November 2, 2016
The DevOps Journey At Sauce Labs
CircleCI held their monthly Office Hours at our San Francisco clubhouse on Oct. 19th with special guest Joe Alfaro, VP of Engineering at Sau...
Thank you everybody for coming. My name is Geva Perry. I'll give a little more of an introduction about myself in just a second. You can follow me on Twitter at @GevaPerry. You can also check out my blog ThinkingOutCloud.com. I haven't updated it in a while, but there's actually a lot of great content in there about exactly the topic that I'm going to talk about today.
I have been dealing for several years now, with this question: What's the most effective model for generating sales, revenues and growing a startup that's focused on products for developers?
Now, this is obviously a big question and a big challenge and I'm not going to address every possible angle of it today. What I want to do today is sort of give an overview of a lot of the elements that are involved in it and some insights that I've had over the past few years into what I think is getting closer and closer to the optimized model. I want to show you the progression of that and what that model looks like.
As Tom just mentioned, I've worked with quite a few companies. You see a few of the logos of the companies I've been involved with. Some I'm an advisor to, some I'm on the board. Actually, the three top ones: Sauce Labs, BlazeMeter, and Upstream Commerce -I'm on the board. You can see a whole bunch of other companies that I've either been or am an advisor to, or have done some work with helping them with a business model, the financial model, the sales model, the go-to-market plan.
As Tom mentioned, Heroku is one of them. That's how I know James. With the Heroku guys, I worked early on in 2009 when I think it was still pretty much the 3 founders. I've done the same with some of the other companies. I'm sure you recognize some of those names here - Twilio and New Relic and some others. That's sort of what I've been doing for the last five years.
Before that, I actually worked at a software company called GigaSpaces. My last role there was Chief Marketing Officer. I was there for five years, I was part of a team that kind of restarted the company in 2004. I started actually doing sales there. It was a product that is used by developers.
But this was 2004, there was still no notion of online as-a-service and all that great stuff. It was basically your classic enterprise sales model, meaning it's top-down, high-touch with enterprise sales guys.
It was $300K deals, $500K deals, million dollar deals that took 9 to 12 months to close and involved a lot of work on our part in customizing the product, in meetings and in wining and dining customers. If you've ever done that or been involved with that, you know what a nightmare that is.
Actually, while I was still at GigaSpaces, around 2006, 2007, 2008 the whole Cloud computing thing started emerging, and as-a-service, infrastructure-as-a-service, platform-as-a-service, software-as-a-service started emerging. It was becoming evident, more and more, even though there were already companies that were doing pretty well since 2000, like Salesforce.com. It was still rare, especially for products for developers, and for selling to the enterprise. All that was still just emerging.
I got very disillusioned with the enterprise sales model and all that entails, and I got enamored with this potential for a new model.
I started thinking about it a lot, and started writing my blog about that topic, I thought, "what does it look like? What are the challenges? What do we need to do?" That's why some people started paying attention to me. Actually one of them was Scott Raney who was Heroku's investors through Redpoint. He called me up and he said, "I've got a company, they've got a lot of developer traction, but how do we turn this into a business?" He wasn't the only one, there were others, and that's how I started working with all these different companies and really starting to think through this.
I'm telling you my story in a little more detail than I usually do because it's very relevant. The evolution that I went through and my own thinking through this is very relevant to what's going on in the industry now.
There is an inherent, almost contradiction between getting developer adoption and focusing on doing sales. At least that's how it was perceived and frankly there still are a lot of things going on that make it very difficult to be successful at both at the same time. And I'll touch on some of these reasons, but I just want to put that out there and address this issue throughout the presentation.
I'm going to do a little walk through history here just to give us all context of what's going on. How did we get here?
Historically, even products where developers were the users, weren't brought into the enterprise via developers. Developers were always somewhat influencers in the whole purchasing process, but if you think about databases and application servers in the late '90's (kind of the Java stuff ) the developers weren't really the adopters of these products. They could just say, "We need a database" and everything was pretty much dictated to them top-down in the enterprise.
