How Startups Work

Jul 2014

(Based on a talk for MAP)

 

“The rules of business are like the laws of physics, neither inherently good nor evil, to be applied as you may. You decide whether your business is constructive or destructive.”

– Randy Komisar, KPCB

 

“If you imagine someone with 100 percent determination and 100 percent intelligence, you can discard a lot of intelligence before they stop succeeding. But if you start discarding determination, you very quickly get an ineffectual and perpetual grad student.”

– Paul Graham, YCombinator

 

The notion of what we know today as a startup only really became possible in the early 1900s. Before then the idea of starting a company was foreign. It still is foreign, but is now foreign in the same way meeting someone who plays the cello is foreign, not that meeting an alien is foreign. One is just rare, the other is seemingly impossible. The reason it was foreign is because it was expensive.

The first corporation dates back to the famous East India Trading Company which was the first company to raise capital by selling shares in itself. What they did was use secret trade routes and ship making technology to build ships that could traverse huge distances quickly with cargo, and so unlocked new industries in the form of trade with other regions. This would have been a hugely expensive and thankless task.

And so companies in those days were too expensive to start for the average person. Not anymore. What happened in between is the cost of creating new technology decreased dramatically. It’s always been a rare, subversive group that created new technology. And until recently that group hadn’t expanded much.

But as the cost of creating new technology drops, that group increases in number inversely proportional to the cost decrease. When the cost of creating new technology – and as a result new ways of doing things – decreases. The cost of starting a company does too. When this happens the number of new companies being formed increases because they are the first to leverage a new way of doing something.

At first, most of these technologies were commissioned by or created for the government. That is because they had the highest concentration of deployable wealth and risk tolerance. The government, usually in wartime would spend huge amounts of money to create new technology which would give soldiers a competitive advantage in battle. The commissions would go to small companies which grew in size and began to create their own products.

These new products would often create self-fulfilling markets because they everyone would want to use them and money would follow. A good way to visualise it is to chart the transition of advertising through different mediums.

Advertising was primarily paper based until the advent of the radio at which point audio advertising took over. When the television was invented, TV ads became popular which were both visual and audio based. When the web came along, internet advertising became popular. Online ads had audio, visuals and were interactive. The internet did the same thing to television which television had done to radio which radio did to paper.

I like the ad example because ads have remained largely intact. They are fundamentally the same thing today which they were a hundred years ago. They’ve just evolved as technology evolved. And so their evolution is a good shorthand for the progression of communication. Each new invention subsequently unlocked entire industries and new ways of doing things.

The way most of these inventions became popular was due to startups. But they weren’t called startups back then. It wasn’t until the last decade that people realised startups are not smaller versions of big companies but a different type of company altogether. Until this point, people applied business theory designed for big companies to startups which is like treating a child as an adult. In fact even the word startup is a new term and there isn’t one comprehensive definition for it.

 

The word first emerged in 1976 in a Forbes article. The most famous definition of startup is probably by Steve Blank in that “a startup is a temporary organization designed to search for a repeatable and scalable business model.”

Finally we had a word for an entity that underwent this process by which new things were created and made their way to the market. You could say a startup is a new entity that commercialises new technology. That’s a fairly broad statement. What is new technology? It can be anything that adds to or improves the human condition. Commercialising means bringing it into the hands of people. There’s little point in creating new technology if it doesn’t get commercialised because then nobody gets to use it.

So putting them together you have a new entity that creates new things and brings it into the hands of lots of people. In exchange it charges money. And suddenly this startup resembles something everyone is familiar with. A business. Because at it’s core that is what it is. A startup is a business and so follows the rules of business.

And so the short hand for a type of company that commercializes new technology and can quickly grow from a small company to a big one is startup. That is what people are typically referring to when they use that word, whether or not they realise it.

The engine by which a startup grows is called its business model. A business model is a fancy way of saying how a company makes money. Money is just a tool. It goes in and then the machine that is the business model converts this money into growth and by extension into new technology. It is fuel and often not the end in itself.

There are N number of ways a startup can make money. Usually they fall into four very broad categories. The company creates something then sells it. The company does something and charges for it. The company connects and facilitates a transaction between N number of people. Or the company gives something away and gets a lot of people using it, then sells their attention to other companies who want it.

Until a startup succeeds it is like a whole lot of potential energy waiting to become active energy. There are a lot of ways to start a company but eventually it has to make money. The key word is eventually. It doesn’t have to happen at the beginning but so long as it happens, and the company doesn’t die beforehand, it doesn’t matter when.

The total amount of money a company makes is called revenue. Subtracting the amount it costs to run the startup from the revenue gives you the profit. This is how much you have left over after paying everyone, making all the products and running the business. However in the internet world is a weird phenomenon. Companies that don’t make any profit.

Implicitly you feel like something is wrong. Because it breaks one of those fundamental business rules. But what is happening is that the company is trying to grow quickly until it has a lot of users; then it will switch on, the way you might switch on a light, the ability to make money later and will theoretically make a lot more because it already has so many users. With some complicated portfolio optimisations by the financiers and vision by the founders themselves, you can create companies that don’t make any profit but still make economic sense.

This is best illustrated with a thought experiment. If you had unlimited money, you could theoretically grow a company until every single person in the world was a user of yours. You could do this until you’d destroyed every other company and the only one left is you. Then you can charge whatever you want and consumers would have to buy from you because there is nowhere else to go. What is happening is a variation of this phenomena. This is a super risky strategy and can only be done if you have enough money to start off with.

 

How can you tell when a startup is succeeding and making the transition from potential to active energy? You’ll know because it has reached something called product-market fit. It is when the perfect product is created to serve a market need. That means there is an intersection between what people want and what you have created and you just have to find them.

