A group of entrepreneurs taught a class at Stanford called “Technology-enabled Blitzscaling ” and this is my attempt to summarise the lectures.
This is the first in the series. For anybody who’s starting up a business or is already the founder of a start-up, this is for you.
The following is the video of the class:
The following are the slides shared by the instructor:
What is Blitzscaling?
Blitzscaling is basically the art and science of building a company really quickly to serve the global market with the goal of having the first-mover advantage at scale. Read more about Blitzscaling from an interview of Hoffman.
The mythos of Silicon Valley
- It is a land of startups.
- You don’t have a culture or a fear of failure.
- You can try it again if you fail.
- You can assemble a group of founders easily.
- There’s access to venture capital.
- There are universities such as Stanford.
The classic story of a Silicon Valley startup
- Your story starts in a garage or an apartment or a dorm room.
- The core team consists of technical founders.
- You create a new product (the app).
- You raise funds for your product.
- You achieve product-market fit and it’s off to the races.
This is what makes a Silicon Valley startup strong and this classic story is partially true. You need to check all the above points to weave the beginning of your startup story but that’s not all there is to it.
You think you have it in you to become an entrepreneur? Here’s a quick checklist to find out if you’re really cut out for it:
- How do you select a founder?
- How do you know if the idea is right or not?
- How do you get your initial financing?
- How do you figure out what the competition is doing? Is information about your competitors relevant?
- Are Google and Microsoft doing that?
But, the most important of all is a comment that Sam Altman, co-founder and CEO of OpenAI and the Chairman of Y Combinator, a leading silicon valley startup accelerator, had made earlier. He believes that you should start a company only if you’re dying to and therefore encouraging people to do so is a bit scary.
What makes Silicon Valley… Silicon Valley?
Various startup communities have evolved in different parts of the world as it is easy to assemble a small group of people, have an initial idea and build on it to create a product or a service. There’s enough knowhow, there’s access to funding and people know that part of taking a bold step of starting up creates jobs. And yet, Silicon Valley has an edge over other startup communities and creates a large number of interesting companies.
The classes would cover what makes Silicon Valley so special and throw light on the underpinnings, the risks, the opportunities and the decisions that go into the ability of companies to scale up to provide “massively differential results” in a consistent fashion.
Take a look at the top 10 companies that are valued at over $10 billion (the instructors were using the 2015 data as the reference). Except for Tencent and Alibaba Group, all of the other companies are based in Silicon Valley.
There were four unicorns in the 1998-99 period (Google, Tencent, Alibaba and Salesforce), which quickly increased in number, an indication of increased value creation. A note of caution: This is not the only way to judge how large or big a company is. Also, do keep in mind that these are paper valuations.
Fun Fact: Aileen Lee popularized the term ‘unicorn’, which is used to describe companies that have valuation of over $1 billion.
The above table only reaffirms the trend highlighted above. However, the kicker is that the population of the Bay Area (around 7 million in 2015) had created half of the unicorns trading then.
The Importance of Networks
The theme of networks essentially means the links between companies, people, jobs, skills and universities and how they all relate to each other. One of the reasons for Silicon Valley’s special occupation on the startup map is because of the amazing networks it hosts and sustains. It is definitely possible to build such networks elsewhere but “there’s some kind of organic luck that comes into it”. You need a set of cultural, communication and ideational things to create networks.
The talent, capital, knowhow and the kind of people make up the Valley’s networks. Another key factor is that the formation of networks began much earlier in the Valley than in other places. So, this has given the region a headstart over the others. It dates back to Bill Hewlett and David Packard who started the Hewlett-Packard Company who were Stanford students and are considered to be the first to take the Dean of Engineering of Stanford University, Frederick Terman’s advice, to stay in the locality and develop it as a high-tech region.
With the birth of the HP Company in the Valley, others such as Yahoo! and Google came in and with the birth and growth of each company, more layers were added to create a region with a huge diversity of roles and talent, from funders to strategists to product people, who understand the importance of networks and invest in them over and over again. Sure, you can replicate it elsewhere but it takes years to do so.