Why was that? They had no ability to influence the process more than just offer their opinion when there already was a sales process with the vendor because they didn't have access to the software itself. They didn't have access to hardware, and I'm not talking about a hardware product that was being sold to the company they worked for. They didn't have access to hardware to try the software even if they could get the software.
Even later, when there were already downloads, the developers had to ask IT for servers to try something out and wait 3 months. It's still very common in the enterprise. That was a problem. They didn't really have a chance to test, touch it, try it out, and certainly not start working with it.
They didn't know what the cost and pricing of it was. Notice I've put a separate bullet for budget because I'm not even talking about them not having the budget to buy it. They didn't even know how much it cost because nobody put pricing on the Web. It was the sales guys who wanted to engage and wanted to negotiate it. There was this huge list price, if there was a price publicly made, and then they would get 80% discount on the last day of the quarter when the sales guys engaged with them and all that good stuff that Oracle still does today.
For the developers, even if they could get their hands on the software, products were very difficult to setup, configure, integrate with other things, so they always had to go to somebody else in the organization. It had to be a concerted top-down decision and top-down effort in the company to get this stuff done. The developers were very minor influencers in the whole process. I'm not even talking about issues of politics and getting budgets and who controls what.
If you look at the little timeline here, just to put it in context, pretty soon a lot of these barriers started being removed. We look at the 1990s. Of course, there was software being sold for developers before the 1990s, but it wasn't really a mainstream product. There weren't that many developers around, not many companies were doing so many software projects as they are today. It was a very niche, unimportant thing, and most companies were creating their own software. Sure there were some database products and some other things coming out, but that was the exception.
In the '90s software products for developers started becoming mainstream. But because of all those reasons that I just talked about, it was a high friction model to get that software into the enterprise. It involved salespeople and meetings, and all that great stuff, and getting servers and doing POC's. A lot of friction. It was the classic enterprise software model whether it was software for developers, or for CRM or whatever, it was all the same.
Then things started to change. One of the first things that happened when the Web came out, something very simple: Companies started putting out their software for download.
One of the first companies that did this in '97-'98 is WebLogic. I don't know how many of you guys are familiar with it. I see a lot of young faces. But, I know those guys really well and they were really some of the pioneers in starting to do this direct connection to the developers via the Web. They put their software for download which was a new crazy thing because, "aren't they going to pay for it?" All these questions came from their VCs which included Kleiner Perkins and others at the peak in those years. But they were developers and they decided "We're going for it this way."
They became a very, very popular product among developers and started succeeding in the model and they were followed by many others. They did it in '98. I'm saying 2000's here is when it became more and more mainstream. Of course, a little later in the early to mid 2000's open-source software really became mainstream. So you had enterprises now using things like JBoss, Red Hat, Linux in all flavors, MySQL, et cetera, et cetera, et cetera. That really was a game changer because not only was it software that was available for download on the Web, it was free basically, at least in theory.
We all know it's not really free down the line especially not for enterprises. But as far as the developers were concerned, it was free. So they didn't have to ask for budgets, they didn't have to ask for permission for anything. They just went ahead and did that. That progression of sort of removing friction from the process continued.
The next phase, I'm going to call that era Web 2.0 because what was really changing about that is the business models. All these ideas of freemiums, and free trials and pay-as-you-go payment models that Amazon Web Services started really making mainstream at that point right around 2006. Also, there were the marketing models, social marketing, marketing automation.
The enablement of all this stuff made it really feasible to market and sell products that were traditionally B2B enterprise products via the Web.
Eventually, the Cloud era which was infrastructure-as-a-service. That by the way, was one of the last pieces that needed to happen. Remember I talked about access to hardware that the developers didn't have it. Now there was access to hardware, cheap, on-demand, scalable, whenever you need it, all that great stuff. Not to mention platform-as-a-service, and software-as-a-service for a lot of the products and all the API. I was just talking to the Keen.io guys here. We mentioned Twilio, and a lot of the other API revolution that happened. Again, a lot of enablement for developers to do stuff they couldn't do before.
