Breaking Even
Christian Crumlish–Break Even!
(upbeat music) - Hello, and Thanks for dropping in.
I'm really happy to be back at Web Directions, or I guess to have Web Directions come back to me, here in my home office in Palo Alto, California. My name is Christian Crumlish.
I'm a Product Consultant at Design and Product and this talk has you may have heard by now is called a Break-even.
Not breaking even, as I once thought it might be called.
That's back when I was thinking a breaking bad logo type thing, probably about a year and a half ago maybe. I like break even better.
Because it's both, it's a moment.
The break even moment in the life of a startup. It's a goal.
And for me, it's an expectation to everybody out there. Product people don't be afraid to think about the money and figure out the revenue.
So I'm gonna tell a couple of stories from my experience and try to share a little bit more about how you can break-even, how you can get a product to the break-even point. Just a little bit about me and why should you listen to me? My name, as I said is Christian Crumlish.
I have a background in product management, leadership and design management.
I spent some time at some big companies like at Yahoo, where I was the last Curator of the Design Pattern Library and the Author of "Designing Social Interfaces", Co-author with Erin Malone.
When I came down to Sydney for Web Directions in 2009, it was to talk about "Designing Social Interfaces". So that was a wonderful experience.
I left Yahoo to go to this up-and-coming internet company called AOL, where I was Senior Director of a Consumer Experience Team working on improving the product and UX practises at the company.
And it was there that I transitioned into product management.
Eventually spending some time as the Director of the Product Management Team for AIM. The late lamented AOL instant messenger that so many people first experienced online sociality or harassment (laughs).
In the early days of the internet.
I went from these big sclerotic companies to startups, where I was eager to get back to the internet pace that I was used to.
Where you ship, you commit code, you push your pages live, you would take a look at what's out there and fix it if it's mistaken.
I was Director of Product Management and User Experience at a company called CloudOn, which is a word processor for the iPad more or less collaborative sharing, editing environment, and gesture based tablet cloud, et cetera.
We sold the company to Dropbox and Dropbox Paper bears some of the imprint I think of the work that we did. And then from there, I spent four years at a company called 7cups, which is what I'm mainly gonna talk about in the rest of this talk.
7cups was originally called 7 Cups of Tea to mental health, behavioural health product that leverages self-help and also community care, a peer to peer support, people learning how to take care of each other, trains people on active listening and offers to find somebody to talk to you or chat with you, text chat with you, for free on demand within a couple of minutes. When you really need to talk to somebody.
We won a Stanford MedX Award for health systems design. We won a World Economic Forum Pioneer Award just last year and helped millions of people while I was there.
And as I said, I'm gonna talk about that a little bit more. Now I'm consulting, as I mentioned, I've got a company designing product I'm consulting mainly with startups, across a whole bunch of spaces, some in mental health, but some in civic tech and just other spaces as well. And I've spent a lot of this year working with a non profit foundation called Path Check on a contact tracing tool and open source product that preserves privacy while trying to help fight the pandemic.
So that's about me.
Let's get into the story and maybe talk a little bit more about 7cups.
As I mentioned, 7cups is a anonymous free service to provide emotional support to people online, 7cups.com.
If you remember the television show Game of Thrones, one of the actors on that show, Maisie Williams, I believe is her last name.
She played Arya Stark, and she was quite young when she started and she experienced a lot of online bullying. So for a teen or pre-teen girl, even a famous actress, it was upsetting to her.
And she did an ad campaign in the UK about bullying and specifically cited 7cups as a service that she used to get support, when she was feeling low.
We used to tell our volunteers, you know, you never know you might be given support to Arya Stark.
So stay on your toes.
So 7cups is really this experiment in trying to build a supportive online environment that was healthy and compassionate and not about tearing people down or competition or bullying and harassing, but instead about being compassionate towards each other and caring for one another.
And of course the challenge from a product point of view is how do you make a living doing that? How do you pay for that? Especially if you wanna offer it for free.
One of the challenges as a product manager is how do you make a living doing that? How do you make money? How do you sustain it? Especially if you wanna be able to offer this service for free, to a vast majority of your customers.
So 7cups was a very lean startup the whole time I was there.
It had raised seed capital after starting off at YCombinator.
And it preserved that sort of scrappy YCombinator, ethos of trying to focus on core metrics and get them to grow, and learn what what's working and be really honest about what's not working and faced the problems. We used to have in the saying if 7cups is the problem is the path.
Meaning, if you look at what the problem is and study it, the answer is right there in the problem.
And if you don't avoid the truth of what's actually happening, you'll find the path.
And that served the service very well.
Even when I joined...
I joined after about a year and a half into the company's life.
After I graduated from YCombinator.
And already as I said, it had an amazing product market fit.
I have this test I use for product market fit, and we used the highest level, which is what the customer crawl through glass to get the experience.
And that's really what I thought I was seeing at 7cups.
Especially because the product itself was a very early sort of cobble together, you know, brilliantly coded, but at the same time organically sort of sprawling assembled product that was figuring itself out, and did not show it necessarily, a sprawling product that was just figuring itself out and did not necessarily know how to present itself to a confused and new person trying to understand, you know, what's there.
