July 25, 2014
The Anatomy of Heroku’s Product Team
Peter is a co-founder of the Heroku Postgres product team. For over three years he has been responsible for designing, developing, and opera...
Thank you very much for coming. I'm really excited to give you the End of Year Speaker Series for Heavybit this year. As I said, James gave this talk at the end of last year. There are few things that we try and do every year, or we're trying to set a tradition of doing every year, at these End of Year Speaker Series.
The three things that we really looking to do for you this evening is, one, give you a bit of a look back at what's happened in 2015. Two, turn our attention to what we think 2016 is going to look like. And after we've gone through those materials, we really want to to talk to you about what you should be thinking about, the things you should be doing, in 2016 to succeed in that environment.
We'll kick off, and we'll start with 2015. Does anyone remember what 2014 was like, first? You know, as I said, James gave this presentation at the end of 2014, and 2014 had been a bumper year, an absolutely phenomenal year for entrepreneurs, for investors, for everyone in the technology landscape, frankly.
You know, we'd seen more investments in early-stage, in any stage. We'd seen more investments in startups in 2014 than we had since 2000. We saw more dollars being invested in startups in 2014 than we had since 2001. And we saw more venture capital funds being started, more money being committed to new vehicles, whether that's an angel list or a variety of seed platforms, as well as a very, very large opportunity funds and late-stage funds.
We also saw more exits than we'd ever seen since 2000. It seemed, at the end of 2014, that the good times were back and were back to stay. And so we entered 2015 with some really, really high expectations, expectations both around a set of technologies and operating market that we thought were interesting, but also around what was going to happen in the capital environment, and in particular, venture capital.
James stood here and we really talked about two things. We made the bet that there was going to be increased activity, increased velocity in a few high-energy categories that we outlined. And then two, that there will be an increased amount of capital available going from the early stage all the way through to exits.
We can go back and look at how we did on some of those things if we take the high-energy categories first. We talked about seven major categories. Number one, continuous delivery.
Continuous delivery is clearly now a dominant software-delivering model. The continuous approach is so pervasive.
We've seen more and more companies now that basically are designing their products around being baked into some CI/CD process. That is something that we thought would continue. In 2015 we've seen more and more businesses start in that space.
The big data just continues to get bigger, frankly. We've seen some very, very big rounds from Heavybit companies in this space in 2015, as well as a number of companies, or a number of large rounds with technologies like Spark and Kafka.
We've also seen some of the big boys like AWS and Google really beef up their offerings in the big-data space. So more and more activity there, this big opportunity there still.
When it came to containers, we were really predicting war. Docker raised very, very large sums of money, there was money going at CoreOS and it could be Netezza and a variety of providers that fit into the overall container ecosystem. And it wasn't clear how they'd do.
We figured that 2015 would be the year that everyone came heads to heads and started battling it up in a marketplace. And that really has not happened.
What's actually happened is we've had a bit of a detente. We've had a bit of an easing of relations. In June/July of this year, we had the Open Container Initiative, and then we had the Open Container Project. Everyone came into big press circles, said "Kumbaya," and said we should all work together.
That's phenomenal, I think, for developers. Everyone gets to develop to a common spec now, and a common runtime. But frankly, we still think the jury is out.
I mean, if we want to continue this war analogy, this is the interwar period between WWI and WWII. There's some big, big move and some big competition that's coming down the line, but it's taking some time to work through.
Number four here, we said there would have been a lot of new application architectures. And we were talking there, primarily, about new ways of building apps both on the back end and the front end. And we were seeing new mobile frameworks. We were seeing new back-end architectures, microservices. And that's continued apace into 2015.
In the last couple of months we've seen a whole new set of application frameworks. React has definitely taken off, but people are now talking about this BFF model, which is back end to a front end.
The next one is hardware and IoT. And the only thing that we really discovered in hardware and IoT in 2015 is just how hard hardware is. You know, we saw the number of companies either building developer products or building hardware products falter in 2015.
It's clear that people are still underestimating how difficult it is to build a full hardware stack or a hardware and software stack.
There is definitely still opportunity there. It's just going to take some time to see who can work their way through those difficulties.
And then the last two, helping scientists and security. Those are the two that probably haven't panned out as much as we'd hoped. We'd figured that there was a lot of activity around new tools for very smart users, like the science community or new big-data tools for people to manipulate interesting kinds of information. And that really hasn't happened.
Instead what we've seen happen is, rather than tools for smart people, we're seeing a lot of interesting experiments or a lot of interesting activity in smarter tools. And so you can point to things like Google TensorFlow, or large rounds for various AI projects, or even Elon Musk's and YC's new Open AI project that's just happened in the last month or so.