Now, if you think about the buying process, and I want to talk about this for a second in this context that I just described, this sort of historical context. I'll walk you through the steps real quick just to make sure we're all on the same page. So status quo, a company doesn't have any reason to buy, this is the current situation. Suddenly something happens where there's a priority shift either some kind of a change in the environment happens, new technology comes out. Whatever it may be. Or a new need arises in the company for multiple reasons, and now they have to shift their priorities, and they're getting closer to understanding they need to bring in a product.
If you think about this priority shift, by the way, it can come from things like either a completely new need — mobile apps. Now everybody needs to build mobile apps. They didn't need to do that 3, 4, 5, 6 years ago. That's a new need. Now they need to buy stuff that enables them to do that, the emergence of Cloud and software-as-a-service.
Suddenly there's a new value proposition out there which is "Yes, CRM like you knew it, but now it's in the Cloud with all of these advantages".
You don't have to install, upgrade, you can add users instantly. You can do all these great things. You don't have to worry about the maintenance. Suddenly it's a new value proposition and there's a new need to buy.
After that they research their options. What do we need exactly? What are our requirements? What's out there? They realize the options and weigh them. They validate whatever kind of final lists, short lists they get to that's interesting and make a selection, and eventually go through the negotiation and purchasing process.
It used to be historically, that the first parts of that buying process were relatively low or no-touch processes. They would happen by marketing materials even before the Web and/or in a conference. But as time progressed and these barriers started decreasing, more and more of this process became low-touch as opposed to high-touch. High-touch meaning sales guys had to get involved, basically.
Companies were in the early days, and still for a lot of companies out there, selling the enterprise sales model. Sales guys get involved even at the point where they're trying to shift the priorities of their customers and are very, very heavily in the account all the time, working the account for whatever angle they can. As time passed, when we removed all these barriers, that shifted.
More and more is done now by marketing online, by the product itself, by a lot of the technology that's out there. At some point we got the ability to even sell things that were, again, B2B enterprise. I'm not talking, of course, about consumer stuff, and certainly developer products that you could complete the entire buying process online. So great. This became a wonderful thing. There are a lot of other advantages.
Early vendors who were doing this became really excited about developers as their customer, because until then, developers were influencers, not the buyer.
One thing that's great about selling to developers is you're getting in very, very early in the lifecycle. When the developers start developing and looking for the tools to work on just building the application or even the prototype, you're at a point where if you sell to them, you're getting into the project really early on.
Sure, maybe there's not cash money in it yet, but you are sort of, "locking them into your product" because they're putting it in the app, it's being used, and it's essentially a lock-in. I found a very fancy way to say it here: De facto decision on purchasing. So, they're getting locked-in to your stuff. From a vendor perspective that's an excellent thing. And the other thing, of course, is that this whole model that I described that's low-touch and bottom-up throughout the whole process is an extremely efficient low-cost sales model.
As this realization came and the barriers were removed, there were a lot of huge advantages to this business model. The pendulum swung from what used to be the attitude in Silicon Valley. If you talk to a VC, some of them even today, but some others 3 - 10 years ago for sure would have told you "there's no money in selling developer tools." And that became a dirty word in the Valley. "Developer tools," you can't make money of off that.
Suddenly all these barriers are removed and a bunch of companies emerged doing really well. I put here 2 stories of 2 companies that came out around the 2011 timeframe. One is New Relic who did this whole PR campaign. You see they're calling it "Death of a Salesman." Nice little play on words there with the Arthur Miller play. And then there's Atlassian that passed the 100 million dollar mark with zero sales people as well. That was wonderful and now everybody was excited about this and a lot of folks out there were trying to implement this zero salespersons, zero touch model. Of course, there were other companies out there that were becoming hotter. Heroku, of course, is one and a bunch of others.
But, after a quick sort of honeymoon with this thing, for many people there was sort of a disillusionment with this approach.
It turns out you do hit a wall when selling to developers and not through developers which is they still had limited budgets. So at the end of the day they still controlled limited budget and they still had limited control about the big decisions in the company.
Sure, you've got developed individuals or little groups, or even departments, buying stuff for a point solution here and a project there. But you couldn't really do these big million dollar, or even 200 thousand dollar enterprise deals to get this really used in the core mission-critical stuff that the enterprise was doing. The developers still did not have that kind of influence. I mean sure there were some flukes here and there, but on the whole it was becoming very, very difficult to do these bigger deals through this model.