Nonetheless, without being a perfected product without a great onboarding in place yet, without a lot of user experience improvements that we were able to bring to it over the years, already people were using the product and the user base was growing, and it was providing something people needed. So we knew that it had sort of touched a nerve and that if we could make it easier to use, easier to understand, easier to become a volunteer with et cetera, that we had huge potential to help millions and millions of people as we ultimately did. We tried a lot of things to help encourage people to participate and to take interest in their self-help. We have this concept of a tea cup that would sort of go down and level as your... If you missed a day coming back and would be restored, if you came to the site and took care of yourself, we would give you a credit for having a streak. If you were able to keep your tea cup full for, you know, a certain number of days in a row. At the same time, the growth was not necessarily that kind of hockey stick or L shaped curve, J shaped curve that people hope for, that looks like the investors are gonna make a tonne of money right off the bat.
There was growth but it felt like it was tailing off. We certainly worked very hard on trying to hack the growth, try to learn new ways to find more customers, better use of mobile, translating into other languages and reaching out to different age groups, doing a better job, of course of converting visitors and finding traffic. Nonetheless, it felt like there was some natural limits that might be in place, at least at the level of resources that we had, and trying to raise a series A without this sort of breathtaking growth was proving a challenge.
And one of the effects of not being able to raise that growth, the user base quickly enough or in escalating enough level, accelerating enough speed was to possibly try to figure out revenue, for two reasons. One is that the seed money we know is only gonna last so long, especially if we could not raise venture capital. So revenue would have to replace that.
And the secondly, if you're trying to raise a round, if you're trying to get VCs to put money into your startup and you don't have...
If you don't have exponential growth, that's going through the roof, but you have increasing revenue, of a few revenue is growing at an exponential rate on top of your existing user base, then you're demonstrating an ability to monetize your users and figure out a break-even point where growing the user base is gonna pay back in revenue, and investment in the company will sort of exploit that calculus that you're seeing, that you're now able to generate this many dollars per user in the lifetime value of a customer, as they say against the cost of acquiring the customer. So there's a number of ways to square that. But ultimately the pressure on us was to find revenue and grow revenue.
If we weren't able to grow our user base as quickly as everyone would like to see, out here in Silicon Valley.
So, you know, the basics were that we had the service that people could use, if they could find it. Anybody could ask for support typically took about two minutes for someone to show up, but people did tend to bounce in those two minutes, and there's questions of trust, who was 7cups? Why should I trust you? Is it really anonymous? Is it safe? Is it private? What are you doing with my data? Who's paying for this? So we felt that there was a lot of opportunity to tighten up the new user onboarding, and not only to drive basic growth at an engagement, but also ultimately to find a revenue path. So, as I said, we were under a lot of pressure to raise money.
And one of the things that I'm gonna do in the rest of this talk, is we're count five different revenue experiments that we tried along the way to looking for profitability. And the first one that we thought about was just simply asking our user base for donations.
7cups is not a non profit, it's a for-profit company.
It's a privately owned company.
And yet it clearly offers a service that benefits people and it's free and where people who need support and people who are looking for ways to help others can find a community that values what they're doing, and gives them a sense of participating in something bigger than themselves and working on something meaningful.
So there's a lot of value there outside of a transactional sort of relationship.
And we thought it was, you know, we considered the question of just asking our most loyal users to chip in, sort of on the NPR model.
We used to say the public radio model, NPR being public radio in the US.
And the question of course was would people donate to a for-profit company? And I always had my doubts about that.
But part of being a product leader in a company working with CEO, is you voiced your scepticism, be pushed back if you need to.
And if, you know, if you don't win the argument, you agree to disagree and commit to trying to make it work. And I was certainly wrong as often as I was right, when I argued with my CEO, based that tension between us often helped to find the right path forward, you know.
But so I was concerned about this.
I didn't think we'd ever make enough money from donations, but I wasn't against letting people give us money. So it seemed to worthwhile as an experiment. We also questioned whether we should let people give one time or insist that it be a recurring donation. Initially we locked into recurring, but over time we relaxed that, especially I think when people came along wanting to make one significant donation.
We played around with something, you can see in the upper right of my screen here called a compassion jar in the...
By the way, that's my avatar on 7cups, crank the mellow unicorn.
That's my anonymous username.
But the compassion jar, two people had donated that day, since the compassion jar had reset.
There's a goal every day, and people are encouraged to contribute if they can or find contributors.
So there was an attempt to sort of gamify that a little bit, and ultimately what we saw not surprisingly, it was that we made some money.
It certainly brought in some money, but functionally it was sort of pin money, petty cash. It was definitely not the kinda money that was gonna keep the lights on at a company with a staff of 10 to 15 people or over time, 20, 30 people.
But it was interesting experiment nonetheless. And one thing that you'll see with a lot of these experiments is that we didn't necessarily turn off revenue streams just because they weren't by themselves making us profitable.
Even if they weren't the major contributor to this day, you can go into the compassion campaign page at 7cups, and you can donate, and you can also share your data as a different sense of donation.
You can see it now allows for one-time donations. And frankly, we're not the only company out there that's tried this, just the other day I was checking out a site that rates Twitter accounts on how much they appear to be a bot, called Bot Sentinel.
And I couldn't help noticing that Bot Sentinel was asking people to donate as well.
Without as far as I can tell being a nonprofit. So again, nothing wrong with asking for money. It's just that itself didn't square the circle. That wasn't the full solution for us, for sure. The next experiment we tried was upgrades, a certain kind of upgrade, and you know, something where somebody takes a free service, and you ask them to pay some to unlock some additional services, this often sometimes called a freemium model. Meaning that there's a free version of the product, and then there's a premium version of the product that people have to pay for, and some sort of paywall.