Where that's headed, people are baking in smarter, the better smarts, into their applications. And then with security, again, we've come off a very interesting period. You know, we've had a bunch of Snowden releases, and security is on everyone's lips.
We figured that that would be a galvanizing force to get people to care more about security. It's not really happened. Security is still important, and it's still becoming more pervasive, but it's definitely not had a significant push back on, say, the adoption of the cloud, or people's willingness to give their personal data to third-party providers.
So, you know, not bad from a market prospective or from a categories prospective. We did pretty well, five out of seven. And if we were to pick a few things that are emerging now, I think there's some interesting stuff happening around server-less computing or event-driven computing that Lambda is continuing to spur.
I think there's some interesting stuff around unikernels happening, as well as an increasing level of abstraction when it comes to APIs. It used to be that APIs would just deliver, compute or do storage. Now, people are really putting very complex business processes behind APIs and delivering those processes or those services in an API-first way.
So that's the categories prospective. What about the capital? Well, we said that 2015 would be a great year for founders, a great year for venture.
We had very, very high expectations about the amount of capital that was going to be deployed across the variety of stages. And that's really proven to be true, beyond the pale.
This is undeniably true. In 2014, as I said, we had more deals and more investment than we'd had since 2001. And in 2015 we blew through those numbers again. And so, from a straight investment, and this chart is showing total U.S. investment, total U.S. venture capital investment by quarter.
It's really exploded in the last five to 10 years. It used to be, back in 2007/2008, that you might expect to see five or six billion dollars deployed by quarter, and maybe 20 to 25 billion dollars deployed per year. Now we're looking at the point where people are deploying 20 billion dollars into venture capital investments per year.
So, we've gone 4X, almost, in terms of the amount of capital being deployed. And that's the broad venture capital ecosystem. But that actually still applies to the developer tools market, where we've not escaped this furious bounty of new money.
This is some research that CB Insights and Redpoint put together about the new deals, the amount of investment that's gone into developer-tools-specific companies in the last five years. And we follow that same curve and that same shape. That's something that, anecdotally, we've also seen just in Heavybit companies.
We've seen some great fundraising for a lot of the folks in the audience here, which we're super proud of. But one of the things that you might have noticed that both of these charts have in common, not just how much they've grown, but it's that gray line in the background. That's the number of deals that are getting done every quarter.
As you can see for developer tools, the amount of money that's been invested has gone up dramatically. But the number of deals that are happening is flatlining. It's hovering around the same place.
And this is true not just for us, but for the entire market. And that's led to a dramatic concentration of capital into fewer companies. Looking back five, seven years, the amount of money going into each check or each round has anything from doubled, so they have tripled.
This is data the NVCA, the National Venture Capital Association, put together that looks at the average round size based on the stage of the company. So, at the bottom there you've got very early-stage companies that says on average they'll be raising little less than $5 million. For growth stage companies, that average has gone up from about $5 million to about $10 million over the last five years.
And then for really late stage companies, that's ballooned. It's like 2.5X. So, more money is flowing into these deals.
Investors are placing larger bets into companies. But it's not just that. It's happening on better terms.
This is great news, in many regards, for entrepreneurs, because they're not just taking on board more money, they're taking it on board at a much, much better price. And again, this is true across the spectrum or across the stage of company.
So, for series-seed companies, you can see that medium valuations have gone up in the order of 2X from about 2008/2009. That orange shading in the middle of that is the great financial meltdown. So, I give you some perspective of where we lie.
For series A companies, it's, again, a pretty dramatic improvement. But series B companies is even better. You know, a median investment, or a median valuation for a series B company, back in 2008/2009 was about $20 million. Now it's twice that. It's $40 million.
And that all sounds great, but that, frankly, pales in comparison with what's happening at even later-stage companies. And so now, if you widen the net and look at series B-, C- and beyond, stage companies, and now that $40 million line, or that series B line, is now the bottom line on this chart.
You can see that late-stage valuations have well and truly exploded. It used to be that late-stage companies would be raising money at 50-to-80 million dollars, and they're now raising it closely to $200 million. And this all sounds great. This all sounds phenomenal for the entrepreneur.
You're able to raise bigger checks at better prices with less dilution. But, there's a big caveat.
That caveat is that it's happened without the exit market improving. This chart shows you the number of venture-backed IPOs that are happening per year since 2005, the last one year. 2014, again, bumper year for exits. We had over 120 venture-backed IPOs happening that year, including some really, really big ones. And it's really fallen off a cliff in 2015.
All of these large checks that are being written, really high valuations into late stage companies, they haven't been transferring well into exits. And so that brings me to my next part of this presentation, which is, what does that mean? What lies ahead now, given that's where we are today. What are we going to expect in 2016? And to that, I want to break it down into two parts.