I wanted to show you something, a little comparison that I did. The reason I did it is because as I was researching a little bit for this presentation — this is stuff I think about all the time — but, I wanted to do something new. I saw an interview with one of the Atlassian founders who was sort of badmouthing people like Salesforce.com. I realize New Relic and Atlassian are developer products and Salesforce.com isn't. Well Heroku is, of course, but most of their revenue is where it comes from, at least in these years that I'm going to talk about, that's for sure. But, I think the principle still holds for what I'm going to say.
He was saying, "Salesforce was never really a kind of SaaS company" or rather he was saying, "SaaS isn't a business model, it's a delivery model" which is correct technically. But he was basically saying they didn't change their business model, that they still had the same enterprise sales business model. I disagree.
First of all, I was one of the early users of Salesforce so I know how it operated. I also talk to and know a lot of people who work at Salesforce from the early days and today, about their sales model and marketing model and what they do. It is true that they did have sales people from day one. So that is true.
It's not true that they used the old model because even though SaaS is just a delivery model, it almost imposes some new things in the business model to happen.
Salesforce from day one had sales people, but a lot of people were buying online, and buying on their own, swiping credit cards. That was happening from very, very early on and especially when they were taking off.
Now Atlassian, I want to say a word about them. If you notice down there, I wrote in 2011, the year they surpassed the 100 million dollar revenue mark that they had 400 employees and they say they had zero sales people. They also say that everybody in the company is a salesperson (except nobody gets commission). So they passed this, they had 400 employees and did 100 million in 2011. That was year 10 since their founding. That's not that great. That is a lot of years wasted and not great growth.
By the way, actually there's on the Web a graph that shows the revenues throughout the years. They started putting these numbers out because they're getting ready for an IPO so they can't be secret about it anyway. If you look, a year before that they were at 60 million so that particular year they had great growth. But, the years leading up to that, not so great. Even that is not that amazing. And it's a lot of employees, too.
If you look at New Relic, for example — and here's a part I just want to say: I did say in the first slide and I did put their logo that I worked with them, but it was several years ago. I know these numbers here from rumor and innuendo in the Valley. Nobody in the company told me. It's not an official number. You can believe it or not. But that they did last year, which was year 5 since they were founded, 40 million dollars in I'm not sure if it's recognized revenues exactly or bookings. We won't get into those technicalities. And this year they're going to do between 80 and 100. At first I heard 80, then I heard 100, so it's somewhere in there.
They have 270 employees and out of those 270 employees, 70 are in the sales organization. Not everybody is a bag carrying sales guy as they say. It's the whole sales organization which today speaks to what I am going to say because it's the modern sales organization with kind of a triaged model of how they handle all that. But I'll get to that. So if you look at that, they're not at year 10 yet, but you could see the trajectory there. It's quite remarkable. Obviously the biggest difference is that the New Relic guys do have a fairly large sales organization at this point so they're much more efficient on revenue per employee and a lot of other metrics.
It's interesting because in 2010 they were talking about no sales guy and 10,000 customers. Well that's changed and they hit that wall. Some of them might admit it to you. I'm not sure. You can see Salesforce, by the way, which is even higher numbers.
I understand that you can't always compare. Not all of these companies are apples to apples. There are many different reasons.
Salesforce was targeting a very large existing market which was CRM with very poor, not poor in the sense of no money, but sucky incumbent vendors which was Siebel Systems primarily. They got to do it very quickly and Marc Benioff was already a senior executive from Oracle so he came with a lot of credibility and all that. So sure, there are a lot of things there that you can't compare, but you still cannot attribute all of that extra revenue to it. And since they were working with sales guys early on, I think that explains a lot of their growth. They also burnt a lot of money, a lot more money than the other guys. By the way, that year where they did the 748 million they had 1,400 employees.