So people who hit the paywall have to pay to continue past that wall, and get access to those, you know, those premium services.
One of the things that 7cups has is a growth path, which is the self-help series of steps that you're encouraged to take, to help yourself every day to maybe do a mindfulness exercise, or work on your breathing or acceptance, or maybe a little bit of other cognitive behavioural therapy that all kinds of different modules available in the growth path, and the growth path bore a passing resemblance to the sequence of mindfulness recordings in the Headspace app, which is another sort of app in the space of mental health and fitness essentially. And Headspace itself has a freemium model, where you get something like nine or 10 meditation or mindfulness exercises for free, recordings, tapes, and then you can pay to unlock the rest of the path. So that's one model where you just say, you get this many steps and then to continue down the path, you have to pay more.
And there's many different ways to do this, but we were under some pressure to make money, and one of the advisors and early investors, angel investors in the company, told us to seriously consider this Headspace model. So as a product head for me, this was not exactly the way I wanted to generate a strategy, but again, I had to roll with the punches and pursue this model.
And I had learned a thing or two about freemium models when I was at CloudOn.
CloudOn, as I said, was a word processor, sort of a word product for the iPad, and later the Android tablets, and then both phones.
And similarly, we had a free version of the product and we had to figure out a paywall freemium model, and exactly where to draw the line.
Now do we charge people to boldface text? Or do we charge them to use a legal line numbering? And what's the cutoff? How do you help keep...
Most people stay productive and use your product and like it, and charge the people who are really willing to pay for some aspect of it.
That's more valuable to them.
At CloudOn, we found that a single payment to unlock all the premium features, primarily things like track changes, was the right paywall that worked for us and produced a good trade off between the free users and the paid users.
But more to the point, I think it's just that I had already had some experience with looking at those models, pricing them and experimenting with them to see what works. We tried this new version of our onboarding, for first-time users that ultimately called the yoga onboarding.
Which is based on this premise that you don't throw people into your entire system on day one.
You give them an easy thing to do at first, like a simple stretch, a simple breathing exercise, and then over time you tell them, "Oh, you've taken a step.
"Oh, now you've taken two steps.
"Oh, you know, these steps you're taking "they're on something called a path.
"Well, you know, if you can stay on this path "you'll get to here.
"And you know, there's other people on the path with you "and you can get support from them." So bit by bit, you reveal more and more of the product, instead of hitting people over the head with everything all at once.
At the same time, we were trying to have people take steps down the growth path, to the point where we could offer them the upsell, where they would upgrade their growth path to the premium version of the growth path.
Which in our case was one that would unlock features that related to your specific concerns.
Depression, ADHD, bullying, coming out to your parents, any kinds of all kinds of different topics, you could choose the specialised path for that subject, if you were willing to upgrade and subscribe on a monthly basis.
So the biggest problem with this approach from my point of view, was that the primary value proposition of 7cups was to offer on-demand chat with a real person when you were in need of support.
And the growth path is a very important part of making it a habit of taking care of yourself in the long run and using the 7cups product to support yourself.
But I felt like it was the secondary thing. And then if we just distract people away from the chat in the interest of trying to sell growth path upgrades, there was a risk of cannibalising the core, mechanism of growing our entire user base and a core product experience of volunteers offering chats to support people who needed help.
So that was a concern, and I think in the long run, it did prove to be problematic.
On the other hand, it did work in the sense that it made money. We made revenue.
It didn't get us to a break-even point by itself, but it took a while to see whether it was gonna do that as we optimise and played around with price points and merchandising and packaging in different ways and trials and discounts and interrupting the user at different places to offer them an upgrade and all kinds of different things. But what we found was that we could sort of eak out a certain amount of money from it, but that number was not gonna converge with again, a sum that was gonna make our payroll by itself. So again, it was a valuable experiment and we learned from it and we continue to have, to this day, you can do growth path upgrades at 7cups. But again, at this point, we still hadn't found the Holy Grail of how to break-even. We also, this is a little bit of just a picture of that first screen showing the easy onboarding of just being asked to do a simple breathing exercise. The sun would pulse in and out and you'd breathe along with it so.
The third experiment we did was a white label version of the product.
White label if you're not familiar with that idea is a sort of an unbranded platform version of your product that customers can present integrated with their own website or their own service, usually with their branding or with co-branding. And we started off working with universities, trying to sell into them, knowing that they had mental health challenges supporting their population.
And also that 7cups being anonymous and happening on your device is a little bit less easily stigmatised.
Let's say, compared to visiting a counselling centre. One of the challenges of this attempt to start to sell the product to as a white label product was that it was a little bit more like enterprise sales rather than direct to consumer, and finding the right price point took awhile, especially because early on, we were sort of using the give us what's in your petty cash drawer the first year, and we'll find a price for you next year, approach. We did get customers this way.
We learned a lot about how to generalise the product and how to work with partnerships, which paid off in some other ways.
And it did represent another meaningful revenue stream for us.
And I'm sure probably it still does for 7cups. I think it started to...
In a different way from how the upgrade path was a distraction from the chat, I feel like the white label product in some ways was a distraction from the direct to consumer model that was working so far for the core product, and potentially creates new stakeholders who pull the roadmap in a different direction. We started to see what some product people are familiar with.
Unwelcome situation where a salesperson tells you that they've already committed to a feature on a deadline that they haven't run by you yet, and that's not on your roadmap.