I want to talk first about what the overall operating environment, what the technology landscape, looks like. And then, second, really to come back to this idea of how the venture market is going to change.
Frankly, the operating landscape is phenomenal, absolutely phenomenal. There have never been more developers. There is just ever increasing amounts of software. There's ever increasing complexity. Systems are becoming more distributed.
There's ever more opportunity for people building smart tools for developers.
When you look at the market research, it doesn't matter who you turn to, but everyone says that this is a phenomenal, phenomenal category to be in. Forrester's predicting the plat market is going to grow by over 35%, year over year. Global investment in infrastructural services is supposed to be up 30% this year. Goldman Sachs is saying that cloud infrastructure and platform markets is going to reach $43 billion by 2018.
Worldwide spending on enterprise software is supposed to be $150 billion in 2015, and that's going to be more than $200 billion in 2019. So, just phenomenal, phenomenal operating metrics.
There are a ton of developers, there is a ton of software to be written, and there are a ton of customers out there who are willing to transition. You know, there used to be a bunch push-back around, like, "I don't want to move my stuff to the cloud," and that's just no longer an issue. In all of the surveys that we've seen, everything that we've read, enterprises are adopting cloud infrastructure, with very little concern, at a phenomenal, phenomenal rate.
That's the operating environment, and it's pretty straightforward to see how that's working. Understanding what's happening to the venture capital, what's going to happen in 2016 to the VC market, is a little bit harder. These are private companies. People like to keep that information as private as they can.
One thing that you can do to figure out what's going to happen is you can just ask VCs. And thankfully, someone goes out and does it on a regular basis. This is a survey that gets done every quarter of how confident Silicon Valley investors feel about the next six to 18 months. They've been doing this for, I think, well over 15 years now.
This is the last 10 years, and it's a five-point grade. You get a grade from one, if you're not very confident, to five if you're phenomenally confident about what's going to happen.
There are few things you should take away from this chart. Number one, VCs are just incredibly confident people. It's never been below 2.5 height. This goes up to five.Everyone seems to think that everything's going to be okay in the end.
Number two, the confidence level is in some sense dropping a bit. It's gone from, whatever, 3.8, to a little below 3.5. And it's rather abstract numbers. But more important than that directional numbers is really the timing. The issue with asking VCs what they think is going to happen to the venture market is that they don't know, or at least they don't know ahead of anyone else.
Does anyone remember Sequoia's presentation "RIP Good Times"? Anyone know that? Yeah, I see some nodding heads. All right, well, Sequoia, a few years back, put out a big presentation, that on the cover had a gravestone. It said, "RIP Good Times."
You know, batten down the hatches, fundraising is over. Cut costs, everything is going to hell. Can anyone guess when Sequoia, arguably the best venture firm in the Valley, can anyone guess when that report was produced? Anyone? No. Okay, that was produced at the very lowest point on this survey, well into the 2007/2008 recession.
Pretty much all the other VCs have figured out that they weren't that happy about the thing, and then Sequoia suddenly says, "Okay, well, now actually, now things are a problem." And so, that two- or three-quarters lag is just not going to help anyone.
It doesn't really matter what VCs think about what's going to happen in 2016. It doesn't really matter what I think is going to happen in 2016. The thing that you really have to look at is what are the fundamentals.
What do the fundamentals for the venture industry look like? And you can look at it from two sides. Pretty simple. This is the money in, and then some money coming out. It's a pretty simple equation.
Venture capitalists raise money. They put into stuff, and then they get it back. And they give it to their LPs. And so, if you look at the fundamentals for how much money is going into the market and then how much money is coming in out of the market, you should have a pretty good sense of what your opportunities are, what your chances are to raise money in 2016.
If we start with the exit opportunities, unfortunately, the exit market doesn't look that great. And this probably will take a little bit of explaining. The way public companies or late-stage companies are valued by venture capital investors is usually through multiples.
Late-stage investors, late-stage VCs will look at a basket of public companies that make similar products. So for us, it's going to be SaaS companies, publicly traded SaaS companies. And they will take their market cap, the overall value of those businesses, and they will look at that, and then they will look at the revenue that they're making. And they'll come up with the multiple.
They'll say, "Well, okay, SAP or Salesforce, they worth $100 billion, and I know that next year, they're going to make about a billion dollars. So, the value of that business, or the multiple for that business, is a hundred times.
I'm going to make a billion dollars next year in revenue. The public market says the stock prices are $50 a share, which is a $100-billion marker cap, so we're going to give it a 100X. Investors, venture capital investors, usually, particularly at the late stage, will take those EV to revenue multiples and they will apply them to the venture capital investments they're trying to make.