I think that in the industry, as I said, the pendulum swung. At first it was on the complete top-down, high-touch sales model, then it swung. I was among the people, you could find it in my blog, we're extremists I would call it, who then being disillusioned from the high-touch model, from the top-down model and being in love with the zero-touch or low-touch, bottom-up model that completely said, "that is dead, the enterprise sale is dead, death of a salesman" and all that great stuff. "It's a new world and who needs those sales guys anyway." Then it kind of swung back into the middle and that's sort of where I am, and I think a lot of people in the industry are.
We're still in very uncharted waters as to what that means, what that model looks like, how it works, and I think it's also a little more complicated than people think.
It's not, as some people have been calling it, a hybrid model. I don't like that term because from my experience what I've seen that works best is not a hybrid model. It's a little misleading. It might become eventually some kind of a hybrid, but still not really because there are two phases. From my experience, again, I want to be careful that I'm not doing that apples to oranges thing. For different companies I have a bunch of criteria I put together as to what might change that, and some of it is obvious.
The first two years, I put down as one phase. That is a phase where you're doing the Steve Blank customer development. So the high-touch there is done by the founders and the management, talking to people to understand what they're doing with the product, what they need, use cases, figuring all of that out. You could call that high-touch, but that's not what people mean when they say high-touch sales. That's not the idea. That's more self-education and customer development.
In terms of the actual sales - that's done as low-touch, no-touch, online sales. I'll describe that model in a little more detail in just a second. There's some kind of a transition period which I'll explain, then you're moving into what I call "triage sales" which involve also high-touch, hiring sales guys, building a sales organization and all of that.
I'm counting the two years from what I'm calling GA, and just to make sure we're all talking about the same thing, GA — general ability, that's roughly when you start charging and announcing and all of that.
I was just talking to somebody who was telling me, "Yeah, we started charging, but we really didn't tell anybody. We just kind of quietly put it on the website." But six months after they're charging, they're planning a big announcement. That's when I'm counting the two years, from that point. It sounds like a lot and it is, but I'll explain what it means, what happens.
I'm going to say something else here. There are other ways to approach this. As I said in the beginning my objective and the thing I'm thinking about is what is the best, most effective model. But it has one assumption in it which is - you do fundraise. You do need funding to make this model work. You can tweak it to work with less funding, but then you'll get less growth. For example, Atlassian for eight years didn't raise money. They bootstrapped for eight years. Good for them. They have a shitload of equity, excuse my French, and that's wonderful. But, the growth and revenue was slower.
Now what happens in the early phase? So that first phase, the marketing, that inbound marketing base, it's a COSEM, social marketing, a lot of that stuff. You could do some PR and all that, but basically you're trying to bring people to the website, have them read what you do. Sign up, start using, some kind of a free trial, freemium model, eventually swipe a credit card, self-service themselves throughout the entire process, zero-touch model.
One of the other things you need to do in this phase is emphasize removing friction. Throughout this whole process there is still friction. You see this a lot. There are companies where you go into their product and first of all, the documentation sucks. Then the product still requires some kind of installation and integration. They don't have multiple user accounts. They don't have the right security features. They don't have this and that, so there's still a lot of interaction, problems, support calls, and even just problems getting up and running. This is an important point. You'll understand in a second why.
There's also the on-boarding. You want to do things like marketing, automation, and measuring everything in metrics, and then automating with it. This is exactly that two-year or so, it could be less, a little more, that great period to get that done.
In terms of the transition and late phases at some point you realize, and the next slide will say how you realize that, it's time to transition to the next phase. That's when you start building the sales organization, the sales processes which includes high-touch, inside sales, people on the phone basically, and field sales guys going out to meetings, wining and dining customers into your organization. You do it in a fairly gradual manner. That's that transition period that I'm talking about. It's mostly, you start with some inside sales people. You can even start before that with one person on the phone doing the hot deals. I'll get to them in a second.
Then the other thing that's very, very, critical is that triage process. Now what do I mean by triage? This is something that companies like Salesforce did pretty early on and a lot of companies do today which is you basically find a way to identify what kind of customer or lead you're dealing with here. Let's say there are three tiers. Some companies have more than three tiers, or three options, in this triage. Some companies have two. I don't think anybody has less than two, otherwise, it's not a triage. The three basic models are zero-touch, low-touch, and high-touch. Okay, what does that mean?