This is how the networks pan out: the graphs of companies look flat for years until something clicks where “network starts to get internal consistency and liquidity” and takes off. This is the essence of what ‘Blitzscaling’ is, a term coined by Hoffman and Yeh.
Key factors founding teams need to consider as they scale up
These are some of the key questions you need to answer not only in the beginning stages of your company but throughout the scale-up phase. The way you think about them and the way you work on them changes at different stages of scaling up.
For instance, Aneel Bhusri, co-founder and CEO of Workday and David Duffield, co-founder and Chairman at Workday, interviewed the first 500 employees for their organization – not on the basis of skills but on the basis of culture fit, customer focus and their ability to be good team players. But after that, it changes because they cannot continue to interview every single hire as their organization grows bigger.
When startup founders pitch to venture capital (VC) firms, one of the things they usually say is that they are worried about Google or Microsoft being their competitors. If you are competing at some angle with one of Google’s 200+ products, then you needn’t really be worried about Google being your competitor. Your real competition is other startups.
When you achieve organizational scale, you’re deploying your product/service, you have a sizable number of customers with a steady revenue stream, then larger organizations become your competitors. Not only will you be competing with other startups but also with industry leaders “in certain ways because they may have a leverage and they may now be focused more on your end”.
Different levels of organizational scale
The instructors have put together a “super rough structure” to define companies at different stages of scale-up – family, tribe, village, city and nation. These stages have been made on the basis of number of employees, number of customers or users and revenues but this is just to give you an idea as to where you might be in the stage of evolution. As such, you may not have the exact same metrics under the sub-categories and your company might be a combination of different metrics from multiple stages.
The point of this is to show that things are very different when there are just three co-founders versus 15 employees versus 100 employees. Again, things are profoundly different when your company is earning nothing versus earning $1 million. As a company evolves through various stages, certain themes are bound to be consistent but the activities or the way work is done changes.
Usually, the number of employees correlates with the number of users and revenue numbers though there might be exceptions such as Whatsapp and Instagram. In 2015, Whatsapp had around 950 million users across the world and this was managed by a small team of 50 engineers. As for Instagram, it had just over 450 employees managing 600 million monthly users and 300 million daily users in 2017.
How to think of Organizational Scale and Functions
Note: OS1 stands for first operational strategy, OS2 for second operational strategy and so on.
When you talk about organizational scale, you need to think about different parts or functions of your company. You need to think about questions such as:
- Are you a single-threaded or a multi-threaded product company? As a startup, you’re always single-threaded but when do you get to being a multi-threaded company and how do you make that decision?
- How is your go-to-market? Is your go-to-market one simple thing? Is your go-to-market a plan that’s enterprise-based or consumer-based and remember, there’s differentiation within those plans. Some enterprise-based plans are classic that involve heavy field sales, some are essentially telephony while others such as Slack are based on an entirely new model.
- What’s your technology strategy? One of the quotes that Hoffman’s friends say he will never live down is “if you’re not embarrassed by your product release, you’ve released (it) too late”. The whole point of that quote is to emphasize the importance of speed and time with respect to getting to your market. Obviously that doesn’t apply to launching a hardware product because if you’re embarrassed when you launch it, it probably means you’re already dead. So, this largely applies to consumer internet software businesses where speed is extremely critical. If you are going to say, you’re going to build a “super robust technology” and that you’re going to have it as part of your first operational strategy (OS1), then you’re likely going to be developing it slowly and you’re not reaching the market fast enough. So, if you’re a consumer internet company, how you launch the product and then rebuild it as you go becomes part of your technology strategy. However, the pattern that works in the early stages of your company will not work when your company is bigger and your business is likely to die. At the later stages, you begin thinking about how is it that you’re building a platform, how the platform makes a development stack that makes everyone productive and how do we get tools to do it the right way and that’s how you begin to look at how the range looks like in the scale-up phase.
Most likely, you’ll figure out your inflection points at OS3 and find out whether or not you have a real blitzscaling opportunity. At this stage, you may end up spending a lot of time figuring things out or you may move really quickly – it depends.