They'll say, "Look, the public market values companies like yours on a hundred times next year's revenue, and you're telling me that you're going to go public in a year or two, and next year you're going to do $50 million. So, based on that logic, I think when I sell your business, it's going to be worth about a hundred times what you're telling me next year's revenue is going to be.
It's pretty bully, some of this math. There's no hard and fast rules here.
The important thing to know is that the multiple that venture investors have been using, this enterprise-value-to-revenue multiple, has been plummeting.
This chart shows you three things. It shows you the median enterprise-value-to-revenue multiple, the maximum enterprise-value-to-revenue multiple, and the minimum enterprise-value-to-revenue. And you can see that, for the last 10 years, your average SaaS company, about 50-60 SaaS companies that we put together, your average SaaS company is worth about five to 10X next year's revenue. That's what the public market thinks it's worth.
Now, great companies, really high-performing companies, the ones that are growing really well, have great margins, great customers, lots of long-term value, lots of stickiness, they're valued more highly.
We're going to assume that you guys in the audience, you fall into that category. But all of those folks, the upper bound on that value has been as high as 30X in the last three years. But these days, it's been dropping, dropping dramatically. So, all of those big, big rounds that you've seen happen are now backed up because venture investors, who have been betting on the fact that the public market is going to buy their business off them at 20X next year's revenue, are suddenly having to rethink.
Because the public market is saying actually it's only worth 10X next year's revenue. And all of this might seem a bit removed from your early-stage business, or your series A, or even your series B, but the problem is there is a pressure. And this pressure, this valuation pressure, travels back down the pipeline.
The guys who are going to start applying this pressure first are the big VCs. Your big VCs: Andreessen, Accel, Redpoint, folks with hundreds of millions of dollars, folks whose funds are hundreds of millions of dollars, who have billions of dollars under management. They see the entire life cycle. They're making small-seed bets, but they're also making these big, late-stage bets.
They will know, when they sit in their partnership meetings on a Monday morning, they will know that the late-stage guys aren't doing these deals. They will see this back pressure, essentially, from cratering valuation multiples.
They will take it back to their early-stage deals and say, "You know what? I'm not going to pay 10X, 20X next year's revenue because I'm not going to get this finance in 18 months time. Because I've sat in my partnership meetings and that deal gets shot down every time.
And so, what you're seeing is a lot of very, very late-stage companies who've raised good amounts of money. You know, we've got a dozen or so up there including half a dozen A companies that are in our space, raise phenomenal rounds at phenomenal prices, and who are now stuck in a bit of a valuation trap.
That valuation trap, that pressure, is going to start flowing back to the really early-stage companies. Now that's the exit market.
If you want to jump to conclusions, you're going to be like, "Crap. I'm not going to able to raise any money," or "I'm going to have to raise money at a worse price." But that's only half the equation.
The other half of the equation is, what does a supply look like? How much money is there to be deployed? And the answer is, supply looks really good.
We don't, unfortunately, have good data for 2015 yet, but the last few years have been phenomenal years for VCs raising their own money. They have cash sitting in their banks waiting to be deployed. And it's not just large-scale VCs, but it's all kinds of vehicles.
AngelList announced, earlier this year, a $400-million seed-stage fund. It's a phenomenally large seed-stage fund. There are big secular changes or big structural changes happening at the early-stage market. This means there's more money flowing into those things.
And so, on one hand, you have back pressure from the public markets, from late-stage deals saying these businesses aren't worth that much. On the other hand, you have these early-stage guys who are saying, "I've got to deploy this capital." And 2016 is going to bea tale of two cities, where you're going to get caught between two posts where people want to deploy capital, but they want to be smart about how they deploy it, where they want to have confidence that you are going to be able to hit the revenue and growth targets that you are talking about.
They're not as willing to make investments based on the overall promise, but they're willing to understand more about the actual fundamentals of your business, and whether or not this thing could be financed in 12-18 months. So that's something to think of From our perspective, in a nutshell, that's really our expectations for 2016. You are all in a fantastic market, frankly.
It is a great time to be building products for developers, building the next generation of cloud infrastructure, building the next generation of enterprise software.
The way that you're building those companies that go to market is working pretty well. Out of the IPOs that have happened, New Relic and Atlassian, both of which take this mall, have been some of the best performing.
If you're going to build a business in this space, there's no better way than the Heavybit way, frankly. This is the right approach to build these companies. But at the same time, 2016 is going to be a really uncertain capital environment.