Zero-touch, the person goes in, you expect them to sign up on the Web, self-serve, swipe a credit card and get going. Low-touch could be, you have to close the deal in 2 phone calls. If you can't close it in 2 phone calls you have to decide does it go down to zero-touch and then just say in a nice way, "good luck." Or you put a high-touch sales guy on it? Then, of course, the high-touch is where a sales guy calls, it's his account, he's on it, he's trying to make the deal happen.
How do they triage? By what criteria? It's based on, sometimes as simple as, an email from a domain name of a big company. That's enough to put somebody in a high-touch bucket, even though they don't always do it automatically. Also, more and more companies are starting to measure user behavior — what they're doing, their log-in, if they're adding additional users like a multi-user account, if they're doing integration, activation. All the phases I'm sure you all have in your product. Something that's an indicator of seriousness, activity, real use, or whatever it may be. And then they triage them. They do automatic nurturing to them that tries to approach these folks. That's kind of what happens in these transitions.
The late phase is building that organization. That's something that, for example, I mentioned New Relic before, that's what they've been doing in the last two years. In 2011 they had they claim 1 sales guy. Now they have a 70-person sales organization. They've exactly been going through that.
The obvious next question is why not transition sooner? If it's such a great thing, why not do it sooner? From my experience there's several problems with it.
Sales guys, once you start hiring them and they've got quotas and they're working really hard to meet their quotas, they are going to push. These are not wallflowers.
Almost by definition these are strong personalities, type A's, and they want to get paid. They're coin-operated. They're going to work hard to make sure they get theirs. They're going to put a lot of pressure on the organization to do what they need and not necessarily what's right for the company.
The other thing is sales guys are not going to work it out with the customer and do customer development which is that thing Steve Blank talks about. Figuring out what they need, what the use case is, all that great stuff. They need a story. What's my story, what's my pitch going in there? Understanding does this guy fit my pitch, doesn't fit my pitch, it fits my pitch. I give my pitch, I close my deal, I walk out. They know how to work the organization. They know how to work people, help people.
I'm not saying sales guys are bad. Maybe it's a little too late in the presentation to say that, but it's not my intention. I've met really great enterprise sales guys. These are, if you're familiar with, solution sellers. Solution selling is about uncovering the customer's needs and then helping them figure out together and all that. The really good sales guys, that's what they do. They're very smart about it. They need much more of a template into sales and at that phase the company usually still isn't just ready. It isn't ready for that.
There are exceptions to the rule, like I said. Sometimes it's a little clearer. For example, one of the easy things for Salesforce was that they did what's called, and some of you may have companies who are like that. They didn't really invent a new category, a new technology, or a new need, that's what's called, Steve Blank and others before him called, re-segmenting an existing category.
So there was CRM, installed on premises. Everybody knew what it does, what it's for. They already knew they need it. They already had it. He was basically saying, "This is CRM but that works better. You don't have to install it. You don't have to upgrade it. You don't have to maintain it. Anybody can connect from anywhere with a browser and that's wonderful and that's it. So, it's just a better CRM."
That's a fairly easy value proposition to explain to a bunch of sales guys.
A lot of companies have it much more complex. Heroku was a little more difficult. Heroku's first meeting, early day meeting, we were asking, "What is this thing? What do you call it?"
And I still don't think they've figured it out, have they? Platform-as-a-service is what most people would say it is now. I don't think it even says that on the website. That wasn't a given. That was a term I think that was already out there in 2009, kind of obscure and used by industry insiders.
You could see how that's a very difficult thing to go tell sales guys. He says, "What am I selling, just tell me what is this thing?" You couldn't even answer that, and that's true for a lot of companies. Not to mention that even if you do a little deeper things, you know what category it is, but it may be an unfamiliar category to a lot of people. What exactly is the value proposition? Who in their organization is the point of contact? Who needs it the most? Where's the biggest pain? What are the best use cases?
All this stuff is still undiscovered or being explored or in transition and being figured out. It's not something you could put sales guys on.