At which point you have to either do it and suck it up because there's money involved and a commitment has been made, or you have to have a very difficult conversation with sales about how that's not gonna happen, and they have to go back and renegotiate at which point, as I said, that the conversation has to go on further into how we do this and what we offer and how our commitments made.
It's not the first time I've been in that situation, and it's not gonna be the last time either, but it was one of the challenges involved in that white label model.
At the same time, it was certainly valuable learning for the product as a whole.
And as I said, it was generating revenue, and revenue at a higher...
I mean, here you're talking about thousands, tens of thousands of dollars potentially a year because you're selling to a population.
So rather than doing 29 bucks a month per person, these are much larger ticket deals, but they have a longer pipeline.
And you're now in the sales business, which is a different way of going to market and then say finding people in the app store or through search engine optimization.
Experiment number four was to try to build a directory of therapists, professional therapists on our site.
To try to build up a competitor to the most popular psychologist directory in the world today, Psychology Today.
So we were trying to imitate them, and you, you know, you might say we were trying to scrape them, although probably legally, I wouldn't say that's what we were doing, but we were certainly trying to use whatever information they were making publicly available, to understand the best way to build such a directory and the best candidates to be in such a directory. But it was an uphill battle because Psychology Today today as a well-established brand, it's popular magazine, the best known, popular magazine of psychology. And as I understand from psychologists, including somebody in my family, the subscription pays for itself if they get one referral during the year.
And apparently that happens often enough that it's an established method of bringing in new business. So to break into that was not necessarily gonna be an easy thing to do.
And in the end, it didn't prove to be easy to do. And we never made a tonne of money from that directory. And there are other reasons to have the directory, like building up authority, showing that the 7cups was an organisation that had professional therapists involved and not just volunteers, and it also was part of...
It fed into the next experiment we did as well, which was to add online therapy.
And when we start to add on only therapy, we also asked the therapist to sign up for the directory, so it fed into that.
But by itself, again, it was never something that was gonna get us to the break-even point.
Again, it was something that we learned from, and it was happening alongside these other experiments and these other revenue streams.
But at this point, we still hadn't quite reached that point. So let's get to experiment number five, paid therapy. So as I mentioned, the main product was free. Somebody's chatting with you as a volunteer. They've been trained in active listening.
They wanna help you, they're doing their best, and 80% of the time they do help you.
About as much as a therapist would actually, because it turns out most people, well a lot of people what they're getting from therapy is a caring attentive listener, giving them space to work through their feelings and their thoughts, so you could get a lot of that value.
But some people need more help.
Some people need a professional, they need a clinical therapy of some kind, and they can afford it, especially if you can make it affordable.
If you can use online techniques to reduce a lot of the friction, a lot of the overhead of traditional therapy. That's what we tried to do.
This was our second freemium model.
This is we're monetizing a different free service. The growth path was free and the premium growth path was a freemium sort of model imposed on that. And in this case, the free support from volunteers and the paid support from trained therapists represented another freemium gradient that we experimented with.
So we reverted again to a form of onboarding that was oriented towards chat, which made me feel a lot better as again, it was the core value proposition.
And now what we were saying was you wanna chat with a person that's great, we have two ways to do that.
You can talk to someone for free, as soon as someone's available, or you can sign up for therapy, and here's how that works.
It did mean some people were diverted away from the free service and into this upsell funnel. Some of them never be heard from again, whether they bought or not.
But again, that seemed like a worthwhile way to divert some of our core traffic into a paid version of the same service.
We started to really evolve the chatbot that we had initially introduced as a sort of a waiting room concierge, to help people figure out their way through the product and to figure out if they want it to pay for therapy. We introduced, in fact, a second version of the chatbot, the first one was called Noni, and that was the onboarding chatbot, a wise grandmotherly teacup.
And the second one was Sophia, who is to help people with the early part of the therapy signup, including gathering information and consent and things like that.
And what we found was that we got some results right away. By the way, I was sceptical again.
$150 is a lot of money to pay on your iPhone for a service. Even if it's a lot of value for a month, much more than you might get out of one session, let's say with a therapist, many, many conversational turns with a therapist throughout the month is pretty valuable.
And yet I was concerned.
And nonetheless, what we saw there were people out there willing and able to pay this and right away it looked like there was something here that had a lot of potential and fit with the overall value proposition of 7cups.
One of the things that we did was we introduced people to the community and gave them an idea of what it chat was like, by letting them see a sort of an idealised group support chat, which they could skip at any point.
But it was actually, there was like 15 minutes of chat there.
If people really want them to sit and learn. But once they felt archimated they could see that they were ready.
At this point, they would then start a chat with Noni and Noni would take them through this tree of options. Noni was also now learning how to chat with you as well. If you wanted to do a little bit of free chat, Noni was something like maybe a smart puppy dog, as far as it went able to respond to you and learn certain patterns, but would quickly lose the thread, if you tried to do anything sophisticated with her. Nonetheless, it was interesting way to help people by the time.
And we did have some users who preferred chatting with a bot to a person.
So you never know.
But then once people said they wanted to chat with a person, we were able to sort of direct them over to the part of the product where they could sign up for the therapist service, if they wanted that.
We took our experimentation very seriously. We tried to be very rigorous about it.
We wrote down what our hypotheses were and we scored them. We came up with experiments that we measured based on their potential reach.
What part of the product they were in? Who they would impact? What's the range of possible improvement might be? What the resource effort engineering effort would be to actually run the experiment.