OS1: 12-15 employees
OS2: 150 employees (Do read up on Dunbar’s number)
OS3: 100s (200-600 employees)
These numbers are approximates and they depend on a number of factors including your exact mode of operation, the kind of global distribution you have, if you’re scaling your customer service or your central product and development organization.
Key Considerations of Blitzscaling
When to Blitzscale: It’s not as if you can say that you have an idea and you go find a couple of friends and say you’re going to blitzscale immediately. This probably works very rarely. The speed at which you’re operating is partially an exercise in judgement and intelligence about things such as what the competition looks like and what’s the way you’re going to win the 10-year-game. These games are rarely one-year-games.What’s the way to get ahead of the competition? (This is likely to be super-intensive in terms of the way you do it.)
Generalists to Specialists: As you’re going through the various stages of scaling up, you’ll find that you’re moving from generalist to specialist. When you first start a company with a couple of people, you’re doing everything from coding to getting office supplies to buying pizza. The kind of people you’d be looking for in the early stages (when you’re trying to figure out things) would be generalists who are more flexible, who are willing to do things they aren’t comfortable with and are quick to learn.
Doers to Doer-Manager to Managers to Execs: But, as you scale, you hire more and more specialists – be it technology, sales or management. This would also mean you will be moving from everybody doing any work that needs to be done in a room to some people who are working on something, some who are managing and some who are doing both.
As a founder, when you get to the executive level, your primary function becomes the organization. How do you compose it and how do you operate as a team? How do you scale people up? How do you do onboarding? It’s not like you can articulate a vision, stand at the helm and point in a direction because that’s what a useless executive would do.
Scale and Innovation: Another trend is that it’s not as if you would innovate first and then everything that follows is thoughtless scaling up. You would be working to preserve your ability to be innovative as you scale up. “You need to innovate on how you’re managing data, you need to innovate on what is your go-to-market and how are you transforming the way that you’re acquiring customers,” says Hoffman.
Choice on adaptation vs. Operational Excellence: Often, you would have to make a choice between preserving adaptability and operational excellence. How to drive down unit costs, how to make it more efficient to provide a service, how to increase your employees’ productivity are important questions. However, as you Blitzscale and speed is a factor, sometimes you end up making choices that side with adaptability and speed, that might result in wastage. (You may say – It’s okay if I hire a lot of people.)
They decided to solve the problem immediately and set up a 200-strong customer service center in Omaha. They ended up churning 70% of those employees in three months because they were going for speed rather than operational excellence.
Global reach: As part of the networked age, you’d be global faster than you can imagine. When LinkedIn was launched, it was present in 15 countries. When people started complaining that their country is not in the list and the drop-down menu got long pretty fast.
Capital requirements: It’s a no-brainer but you cannot Blitzscale fast if you do not have enough capital. There are two ways to do it: have a good revenue model so that you can reinvest that money in your business or tap the capital markets (such as venture capital, debt and IPOs).
Speed is relative – but competition is global: You can even Blitzscale in down markets because speed is a relative metric. It’s more of a question of how do you move faster than your competitors in a networked age.
In 2015, ‘burn rates’ was one of the hot topics under debate and one that’s still relevant today as well. The environment was overheated and companies were raising big rounds of funds. Investors were concerned about the burn rates and the tension will be between how much you spend and if you can go win the market.
“And so, you get into this funny cycle of trying to win the market, but also not trying to increase burn rate, which is how much money you’re sending out the door every month. And so, it’s a real tension and it’s a very, very current question,” says Lilly. You need to look at how much money you make or lose on each transaction versus how much money you use up to go to the market. You may want to read about Uber, a classic example of the delicate balance between burn rates and viability.
This is a rough way of thinking of things when you move from one stage to another:
LinkedIn’s Blitzscaling experience
LinkedIn was one of the companies that did not Blitzscale in the beginning; it spent two years in phase one, after which you see its graph moving up consistently.
Stage 1: In 2002 and 2003, the core team set out with this idea that if it could build a professional network with reputable relationships, it would be useful for the different things that professionals do on a regular basis. They made it a searchable network. The idea was that if they get to a certain number of users, then people would start using it regularly and that it would be a new way of doing business.