You're a founder or CEO, you're worried about fundraising. I think you can really expect two things. One is you can expect smaller rounds. There's no way people can keep deploying money at that scale, in those check sizes. And two, you can expect evaluation compression.
Now, it's probably a little early to shout, "Batten down the hatches!" and start saving money, etc. But it's definitely going to be, I think, a rocky 2016. So, with all of that said, what should you do?
We spent a lot of time on this last bit. I met with James. I tried to meet with some of our founders. We talked to our team. We talked to a lot of our LPs.
Out of all of that research, out of all of that thinking, we've come to one very simple, very concrete conclusion, that's really that I don't know what you should do.
I have no idea, and it's wrong of me to try and suggest that I do. Number one, what you should do is going to be incredibly dependent on your stage. We care very deeply about giving really stage appropriate advice at Heavybit.
The thing I could tell you is probably not what would apply to the person sat next to you. And even if we were at the same stages, every company is different. And so, stuff that works for one company may not work for the other. And the advice that anyone can give you, you always need to take it with a grain of salt.
It's really a set of biases or just a set of anecdotes that they've built up over their own careers. And that's a bit of a cop-out, and it's supposed to be the End of the Year Speaker Series.
Rather than just say I'm not going to give you any advice, I'm going to try and give you three things that you should probably not do, instead of the things you should.
When it comes to things you shouldn't do, these are really some of the biggest mistakes, or some of the common pitfalls that we see happening repeatedly at companies in this space, some Heavybit, some not. These are the things that I think you guys should all focus on and try and stay away from.
I want to start with what I think is probably the most common and the most obvious pitfall, and that is waiting too long to start selling. I could point a couple of companies that are in our own portfolio who I think, themselves, would admit they're guilty of this.
They've built fantastic, phenomenal developer communities. They have a great bottom-up adoption. And they don't spend the time and the energy building their sales organization early enough.
I don't mean they're not going out and hiring enterprise sales people. I mean they're not spending time calling customers. They're not spending time building their inside sales, or they're not spending time trying to convert that $100-a-month customer to the $2,000-a-month customer.
Because God knows it is so much easier to get someone to spend more money with you than it is to find another 19 people who are going to spend the same amount.
The reason that this is such a big problem for folks is simply wanting growth. If you're not investing in your sales organization, if you're not investing in your sales team, you're not growing as fast as you could. And that is going to come back to bite you, particularly in 2016, when there's a tightening capital market.
Secondly, the other impact that that is a knock-on thing is that you have few options. If you're spending money building product, and not on sales, by the time you've run out of money, you don't know what the market wants to pay for.
You don't understand exactly the kinds of things that are valuable to your customers. And you get to the end of your runway without having tested a lot of options. So, you have to really watch out for that.
Some of the early warning signs that you're not doing this right or you're falling into this trap, for me, it's just a simple head count. It's like, how many people on your team are focused on sales, particularly in proportion to the number of people who are focused on engineering.
I'm not quite sure what the ratio is, but suffice it to say that even at a seed stage or series A stage, seed-stage company, you need multiple people focused on that inside-sales function early.
The second early warning sign that I think we've seen across some of our portfolio is that they treat everyone the same. They say they want massive developers using our product. "Yeah, who cares that some of them are paying us more, some are paying us less. We just think of them as one blob."
They're not spending the time picking up the phone, saying, "Hey, tell me a little bit about what you're using our product for." Or they're not spending the time doing customer support or responding to tickets in a prioritized manner. They're not giving potentially high-value customers separate treatment.
If you're not investing that time and energy to figure out who is a high-value customer or a potentially high-paying customer, and who isn't, then you're going to have problems.The knock-on effect of that, frankly, is churn.
All of a sudden, six months in, you've got a company or customer that's been paying you $500 a month, and they disappear for no good reason. And that is directly your problem because you haven't identified that there's more opportunity there, that they actually want some support, or there are other things they are willing to pay for.
If you're getting that decent size churn of larger customers, that's going to be a real knock-on. A simple ways to fix that is just hiring.
Try and find your sales people early. Mark Leslie has a great piece on the sales learning curve, which everyone should read. Doug Leone's talked about hiring the right sales people. Same with Trenton Truitt. We've had some phenomenal speakers here talking about the sales hiring process.
You have to get into that mode of hiring sales people really early into organization. And two, is a culture change. A lot of developer-first companies have this culture of "We don't need sales people." They revel in this idea that, "Salespeople, we don't need those bros," or "Salespeople are outmoded or outdated relics. And that attitude needs to change from the top.
You have to have a very, very different approach as a founder to say, "Actually, you know what? Sales people enable us. Sales people are the folks who will make sure that we can grow fast, that we can add more engineering people, that we can build a better product, because they're getting us in touch with customers and making our customers successful."