And we use that to rank and prioritise them, and to pursue our hypothesis and test them out. And if a hypothesis failed to decide whether to try another test that might figure out an aspect of the hypothesis that was valid. And at which point to say that as far as we were concerned, our hypotheses had been invalidated and set them aside. We in AirTable, we built up essentially a database of these past experiments, because even the ones that failed had learnings in them, that we tried to extract as best we could.
And that we would review whenever we were trying to decide which new experiments to do.
That helped us stay on target.
And also what we used to say, we used to call it stacking up wins.
So you might try two or three different things to convince more people start a free trial with therapy, to give their credit card, to even embark on the initial onboarding sequence. And maybe one or two of these things don't do anything, that you do an AB test, you test them different ways, and you don't see any appreciable difference or maybe it underperformes what you had there before. That still tells you something.
You learn from that experiment.
You say, "Okay, it's not a matter of this." Or "This wording doesn't improve at all." Or "The button colour doesn't make a difference." Or "It doesn't help." Et cetera.
And then there's the times when you do something and it makes it clear difference.
An example would be outside of the revenue stream was we were playing around with the header buttons as you do.
And we had...
Our volunteers were called listeners, because they were trained in active listening. And we had a header call become a listener, which was the one that the main way that a new person or experienced person could sign up to a volunteer. The problem was that a new person has no idea what a listener is.
Cause it was a term of art in our product and not a common term that ordinary people associate with anything specific.
So it might be intriguing, but become a listener doesn't answer somebody's question. Nobody's out there searching for how can I become a listener? We experimented with a number of different phrasings and ultimately what we...
The hypothesis that has that proven true was that using the word, volunteer, the verb, volunteer as a listener or a volunteer with us or things like that.
They got a lot more clicks.
Because there are people out there looking for places to volunteer.
And even if they don't know what it's called there are people at 7cups looking to become a volunteer.
So those kinds of experiment...
Those wins, we used to call them, we used to say we would lock them in, and then try to stack up new wins on top of them. And time this worked.
We experimented with having people sign up with free trial, because that's what a lot of the competitors do. A three-day trial that would kick over into a monthly subscription, if not cancelled.
Which has all kinds of issues in terms of people being concerned, if they missed a deadline or didn't, after the fact are trying to cancel or trying to get a charge back.
We as I mentioned before, created a second bot to help with this onboarding. And as opposed to say a form filling out experience, which made a real difference.
We also instituted something that in retrospect, we didn't call them this, but I would call them growth sprints, which is that we had a core group very much focused on just tightening up that onboarding revenue funnel that signing up for therapy, subscribing, staying as a subscriber, helping the therapist enough value that people would continue to subscribe.
All the things that we were focused on to try to make the therapy business successful got focused on this little task force that met once a week on Thursday afternoon to review the last week.
So this is the sense in which it was a sprint. We'd say, " What were the experiments that we've been doing? "How are they turned out? Are they working? "Are they still running what's next?" And we would in that meeting brainstorm the most important things to do next.
We review our ranked experiments.
We try to come up with new hypothesis.
We would play, you know, an example would be we had a queuing algorithm for which therapist gets the next referral. And it can't just be that you give everybody the most popular therapist, because that has a sort of power law effect, where the rich get richer.
On the other hand, if some therapists are more successful or better, you wanna give them more than an average number of turns in the queue.
So ultimately we had an algorithm that decided whose turn was it next, based on a number of different factors.
And sometimes in these meetings, part of what we would talk about was playing around with that algorithm to try to optimise response time or therapist reviews, or ultimately retention as a therapy subscriber. And then the engineering lead who was part of that core team, Ezra, would stay up all night, hacking the algorithm for the queue.
And, but we go for week, we did this very tight cyclical review process and these sprints, and we found a path forward where we were constantly improving.
And stacking up the wins as I said.
We used them this product analytics tool called amplitude, which I really love.
As a person with a UX background, it's one of the best analytical tools I've seen. And I remember when I was at CloudOn and going from a UX to a product professional, this side of things was rare I really had to go to school. I had to learn about the business side, going to market, making money, charging, you know, looking at costs, figuring out what revenue is gonna come from, looking at lifetime value of subscriptions, looking at cost of acquisition of customers, looking at coefficients of vitality, a lot of math.
Nobody ever said there would be no math, in being a product manager.
There's definitely math.
You have to like it.
And I do like it enough that I could spend all day slicing and dicing charts and looking for the nugget of information and the data that can tell me what I should do next, or at least help me figure it out or test out my theories. So we had both high-level funnels, where we looked at major milestones along the signup flow I'd say, so that we could see in chunks, how each section was performing over time, whether it was improving, whether there were unusual drop-offs, more than you might expect.
You might expect to lose 10% of people at every step. So you can reduce the number of steps.
You expect to lose a lot of people when you ask for money, but you can try to make that experience softer and easier and more trusted, et cetera.
And we studied these funnels and looking at those drop-off points very, very carefully.
This is the point where you asked for money. Suddenly, almost everybody goes away.
And we looked at them over time as well.
So you might say, "How many people are ultimately making it to the end?" And you say, "Okay, in this timeframe, "it's a little bit over 1%." And over the course of doing these experiments over time, average that over the course of the month, it's slowly, constantly, gradually growing. And soon we've gotten to 1.5%.
And then in this most recent month from when I made this chart quite a while ago, it's getting up it starting to approach a 1.75. And ultimately it could possibly, let's say approach 2%.
