,000 subscribers at $30 a month.

So we ask ourselves, do we think we can reach hundreds of thousands of prosumers? And our answer is, absolutely, yes, we can.

So, if something is working, I now believe that it is almost always better to double down and do more of that thing than to spin up something new. When every single team is fully staffed; when you have an excess of resources and personnel; when, to quote Marc Andreessen, with his famous description of product-market fit, “Your checking account is overflowing with money because you’re just making so much cash.” That’s the right time to take on the focus risk and to start an adjacent product.

David: When I think through the iconic products of our time, almost all of those companies then tried to create adjacent products. And in fact, the two things exactly that you don’t want to have happen, happened.

First, the second products generally did not take off the way they had hoped. And oftentimes, the first product sort of stagnated. There’s products that all of us have used where now, maybe five-plus years later, they’re finally going back to the core product and just incrementally making it better. What they’re doing is expanding the surface area of that core product, as opposed to building these adjacent products that are not based on the same core use cases and audience of that initial product.

Rahul: I think when you can do this also depends on the structure of the business. Do you have a moat? Are there network effects?

Now, for most SaaS companies there are network effects, but they’re relatively weak by comparison to a Facebook or more classically, an eBay or a Skype. And so inherently, SaaS companies just have to spend more time on building features and staying competitive. Arguably, if you’re building, let’s say, a classic social network or a company with that kind of network effect, once you’re up and operating and at scale, you do have a little bit more latitude to go and try something else, because there’s inherently more defense built into the products and the ecosystem that you have created.

And for founders who are interested in sort of thinking about this a little bit more Hiten Shah wrote a really good article when the acquisition of Trello was announced. He posed the question, “Well, why did they sell even at half a billion? Why didn’t they keep on growing?”

And he posited that it was because the thing that they had popularized, the drag and drop kanban board, had simply become a feature of things like GitHub, and Asana, and Jira, and so it was no longer competitive. What they failed to do was to constantly build and focus on their core products.

David: Now that you’ve said you’re committed to doubling down on the core product and adding features that make it better, that opens up the discussion of what about packaging? What about pricing? Is pricing something that should be optimized? If you’re just reinvesting in your core product and adding features, does everybody get those features?

Rahul: It is impossible to truly separate pricing from packaging from positioning. Now, I’d go to a cocktail party, back when we used to have those, and I’d listen to people talking. By the way, this is a great way to learn how to market your product. They would invariably say two things. Number one, Superhuman is super crazy fast, and number two, Superhuman is $30 a month.

The second thing that they would say was the price of the product. In our case, the price is literally part of both package and positioning. And when founders ask me, “How do I figure out what my pricing should be?” My answer is always, “You should first figure out your positioning.”

So, we started with this article by Arielle Jackson who, for folks who don’t know her, she actually happened to be the first product marketing manager on Gmail.

She wrote this article called “Positioning Your Startup is Vital – Here’s How to Nail It.” She advises using a formula – and it’s a little bit of a Mad Libs game for a target customer who has a need or an opportunity. [Your product] is in a [product category] that provides [some key benefits], and unlike [competitors], [your product] has these [differentiators].

Folks will notice that was a very tight definition. You can then use that to figure out your pricing.

In the book “Monetizing Innovation,” the author, Madhavan, advocates for developing pricing alongside product in a way that supports the positioning that you’ve just come up with. Now, he covers a lot of ways to do this, and we used one of the easiest methods, which is the Van Westendorp price sensitivity meter.

In late 2015, we asked 100 of our earliest users these questions. Number one, at what price would you consider Superhuman to be so expensive that you just wouldn’t buy it? Number two, what price would you consider Superhuman to be priced so low that you would feel the quality wouldn’t be very good, and that would be concerning? Number three, at what price would you consider Superhuman to be starting to get expensive so that it is not out of the question, but you would have to give some thought to buying it and you still would? And number four, at what price would you consider Superhuman to be a bargain, a great buy for the money?

Now, most Silicon Valley companies orient around question number four, the bargain question. And that’s because most tend to be selling into a new market or a new buyer in an existing market. But for us, it made sense to adopt that premium position. And the price points that most directly supports the premium position comes out of the third question, when does it feel expensive, but you’d still buy it anyway? For us, the median answer to that third question was $30 a month, and that’s how we picked our price.

Once you’ve picked the price, it’s always a good idea to do a quick gut check on market size, can you grow into a billion dollar valuation?

David: So, this has been an awesome discussion. We started at this really exciting place of really understanding the difference between power users and prosumers, and how do we turn prosumers into power users. And then how to make sure users are engaged and are net promoters, and ultimately, how you think about pricing and packaging.

So thank you, Rahul, it’s excellent that you spend your time talking about these things for other founders out there.

Rahul: Well, thank you so much for having me.