Then they start taking you in dangerous places because they take you where their biggest deal is. So they hunt for elephants, that's what they call it. They hunt for elephants. They get the biggest commission from that. They perceive it as the same work as a small deal. In the old model it was true by the way. In the old model doing a $20K deal and a $200K deal was pretty much the same effort. So if you're going to do that, you might as well go after the $200K guys. That's what they do.
They take your company into a place which is not necessarily the right place for it to go. They influence marketing and they put out there who to target and how to target. And they certainly influence the product a lot.
The other thing that happens, they come with a big fat deal suddenly. They've got a $400K deal that is very, very tempting to everybody to close that deal because let's say we're a company that's doing, I don't know, a $30K MRR that month. This suddenly takes the company to a whole new level so everybody is saying "Wow!" But it comes with some strings attached. You never do a deal like that without having to do some customization, roadmap promises, putting a couple of people on it. All that is commitment of resources and it's taking the product not necessarily in the right direction for the company. I see that all the time. That is something that is very, very damaging to a lot of startups.
The last thing that's problematic about hiring sales guys too soon is they remove the focus on removing friction from that whole process, and from the product. If you put a person on it, and a very motivated salesperson, they will make up for a lot of deficiencies in the marketing process, in the customer acquisition process, in the on-boarding process, in the upgrading and up-selling process, all of that. The problem is, it's a much less scalable and efficient model, as I said in the beginning.
We're kind of losing all of that great stuff that we had, we had the best of both worlds, with this new model. You're losing because if you don't feel the heat of wanting everyone to be able to sign up, or add a user, or do an integration, and you depend on a sales guy that's going to run around the company and get the answers for them so we could close the deal.
If we have that, that takes this sort of monkey off our back of needing to make it work good, and it reduces our motivation. We're not going to do it and we're going to be less good at it.
This is the flip side of companies that start with high-touch sales guys early on. You can see in their products, including Salesforce, that they're not as good at that stuff as companies who really started with a zero-touch model early on and stuck with it for a while.
When to make that transition. I'm going to jump back just to that little chart. You see this little chart. Let me explain it a little bit. Obviously, the horizontal line is time, as I said relative to what I call GA, and that one is revenues. This is your classic hockey stick and, as I said, that first period there until you hit that first transition line is about two years in many companies that I've seen that sell developer-focused products. Then you start jumping from there. That kind of transition period could take several months to a year. There are other indicators.
I said don't jump on it too soon, you'll get burnt in other ways. Don't miss the boat, or you won't grow as fast.
There are some things that give you some indication. From my experience, and like Tom said in the beginning, and I said, its been dozens of companies, I think at least two dozen who are developer-focused companies in the last 3-4 years . From my experience it happens, I said 2 years, but really it has more to do with the MRR, with that rate of revenue. It's a fairly big range I'm going to give, I'd say in my experience it's more towards the end of the range, but there are a lot of exceptions. It's $50K - 100K in monthly revenues. That's when things start changing for the company. I think it's closer to 100 which is kind of 1.2 million in revenue. So a little over a million dollars in revenue trailing 12 months is when you start feeling that something is happening in this company. Things are changing.
There are more qualitative things that give you a much better indication. One is you as founders, as managers, as guys working in the company feel that you understand your positioning. We know what our positioning is. I see and I talk to a lot of companies. The first time when I walk in I ask, "what's your positioning" and they think they know. As we engage in our conversation and what I mean by positioning, they realize that they know less and less.
Positioning means you know exactly who you're selling to.
When I say who you're selling to, it's not developers. You can't focus on selling to developers. It's what kind of companies and what use cases and companies? What are they doing? It could be what the project is, type of project. What's the real need there? What's driving this demand for this thing? It's who in the organization is the one with the strongest need and the one we need to get to and make all our messages appeal to? It's, like I said, the use case and it could be more technical things.
Like if this is just a product for Ruby developers, great. That makes us more focused, that takes us a long way, but it's not enough. It still doesn't tell us who are we selling to exactly, what are we selling, and how are we different from everything else. Those are the key things, what the value prop is.