And so at that point you you've gotten a two X improvement. And did you get it fast enough? Well, I don't know.
But it certainly changes the equations at that point. So in the long run, what we found was that we got the revenue that we needed. We at first were making revenue from upgrades and it was not on track to raise enough money. And then we started to get some white label deals, which were larger amounts of money in a given month, but we're still not game-changers yet.
And then we introduced therapy.
And as we started to figure out how to onboard new users and how to drive people into the therapy funnel, as we kept working on improvements, we stacked up higher and higher revenues month over month. And ultimately right around this point, given what our run rate was and our payroll, we broke even.
So we finally achieved break-even, which is sort of the Holy Grail of startups. And this is something that I wanna take a moment to talk about.
Because when I grew up in New York City, and there's a saying in New York, which is that most restaurants fail in their first three years.
That's the say.
And I feel like startups are the same way.
Most startups fail.
And there's no shame in that.
A startup is inherently experimental business, trying to change...
It's an upstart.
It's trying to break into a space that either doesn't exist or has incumbents and odds are not good.
Generally the challenger doesn't win.
Generally the incumbent wins over time.
You know, the house always wins at the casino. It's that kind of a thing.
Now we know, there's...
You hear about the successful startups, the ones that flout the rules, that changed the game that become unicorns, that become billionaires, et cetera, and they undeniably exist.
But you have to question the ratio, how many startups actually become unicorns and what they call winners fallacy.
Where you often look at the successful companies and say, "Whatever they did is the path to success." But what you don't do is roll back the clock to say the first year of Facebook moved out to Palo Alto and compare all the other companies that lived in a house together, or, you know, only hired engineers or ate a lot of cereal or whatever you thought was their secret and did not become Facebook.
So my point is just that it's nothing to be taken for granted.
And I know I'm talking about startups here and not all product managers work in startups, but a lot of products have revenue and a lot of products to pay their own way or justify their existence inside the company. And I think all of these techniques, they're relentless looking for a model in the first place, the investigation of how can you make money? What might people give us money for? Where's the value in this product? If it's valuable, if people want it, some people should be willing to pay for it. And looking for the place where people will pay for your product is actually kind of a fun thing to do.
It's actually one of the creative parts of being a product manager.
And then getting a hold of, you know, that first little spurt of oil, when you strike oil and realise there's something there. How can we get it optimised? How can we get the most out of this? And when we start to optimise it, is it on track to be enough or not? And if it is, great.
Let's keep working on it.
Let's work that vein in the mind until we've got enough. And if not, let's take our lesson there and move on. So I wanted to celebrate that moment.
I think it was an amazing thing in the life of a startup. You have something called a runway.
And the runway is how much money you have, divided by your burn rate of how much money it costs you to keep the lights on every month.
Which is almost all in a small startup, typically HR costs with a certain number of... Sort of overhead for technical ops, in business costs and hosting costs and things like that. So getting, you know, so you know what the costs to keep the company alive at the current rate every month and you know how much money you have in the bank. And so you have this many months to live.
That's why startups need to raise investment capital often. What happens when you break-even, when you start bringing in enough revenue to pay your current bills, is that the right in a way suddenly extends out to infinity. It's no longer sloping down to zero, it's staying flat. And in theory, all things being equal in a perfect frictionless universe at equilibrium, you now have found a perpetual motion machine. A company that can pay its own way and live forever, or a product that can do that.
Now, in reality, of course, it's life is more complicated.
The competition doesn't rest easy.
There are externalities you don't control.
Typically when your runway seems to be infinite, you start thinking about hiring more people, which immediately means you might not be breaking even anymore.
And it also makes you think about what you're gonna do with that money to continue to grow.
Because to be honest, plateauing forever doesn't seem possible.
Typically you're either growing or shrinking. Plateauing forever it's like walking on a tight rope. It's actually easier to grow than to stay exactly flat. And the easiest thing to do is to shrink, but you generally don't wanna do that.
So we were at this point that in the businesses is sometimes called ramen profitable.
Meaning, you know, we were a scrappy startup, in self-help space, in a sort of mission-driven company with people trying to save the world.
We were paying ourselves below market salaries. So we broke even, but we were cheating a little bit. You know, we were kind of letting our employees subsidised the company.
And we were all willing to do that.
Because we wanted 7cups to succeed.
But it's not necessarily a recipe again for growing the company.
It's a moment.
It's a breather.
You can say "That wolves are not at the door anymore "at the moment or surviving." You know, we can go month to month like this and figure out the next step, which is an incredible blessing.
Something, any company I think would be happy to reach that. To achieve that.
That plateau.
But it's not the answer to what to do next. So we now have to think about strategically, what to do next.
There were really at the time three different directions I think we could have gone from that point. And it's not that I knew.
It's not like I sat there, like the all wise head of product telling everybody we've got these three paths ahead of ourselves.
It was not like that at all.
I was finding my way like everybody else, maybe with a little bit of a flashlight, trying to show people the shape of things to come as best as I could see them.
But I think there were really three things that we could do. And I'm seeing a typo on my screen right now, but then again that's life.
The first thing is that we could have just continued to keep organically growing the company.
So in other words, we had found our way to a break-even through this very scrappy method of experimentation. There was no reason for us to necessarily stop doing those things.
We could have continued to tighten up the onboarding, continued to tighten up therapy, signup, upgrade process, continue to improve the training and the support that we had for the therapist, the dashboards that we gave them to show them whether they were doing well or not.
And to avoid discouraging them.