All the people the company hired initially were friends or former colleagues. The organization was as lightweight as possible and was operating with the absolute basics. All their efforts every single moment of every single day were directed at solving one problem: What is going to be valuable for the user? That’s the only thing they thought about and everything else was done at the minimum level possible.
LinkedIn released in May 2003 and immediately, they began learning about what product-market fit actually meant, which basically means you’re providing a product that provides value to the market at scale. Earlier, they had a theory/hypothesis about what their product-market fit was but it turned out to be different from what they had envisaged.
When they launched, it was immediately obvious that recruiters would love it but they had to reach a critical user base for the recruiters to search for potential hires. Their earlier notion of product-market fit did not factor in user base, which would essentially drive their entire idea.
Stage 2: Once they realized what their exact product-market fit was, they put together the team shown in the picture below to achieve that. The company found itself in a place where it had to go from 12 employees to 30-40. They added a set of functions such as customer service, sales, go-to-market components and general administration that they didn’t need to worry about in the first stage.
Stage 3: In the third stage, they had to go to the parking lot to take a picture as the team had grown considerably bigger. LinkedIn had 13 million users at that point. At that point, they tried to do two things simultaneously: take advantage of the existing fit (to build a get recruiter business) and to explore additional fits.
They had a great fit with the recruiters but they represented only a small percentage of all professionals. The team knew they had a value proposition for every professional. How did they go about building that? In 2007, they went from one track to five tracks. Each track supported either the growth of their existing business or explored new opportunities. This was LinkedIn’s way of balancing operational excellence and adaptability.
Specifically, one track supported revenue, another focussed on growth while the other three looked at other options that would continue to allow them to grow. At this point, LinkedIn had 120 employees and needed a different kind of organizational leadership. They appointed Dan Nye as the company’s CEO in early 2007.
It’s extremely important to know which people to put in what jobs, especially while replacing the founders with new CEOs.
Stage 4: In 2009, LinkedIn began Blitzscaling. “We hadn’t really done it up to this point because remember, growth was our main limiter. If we didn’t have a big network, we weren’t gonna be able to drive stuff. So basically, we had to bring all these people in for us to be able to basically take advantage of the growth we had achieved up to that point and continue to drive our efforts in those five product lines,” says Blue.
At that point, the expectations of their users were changing in terms of the way they were using LinkedIn. The company had offerings including a recruiter product, a sales product and marketing solutions product and the bar had to be raised on all of those. In addition, they need people to manage the new relationships, the sales team had to generate new things and a lot of their customer growth came from overseas. To go through these layers of transitions, they brought in a new CEO again. Jeff Weiner, who addressed the class a little later, was appointed as LinkedIn’s CEO in 2009.
Weiner led the company from Stage 4 to Stage 5. The first thing he did after taking over his new role 11 years ago was to prepare LinkedIn to Blitzscale.
“So one thing we had never done and this is a great learning for us, is we’d never written down our company culture. We hadn’t written down our strategy because we were too small. We didn’t have to do that. You’ll find, if you haven’t worked in a startup, that when you’re a startup, everyone knows everything all the time. But when you’re a big company you have different management and executive leadership needs. So Jeff came in, he wrote all that stuff down and it’s still the way we run the company today,” says Blue.
LinkedIn doubled in size year-on-year from 2009 to 2014. Another thing that happened during this time was that since the company had worked on the same codebase for six years, they had to make “tremendous technical changes” to make sure they had the necessary technical platform. The company had to change their technology strategy at this point and they had to consider a bunch of things such as building for scale, for flexibility and developer productivity.
They also had to change their financial strategy during this time. LinkedIn had become profitable in 2006. However, at this stage, they had to think about capital needs to do things such as acquisitions. So, not only did they go for a seed round in this stage but also went for an IPO (initial public offering) in 2011.
“This is a very, very high level description, but there’s no clean path which gets you from one thing to the next, but knowing these changes and how those needs change over time is the main thing that we wanna get out of this (class),” says Blue.