So that's the first big mistake I would try and watch for in 2016. Don't wait too long to start selling.
Second big mistake that I want to highlight is the big mistake that we see is that founders think they understand what enterprise customers want. A lot of these mistakes come from over-confidence or some element of hubris in them.
This is definitely an example that, to get to raised seed money, to get to a series A, to get into Heavybit, you have to have built a product developer's love. There's a self-confidence or a self-belief that comes with that. It's like, "I know what the customer wants." And usually it stems from the fact that you've faced that problem, and you've specifically built that solution, because you come from that world.
The issue is that all of that confidence, all of that belief, all of that learning, does not apply when it comes to businesses. It does not apply when it comes to enterprise customers. The stuff they pay for, the stuff they value, is completely different to what developers want.
And it's very easy for founders to say, "We'll just keep doing what we're doing, and developers will wake up to our vision." And unfortunately, it doesn't work like that. You have to be incredibly disciplined about doing your customers' discovery process again.
This is the very particular issue, of crossing the chasm, that every company at Heavybit faces. All of your developer users, they're earlier adopter users by definition. Like your large enterprise customers, some of them might be early-adopter customers, but the majority, the early majority, they want different things.
You're crossing the chasm not just from the early adopters to the early majority, but you're also crossing the chasm from the developer to the business wire. And that's a really hard thing to do without a lot of focus.
How do you tell if you're falling prey to this trouble, you're guilty of this? I think the first thing is that you're finding it really hard to get your ACV up, your Annual Contract Value. If you've built a product that enterprises love, it should be pretty straightforward.
That's maybe overstating it. If you've built a product that enterprises love, you should be able to extract hundreds of thousands of dollars from them. It can maybe not straightforward, but you should be able to find a few customers that are willing to pay that amount, that have budgets that can support that spend.
If you can't, you're in real, real trouble. And you're in real trouble for two reasons. One, you have this little growth problem, but two, you have a very, very hard limit on your scale.
Most of the companies in this space, most of the companies building these kinds of products, they follow a classic parallel distribution; 80% of your customers will be about 20% of your revenue. And 20% of your customers will be about 80% of your revenue. In many cases it's more.
In many cases, 10% of your customer base, that handful of very, very large clients, make up 90% of your revenue. And if you can't extract a lot of money from those very large customers, you can't address that 90% of your revenue, you simply cannot build really high-scale venture business. Watching for ACV, watching for how much you can extract out of customers, is really key.
The second thing is feedback from the sales team, or feedback from the folks doing customer development. If you're going into large enterprises, and they're coming back to you and your sales team saying, "They want to sign. They're excited. All we've got to do is build these 15 things that I've told them we've built." That's a problem. That's a really early, good early indicator that you're not ready to start doing that enterprise sales prices.
You're going to go down a lot of rabbit holes, building custom stuff, random, like early adopter enterprise. But if your sales people are having to over-promise what the product can do to get people across the line, then that's something you need to watch out for.
Fixes: how can you stop this? How can you stop yourself from falling into this trap? Again, hiring. You need to get some enterprise DNA. Most of the folks we know building these businesses, you hire from your network first. You hire folks you're passionate about, building great products that developers love and maybe are not so excited about building a single sign on features.
Getting some of that DNA from large enterprise, with people who've sold large enterprise software deals before, into the business, is key.
The second thing that can really help is better product management. You need to be disciplined about doing the customer discovery process. You need to be disciplined about having enterprise product management, or product management for your eneterprise version.
For most people, for most early-stage companies, product management is innate. You understand exactly what needs to be built and the order it needs to be built in. And so you don't have a formal product-management process. There are no user interviews, feedback cycles, all the rest of it. And getting into that process can have a big, dramatic impact on whether or not you're falling foul to this.
The last mistake that I want to highlight for 2016 is this: don't leave your developers behind. This is easily the most insidious mistake. This is the mistake that creeps up on you, subtly enters your organization, and then two years down the line, fucks you.
The problem with leaving your developers, and when I say developers, I really mean just the end user, not caring enough about the end user. The problem with leaving your developers behind, long term, is that you have a worse business, that it costs you more to find those customers, because you don't have an engaged community of people who are advocating for your product, who are talking about your product, who are distributing your product. And it also long-term leaves you open to a lot of competition.
One of the reasons we love building engaged developer communities is to essentially protect against the next generation of startups, their coming out and building great developer products. If you have developer mindshare, it's much easier to get large enterprises to stick with your product than it is to have to get disruptive from the bottom-up again, when the generation of tooling comes out.