We could have brought back people who dropped subscription and try to get them to come back again, and over time, you know.
So that's potentially something that we could have done without any kind of dependency on anyone outside at all. Another thing that we could have done at this point was in some ways a more conventional Silicon Valley VC approach.
Which is to now go to some of the investors who were sometimes hovering around us at this point and say, "We have figured out a business model." We developed a community that draws in leads, people who would potentially pay for therapy and just using very big ballpark numbers to give an idea of the range without trying to be exactly precise about dollars and cents.
I would say it this way.
What we found was that for about every thousand people who visited 7cups, one person would sign up to subscribe for therapy. You know, 0.1% conversion rate of the visitor base. And that was the rate that was enough to keep the lights on. Which literally meant that in theory now, of course, some people bounce, but in theory, that meant that 999 people could get a free service because person was willing to pay for their service, which is a pretty amazing ratio.
We knew how long this point roughly after several years, we knew how long a therapy subscription would tend to last or with the tail of that looked like.
And from that we could calculate the lifetime value of the average subscriber.
We now knew roughly what it was costing us to acquire a new customer.
So we had an equation here that Silicon Valley VCs are very familiar with, and we had successfully changed the subject. As I mentioned earlier in the talk, from are we growing like gangbusters to have we found a business model that can make money off our somewhat large base? We did have a million users a month.
We weren't chopped liver.
So this meant that for some users, for some investors, for some business leaders, what they now see is our lead generation business. We figured out a way to drive in customers who are good qualified, you know, the good Glengarry leads, the qualified leads who might wanna buy therapy. Who found us by searching for online therapy or free therapy, maybe, but also maybe just online therapy.
And if we know that we can convert a percentage of them, potentially there's a model there where you can start buying up ads and you know the cost of an ad and you know what it costs to acquire a customer, and you know, how long it will take to earn that back per customer.
And you can almost break even on the ads.
And certainly you can amortise them over time. Which means that you are practically printing money. You can spend money to grow your user base, make the money back by converting enough of them to subscribers and, you know, sort of immediately start getting money back out again, essentially.
And I know I'm grossly oversimplifying this and anybody who's a real CFO or sales genius will probably laugh at the way that I am describing this, but I'm actually getting it basically about right. In fact, one of the things that essentially that we considered as a company at the time when we knew that that model was potential was that sort of turning things around the wrong way. We'd been looking for a way to support our free community. Selling therapy kept the free community thriving, which was our core, the heart of what we did.
To frame it as the free community is simply a good way to generate leads for the moneymaking business.
It felt a little bit like potentially the tail wagging the dog.
In the end, I think these are just different ways in looking at the same mechanical relationships, and it does matter how you describe them and how you view the different people in your product system and how you value and cherish them, and where the money comes from and how that affects your priorities.
So I don't mean to dismiss that we were concerned about that perspective, but at the same time, I think that, again, you don't look a gift horse in the mouth.
If you've found a money machine, a model that more or less pays it for itself gets you and get you revenue and interest investors, that seems like something worth considering. And the third direction that we had in front of us was beginning to be in about the third or fourth year that I was at 7cups, the prospect of some very big deals that involve partnerships and lots of money, and that didn't involve necessarily further optimising and perfecting the therapy business nor turning it into a core kind of revenue stream, but actually took us off in yet another direction. So the talks break-even, we've gotten to that break-even point that I wanted to celebrate, but I think...
And if I'd given this talk, when I first probably pitched it to John more than a year ago, I think that would have been the end of it. I would've said we broke even, we're doing great, and we've now, you know, got these amazing prospects in front of us. And that's a great story.
And in fact, you may want us to stop right here. You may say, I've learned what I need to know about this. I don't necessarily wanna hear the rest.
But also I wanna say that, you know, like I mentioned, that plateau is theoretical.
It still matters what you do with it.
And the last year of the company, we pursued some of...
What I would call those a hundred X opportunities. What happened was there's something called the California Mental Health Services Act. Which as a search an income tax search charge on California residents who make more than a million dollars a year, it's called the billionaires tax.
And that money is put in a fund to pay for mental health services.
More or less for the publicly supported population for the most part.
So in California individual counties administer mental health services to people in their community, through what we call a Medi-Cal or Medicaid. That tax produced billions of dollars.
California is a big state, bigger than most countries in the world.
And a percentage of that, I think 10% or 1% of that fund was earmarked for innovation, for counties trying new innovative things in the mental health space.
And that money was not being spent 'cause innovation is hard and government innovation is doubly hard.
And so we got a whiff of that.
And through a company that we were talking about partnering with at the time called Mindstrong and decided to put in a proposal for an innovation project. And we had the idea of getting $1 million, which sounded like a lot of money for one white label partnership for us in one project.
One of the leads in our company, the Clinical Lead named Laura Gregorio was actually flying down to Southern California for a meeting to discuss this project with one of the counties down there, and somebody at the company we were working, we were discussing this whole project with at the time, Mindstrong, told us that we really weren't asking for enough, that a million dollars was small potatoes, and that we should ask for $20 million.
And we said, "$20 million." You know, we're in the money now.
Now we've really made it.
So we did.
We went after this.
And we landed the deal.
We got ultimately signed at $19 million deal to work with a number of counties to do this innovation project.
And you know, it had some good and bad things about it. For one thing again, we kept learning.