Those are the key things about positioning. It takes time to figure that out and a lot of people realize when they figure, "Ah! I know what we are now." And you realize when you don't know what we are now.
The other thing is that we start getting a more consistent pricing, understanding of what the right pricing is for our product and what the usage metrics are if it's a SaaS or API or something like that. We start seeing that there are patterns because I can tell you in those first two years this is a very steep learning curve and it can change rapidly. In the beginning you're influenced. You're so small that you might get a lot of influence from very small events that suddenly do a burst of revenue because of some oddity, or something unusual that happened. It really doesn't teach you much about what is the repetitive, true calling of the company; what is the repetitive sales model and the true calling of the company.
We understand how to triage because we're starting to see a pattern that emerges what a big customer looks like, what medium-touch customers look like, when we can close it on the phone, when can't we avoid going there and doing the meetings or doing other high-touch stuff for them. It doesn't actually have to mean physically flying.
Higher-touch customers or the bigger deals -- they're not going to want to do your click-through agreement on the website. They're not going to swipe a credit card. They're going to want a PO. They're going to want terms, discounts. They're going to want indemnification. They're going to want all sorts of legal terms. They might want this or that and sometimes they also might want some promise as to the product roadmap, or even customization.
There's no need to be extremist about it, just make sure that sales guys aren't over-influencing it incorrectly. We need to understand what those guys look like so when we start triaging we know what we're triaging based on.
I just want to add one more thing. There's another thing that happens, that happens naturally in those two years. Actually 2 points. One is, it's sometimes hard to know when it's right, but there's also something very natural that happens at some point that makes it very clear. This is time. Like all things in life, or almost all things in life, it's not exactly black and white. When I say two years, I'm just saying that's my experience from what I've seen in a lot of companies, is when that starts happening.
One of the biggest indicators is, and this is the beauty of this model of starting with this low-touch model, your customers will tell you when you're ready.
They will self-select and raise their hands and give you a call even though no sales guy is calling them and they'll say "Hey, I've been using your stuff for a while. I've been spending like 800 bucks a month on it, but we've now made a decision that we want to go much bigger with this. We are going to use a lot more and we can't pay you the list price."
First of all maybe your pricing and plans on the website don't even address that kind of size. It may say "Enterprise? Contact us." or not even that for earlier stage companies.
"So we want a deal. We want a discount. We're going to be using this a lot. We're not going to pay whatever that pricing plan is." And they're going to say also, "We can't do the click-through agreement. We can't swipe a credit card. We want to get PO's and send you checks." All that stuff. You will start getting those.
Some of those happen for some companies pretty early on, an occasional odd call like that, which is a wonderful thing to happen. In the beginning usually the founders deal with it or if there is already some other kind of executive, the marketing hire, or whoever it is. It's a volume where you can deal with it and still do your regular job.
At some point though you'll start feeling the heat and often it will happen when you start feeling you're actually losing deals and things are slipping through your fingers because you're not following up properly.
Frankly, also, a lot of founders, especially technical founders which I assume most of you here are, they're not sales guys. They don't know what to do and how to do it and they don't have the personality for it. They don't have the training for it, the experience, et cetera, so deals will fall through the cracks. You will start feeling that happening.
Before that's happening, my advice is don't start building a large sales organization to make it happen because then you're just setting up everybody for failure.
The company is not ready. The product probably is not ready. The sales guys will be extremely frustrated because they won't have a pipeline. They'll come, they'll leave, they'll bitch. They'll do all sorts of things and it will be a big mess. That's why I'm warning against that.
That's it. Like I said in the beginning, this was almost an overview of where we've been and the model as I think we're kind of getting to understand it. I'm happy to talk. I could do basically a university course on each slide here on this deck. There's a lot of stuff going on here.
But actually, I wanted to say, hopefully these guys will invite me again and then I can zoom in on some of the topics in here — just exactly how does that model looks in that phase, what those numbers look like in each of these phases. A lot of people want to know that -- what do the revenue numbers look like, the MRR, the leads we're getting, all that stuff. Maybe we could get really deep in there. But I just wanted to start with sort of an overview, of the big picture.
So we have some time for questions?