In this stage, which would be covered in detail in subsequent classes, the important things to think about are:
- Is your product any good?
- Does anybody care that your product is good?
- Whom do you hire
- How do you get employees to care about what you’re building and work for you instead of going to work at other great startups or companies?
- How do you ensure that you can pay those employees?
One of the things that is important at this stage is not only which problems you solve but which ones you don’t. “Part of the entrepreneurial journey, and that happens even when you’re at scales of thousands, is there are fires burning when you’re going home. That’s fine. You have to know (for) which fires it’s okay to go home, it’s like, yeah, we can deal with that one next week, right, that’s fine, and which ones you can’t,” says Hoffman. “Letting the wrong fires burn obviously is another fatality.”
The following are not really relevant or very important at this stage:
- Board of Directors
- Analytics & Dashboards
- Everything else
In the first stage, there are certain things that are extremely critical and you need to be obsessed about them everyday. But you also cannot pre-solve problems. It’s extremely rare, in the family stage, that data would be key to your success and it’s very likely that this is the case as you get to the ‘village’ and maybe, even the ‘tribe’ stages. At this stage, you probably have a basic dashboard to see how many people signed up or downloaded the app and similar metrics. As your company evolves, data is probably going to move up a few places in terms of priority.
Though strategy is important, if you’re only thinking and talking about strategy, then it’s not going to work. You need to “actually have a disposition for getting into the fight, for doing things”. There are a very few product geniuses who just think of an idea, work on it, launch it and it works. If you presume you’re one of them, it’s a bit of a high strategy; maybe, it will work, probably won’t. So, you might want to make people care about your product than worry about data and systems at this stage.
The Mozilla Story
In Mozilla’s case, they figured out how to build a quick, fast browser right at the right time when Internet Explorer was starting to be a not-so-great product. Mozilla continued to build asset after asset and when “context happened”, it could Blitzscale. Once they hit the Blitzscaling phase, they grew to have about 400 million users over the next 3-4 years. If you see the chart above (most of them would look like this), it is slow, slow, slow and then fast.
The first few years are about building the organization, the uncertainties, going through the fog of what’s going on in the market and all the rest of it. And, it also becomes a part of your decision as to when to hit the accelerator. If Mozilla had hit the accelerator at any other time, it just wouldn’t have worked and they would have just ended up spending money.
Hitting the accelerator is uncomfortable and when you start to Blitzscale, it’s uncomfortable all the time.
Are Blitzscaling strategies equally applicable when you get out of the software realm?
This can be answered in three parts:
- It’s all about speed differential – what’s going on in the markets, what the competition looks like and how you play.
- There are many businesses and industries where software is the key part of the differential. Think, software and medicine or software and genetics. What makes Tesla interesting is that it’s a software car and not so much, an electric car.
- In some cases, there might be things that make you different from the others. For instance, it could be how do we move faster than the competition with a low error rate as opposed to accepting a high error rate as a way of doing things as seen in the software industry.
Some of the critical questions to consider in this case are:
- Are you selling your product?
- Are you hiring well enough?
- Will you own the market opportunity?
- Are you moving fast enough to do that?
- Are you building a foundation that’s durable?
If someone else owns the market opportunity before you, then you’re not going to own it. So, when you get to a point where the competition is not close to you, what you would actually begin to think about is how are you building something that will compound over time and not how do you get there fast. One metric to go by is if you’re moving really slowly, the fact that you’re going to compound and become big later becomes a little bit in doubt and not necessarily fully in doubt.
There’s a lot of judgment that goes into each specific part and that’s the reason why providing a roadmap for how to go about it isn’t possible. Part of what you’re doing as a founder/executive is you’re exercising judgment. One of the things you should do is talk to smart people, not just to the people who have done it before (build a company) but also others because that would help in making your decision crisper and better.
At LinkedIn, they did not focus on the competition in the early stages because they were concentrating on putting together something that would completely win the market down the line. The idea was that if they were able to build the largest community of professionals, the competition didn’t matter. Most of LinkedIn’s early competitors were companies that were focussed on selling the corporations and they don’t exist anymore.