Told us how our product would integrate with government and mental health services, which was one of the long-term goals that we had. And it was built somewhat on our white label product, where we learned a lot from the university. So suddenly now something where we might've been getting, you know, a couple thousand bucks here, 10,000 bucks there, for a different deals with colleges, we're now talking about very large California counties and millions of dollars.
And it seemed like a pretty interesting escalation. One of the big challenges though, was that, again ,it was changing the model of the business. This is enterprise business and government and, you know, programme management and contract service and things like that.
When we had been a direct to consumer business growing our base directly.
And I felt also at the time there wasn't really the direction of the company that I was most excited about.
Again, I was agreeing to do it, and I couldn't deny the value of taking revenue and maybe never having to take investment from anybody, and just growing a company off of revenue.
That's a dream.
I challenge anybody to start up, that's breaking even and is ramen profitable to say no to a partnership that's gonna bring in millions of millions of dollars.
It's a pretty amazing thing to contemplate. And at the time I thought we've got it made now. At the same time, I had some thought, you know, it took a while to close the deal. Working with government contracts is not easy. And it was challenging because what we were doing was innovative and involved volunteers and anonymity and but, you know, challenges around how do you keep it safe and how do you make sure people are comfortable with it? And how do they know whether they can trust it and you know, who's liable? Or how does it look if something bad happens in the press and stuff like that.
So in the end, you know, before around the time that I left, the project for California was in trouble, and scaling back.
And suddenly the company needed to depend again on I think the therapy revenue, that had been a mainstay all along.
Which probably, well, I know for a fact had been neglected. We had not been optimising and perfecting it because we'd been working for a very large deep pockets partner, you know, the 800 pound gorilla in the room. It gets what they want.
So in the long run, it was time for me to go. I'd been there four years.
It was a wonderful ride.
7cups is great.
It's helping people to this day.
We've helped millions of people, including Arya Stark. But it was going off in a different direction, and I had new new things to do on my own.
And if I had to do it again, and if I had the ability to learn from my mistakes, I would probably have chosen option two.
And tried to continue well, one or two.
A version of two (laughs).
Where we continue to optimise the therapy business, and not just get it to two X, but get it to five X, get it to 10 X.
And ultimately turn it into a great business, learn how to drive more customers to it, and then go to the insurance companies or the governments with a thriving therapy business and have them become payers, because they pay for way more for therapy than individuals do.
So get down the weeds here with how you make money in the mental health business, in the USA and around the world.
But those are my retrospective thoughts on having gotten to break-even, having company reach across roads, what to do with that opportunity, and having to go through a very exciting period of growth and expansion, and then in a period of having to contract and reconsider. And nonetheless I wouldn't take any of it back. It was an amazing ride.
And I think, you know, well, a couple of other thoughts.
First of all, I think sometimes, and I'm guilty of this too, having come from UX.
It's fun to focus on the design aspects of product management, the creative parts.
It's fun to focus on even, you know, the talking to customers part that really matters. And it's fun to geek out on the technology. Sometimes we sort of put down the business side, we act like, "Oh yeah, yeah, business.
"That's the dumb part of product.
"We're the cool people who do product now, "not those MBAs from Dilbert or whatever." Which is super unfair because there's no being a product manager without dealing with the business side and dealing with the money.
So that's part of it, is that, you know, don't be afraid of the money, and don't be afraid that thinking that part of your job might involve trying to figure out how to make a product pay for itself.
The other thing that I would emphasise most of all from this talk and from my experience is that notion of experimentation, whether you're in a startup or a huge company, the idea that you're constantly making choices about what to build, what to fix, what to change, what to test and it's those hypothesis and those experiments, which is where you find the breakthrough pathways to success with products.
So I don't think there's any kind of a product job or that experimental with mindset is an asset. And, you know, I think another thing that I would say as a takeaway from our experience at 7cups is to be willing to experiment with wildly different revenue models and to possibly keep several revenue streams going at once. As I mentioned, we still have those upgrades. We still have people upgrading.
We're also when we made a deal with the white label partner or ultimately with a California County or a city in other parts of the country, we were able to say, we will give everybody in your county, everybody at your school a free upgrade, a $29 a month value.
So having a product that was a loss-leader that wasn't making money for us in a significant way, still put a price on some of the valuable content that we could then package in one of the other revenue experiments.
And I think, you know, being rigorous and thoughtful and careful about experiments and being a little bit of a scientist about it pays off.
Being very honest with what's not working and being willing to throw things away or challenge them if they're not getting the results that you were hoping for is super important.
And then as I said, and I spelled it infinite correctly on my last slide, if you ever do break-even, and have that moment of infinite runway thinking long and hard about what you're gonna do with that opportunity.
So I guess my last thing I wanna do before I sign off is plug a book I've been working on called a "Growing Product People" it's coming out from Sense & Respond Press in next year, hopefully in the spring.
It's a book for executives and business leaders about how to build a product team, how to cultivate a product mindset among everybody in your company.
Not just your product managers.
Your product designers, your product developers, people who sell your product, people who support your product, customer success people, everybody, we were all product people.
Right.
Thanks very much.
I appreciate you taking the time to listen to me. And I look forward to a Q and A.
I believe that...
A conversation with John as well on the programme. And if you're interested, get in touch with me can find me online at Designing Product.
We've got a community as well.
Slack community for people interested in the intersection of UX and Product, that I'd be happy to add you to just let me know. Thanks again.
I'm gonna wrap it up here.
So I'm looking forward to the rest of the conference. It's wonderful to see everybody, or at least to be seen by you all.