This mistake, of forgetting about your developer, it creeps in subtly. You don't tend to notice the impact for a couple of years, and then all of a sudden, you're in a real, real state. Because this cannot cause significant damage. It can cause, frankly, irreparable damage.
There are great examples of companies. I mean, you could just look at Twitter. Twitter has basically, actively, set fire to its developer community in every which way it can. And now it's turned around and said, "Please just forget the last five years of that, and we'd love to have the community back." And it's trying very hard to do that.
IBM has some really interesting products, Bluemix, Watson and all kinds of stuff. They would love, desperately love, to have the engaged developer community that Heroku and Salesforce, or Google, or AWS have. They would love to get that, but it's really, really hard to back into that.
Once you're an enterprise shop, you've really just lost that developer crowd. You can't fix that branding, or it's extremely hard to fix that branding, after the event.
How can you tell if you're falling into this? It's really about the deteriorating developer experience. If your documentation starts getting out of sync with your tooling, if your developer doesn't know how to sign up on the homepage and has to cut through mountains of corporate jargon to understand what you actually do, these are all problems that your long term developer community is dying off.
The second point is that it dries the sales-and-market funnel. It's almost too late if it gets to that point. If it's now at the point, if you've lost the developer credibility, if you've lost the developer community to the extent that the sales team is saying, "I need more leads," or the marketing team says, "I don't know where to go and acquire developers," it's almost too late. And so you should keep a real eye out for what your sales-and-marketing people are telling you.
The well-qualified leads that you can get are from good telemetry about your developer product, about good usage data about your developer product. That is the fuel for your sales engine. Your sales guys, they want to know who's using most of the product. They want to talk to the customers or the developers who are doing the the most API calls. And providing them with that information is key.
How do you fix that? It's tough. It's really, really hard to back into after the fact. But the two things that we probably advocate for here, one is greater openness. And that means greater openness in terms of software development, in terms of communication, in terms of organization.
The more you shut the org down, the more you develop in private, the less open-source contributions you make, the less your communications are open, the harder it is for developers to get to know you, to know your brand.
And so the more you can do of that in the open, the more you can blog about it, the more you can contribute to open-source software, the more you can open source interesting parts of your platform that you've built, the better off you'll be.
The second thing is really design. As I said, the deteriorating developer experience is a good indicator that you're forgetting about the end user. And having your relentless focus on the design, both in terms of the visual design, but also more importantly, the user design, the overall experience, is key.
And we're actually giving a little plug, but we're having a conference at the beginning of 2016 specifically about building that great developer experiences and the design impact of that.
So, there you go. Three things you probably shouldn't do. Again, it's a bit of a cop-out. But, bare in mind, please don't wait too long to start selling. Please make sure that you really, truly understand the enterprise customer. Be really disciplined about doing that customer discovery process again. And then just don't forget, don't forget where you came from. Don't forget the folks, the individual and user who's going to be building your sales-and-marketing funnel.
With that said, I just want to leave you with some parting predictions for 2016. We think it's a phenomenal operating environment.
The fundamentals of software development, the way it's moving to the cloud, the bottom-up adoption approach that go to market, they are all in you favor.
There has never been more attention or focus on developer productivity, or recognition of how successful developers can create successful companies. And so I think, out of first principles, or from the fundamentals perspective, you're all in a great environment.
But 2016 is going to bring with it a really uncertain capital environment. I think you can expect to see smaller rounds at tighter or tougher valuations. And I think you can expect to see investors caring more about your fundamentals and less about your vision, particularly the folks who are at the series B, and maybe at the series A, but that pressure is going to knock back from the the D rounds, and the C rounds, and the B rounds.
I think you will see it primarily from large VCs first, and then maybe from the seed folks later, but it's something you've definitely got to watch out for. So, as a last suggestion, I can really only recommend that everyone keeps their head down, and you keep going.
It's too early to start conserving cash or to start cutting. Everyone needs to double down in 2016, really invest very, very heavily on growth. Look at your revenue opportunities, talk to your customers more, and focus a lot on your underlying business metrics. With that, thank you.
That's really hard. There's a public market sentiment. At a very basic level, it could be in part that public investors don't believe, no longer believe, the growth targets for these public companies.
They've been telling them, public companies have been telling them, that we'd be growing 20-30% for the next five years. I think that's become more and more clear to the public investors that 20-30% for a public company is just not sustainable for a period of more than a couple of years. And so, consequently their valuation multiples are all getting pushed back down. But it's very tough.
The public market is even more volatile then the venture market. It's hard to figure out what is just general sentiment versus specific concerns. Overall, if you look at the NASDAQ, NASDAQ has blown past 5,000. I mean, if anyone who's older, like me, remembers the time when it was, like, half that.
The overall stock market has been on a phenomenal tear since the end of the global meltdown, a phenomenal, phenomenal tear. It just so happens that a particular enterprise software space, or SaaS technology companies, actually under really a lot of pressure for their own share prices.
So, the fundamental problem here is an over-confidence, or is that a lot of founders have not been forced to go through the customer-development experience. And the reason they haven't had to do that is because they're building a product for themselves.
It's a technical founder building a technical product for a technical solution, or solving a technical problem for a technical solution. Without having actually done that, it's a tough skill-set. It's something that founders in other markets have to do all the time.
You see founders building software applications for teachers, and they're not teachers. You see software engineers building social networks for young mothers, and they're not young mothers. But in each of those cases where you're building something for someone you know nothing about, you have to be really disciplined about doing the customer discovery, going and doing your user interviews, going and talking to your customers.
For our founders, a lot of them, they just get it. And they apply that same belief to the enterprise. "I just get it. I know what they understand. I know what they want. I get it." That is only true about 25% of the time, and the remaining 75% is like, they really don't.
Maybe it's because they haven't spent five years buying enterprise software from other vendors, or they've never worked in a very, very large enterprise like GE, or Coke, or whatever. And so, they don't have that DNA. Without that DNA, without that in-built talent you have to defer to process, you have to defer to rigor. And it's really, really tough to do that.
Again, it's not just this personal defect, but it is a genuine market problem because you're making two transitions at once. You're jumping from early adopters to the early majority. You're crossing the chasm with developers.
There's the developer that has built everything on like the last version of Nogs and wants to rebuild the site and mark down. He hates WordPress and is doing this every three months. But most of the developers out they are like, "I've got a great job building Java EE apps for my tiny company in Bavaria.
And so you're jumping from those early-stage developers, or those early-adopters developers, to early-majority developers or the rest of the developer market. You're making that jump, but at the same time, you're also jumping customers.
You're not just fixing stuff for the individual developer. You're actually now fixing it for an entire different customer that you haven't really served, or you haven't served directly.
You only served them through to the developer. And that's why it's doubly hard, right? Not just crossing the chasm across the same customer type, but completely changing your customer type, completely changing the incentives and the motivations for your buyer or for your user. That's why it's hard.
So how you do it? Like I say, hiring! And in fact, that's one of the interesting things, that when we talked to some of our latest-stage companies, I mean, everyone knows hiring is a challenge. Hiring is fucking tough.
One of the things that latest-stage CEOs have said to us is that the best decision they made when it came to hiring is hiring a hire-er, hiring a recruiter, as soon as they could.
They all say, "We wish we'd done this sooner. We wish we had someone full time who was just doing nothing but recruiting for us." And that speaks to how difficult it is to find good people, how difficult it is to find people with the right skills. And really how much leverage that has on the success of the business. If all of our CEOs are saying, "Just get someone to find good people." That's a dramatic change.
So, I think for anyone who really wants to try and understand, or really wants to understand the enterprise, go and find folks who have sold, done customer development, who've built new products for large enterprises, who've sold into the likes of Coke, GE, and Shell, you know, whomever, and bring them onboard. Don't just assume you have the enterprise DNA in your team.
I still think the phasing that we talked about for Heavybit remains the same. I still think you need to develop a good base of developer interest first. Because I think it's actually really hard for sales people to operate without that. They have no folks to talk to. They have no way of getting feedback on how they should be selling, what they should be selling.
And so we talk about building companies in series of learning curves. Everyone knows the sales learning curve, but there's a product and a community, and a marketing learning curve, as well. The order in which you do those things is important.
For us, it's build a product for developers first, then come up that learning curve. The way you come up is by going and talking to lots of developers when you get to, probably thousands, if not tens of thousands, of developers using your product. I really mean engaged developers, not just registered developers, people who've actually taken it, put it into production, put it in service.
If there are enough people, then you need a sales guy or a sales girl to come and call them. Or one of the founders needs to be doing that. Yeah, 100 is too few, I mean, you can think of it in terms of what the day-to-day looks like of an inside-sales person. There are some good activity metrics about how many customers they should be calling any given day, how many people they should be up-selling, how much feedback they should be giving to the team.
But in general, it's like engaged active users into the thousands that you need before you can really have a sales team. And again, it's not about hiring the enterprise sales guide. It's to have hired the most inquisitive sales person you can find.
Maybe you call them for deployed engineers. Maybe you call them business-development folks. Maybe you call them custom success engineers, but they're people who, at the end of the day, can and want to sell stuff.
It's making sure that someone is on point in your team for when you get into thousands of registered developers, thousands of active developers. Great. Thank you all. Big round of applause for me.