In practice it’s when the demand for your product is so great it outstrips your supply to produce it and you are in a constant game of catchup. When this happens to you, it feels like you’re getting torn apart from the inside. You just get overwhelmed and struggle to keep going but you have to. It’s like the growing pains felt during puberty which occur in the transformational stage as a startup becomes a big company. So you might say the role of a yet to be successful startup is to search for this magical place that is product market fit. Once they’ve reached it, they’ll know because they won’t be able to stop growing.

A good sign of a badly designed or badly positioned product is when it is difficult to get people to use or sell. Selling is hard but when something seems nearly impossible to sell it is probably because no one wants to use it. The same works if it is difficult to get customers to return. Consumers intuitively know what’s in their best interest. One sign a product is truly great is when it is easy to sell and customers naturally come back to use it over and over again.

When the business model fails is when product market fit hasn’t been reached. So the company is going to fail. A company that doesn’t grow, eventually over enough time, is going to fail. So if a company stays flat and doesn’t grow, sooner or later a growing one or a big company will come and steal its lunch. So the best defense is a good offence and the safest position to be in is a growing one.

After the startup has grown large enough it stops being a startup anymore and is a company. Now when people work for it, they say they work for X company instead of X startup. The work they do is called a job. Most big companies used to be tiny startups once upon a time and nearly all jobs are of this nature. The jobs are part of the organisation that is the larger slower bureaucratic version of their startup selves after they have gotten big.

In the startup phase, they were very close to the market. But now there is a layer of organisational abstraction separating the market from its employees. And it’s more difficult to see the effect of your actions on the larger organisation.

The place people who have jobs work is called an office. An office is a relic of a time where communication had to be face to face so the easiest way for employees to intercommunicate was to put them all in the same place. Technology freed this arbitrary barrier and now you have a recent phenomenon of a remote geographically dispersed workforce.

The people who work for a company help to create this thing which is called a product. The startup gives other people or companies use of this product in exchange for money. The people who give you money are called customers. The reason customers pay for the product with money is because the product makes their lives better. And their lives being better is something worth paying for.

It doesn’t really matter how a company gets their money so long as it does. This might sound unscrupulous but in practice it’s true. Most consumers don’t really care about the morally questionable actions of a company so long as it makes something that solves their problems. There isn’t a terrible amount of morality in the business world and I think it starts with customers. If customers didn’t buy from evil companies, then they would go out of business. So long as people keep buying from them, they will still be around.

Usually this is done by gaining their trust. So the customer trusts the company. More specifically what they are trusting is called a brand. It would be too difficult to show a consumer the inner workings of a company and most companies probably have things they wouldn’t want people knowing anyway. So they package this all up and display it as a brand. The same way a dysfunctional household usually puts up a happy pretence around dinner party guests.

So when customers refer to the company, what they are actually referring to is the brand. The brand is the sum of the actions of the organisation. So if anybody working for a company does something good or bad, it reflects on the brand. It’s pretty important to have people like your brand because if they don’t, they also don’t like your company or your products.

Notice the key word is product. All aspects of a company are centred around catering for this product. There are people who make the product, people who fix the product, people who convince other people to use the product, people who help solve the problems of those using the product, people who decide what the product does, how much to charge for it, what to add etc etc. The decisions are endless but the key thing to remember is it is all centred around the product, so anything not product-oriented is ergo unnecessary.

This is a huge change. Because most companies are not product companies. They are service companies. So the value is a function of the time value of the person investing their time. You’ll be able to recognise these companies because their employees charge per hour. But each persons value is distinctly separate from the organisation and there is a natural ceiling because there are finite hours in a day. The way these companies scale up is by adding more people.

Whereas in a product company, it collectively pools the time investment of a lot of people into one big entity which represents everyone in the company. The entire company lives or dies of the success of their products. When this fails the company dies.

But when it succeeds, there is no limitation to how big it can grow. The product can grow as big as the need for it and the capacity to produce it. So you can create X instances of the product for adding each new person. Soon the number of people using the product and how much revenue it makes far exceeds the cost of the number of people it took to create it.

It is like the idiom of putting all your eggs in one basket. Conventional wisdom says not to do that, which is why the creation of new companies is such a counter-intuitive process. Because you do need to put all your eggs in one basket and then make that basket succeed.

Eventually I think all service companies will become product companies. Because the cost of product creation has come down so dramatically, eventually a product will be created to cater for nearly every niche, which currently service businesses are solving. Every inefficiency will have an entrepreneur creating a product company to solve it, if the inefficiency is great enough.

This will be a beautiful thing because of what technology does. It makes peoples lives better and frees up their time to do other things, like create more. So the end result of increasing technology will be flourishing art and creativity. I don’t think it’s coincidence that the amount of art production and consumption seems to be increasing as technology is getting faster. They are a function of each other.

 

Hidden in the notion of a startup is also the belief in a meritocracy. A meritocracy is where people are rewarded in proportion to the value they create or their output. This is a very good thing. If smart people couldn’t get rich from starting companies they probably wouldn’t do it.

I think wanting to be rich is a very good ambition to have. Or more aptly, not wanting to be poor. I don’t think the allure of wealth alone motivates people into starting companies. It’s also the fear of being poor. Not wanting to be poor is a very motivating ambition.

In every startup is a number of unpleasant business things that have to be done. Your average inventor is probably not capable of bringing their invention to market without it. Because it’s so hard, you also have to really want it or it’s not going to happen.

But very quietly hidden in meritocratic ideals is a prejudice that people aren’t built equally. That some people deserve more because their output is higher. And while it may sound unpleasant, this is entirely accurate. Because it’s not people judging and allocating wealth. It is the market. And markets always arrange in certain patterns. A common one is that people who produce more will always be wealthier.

So to say that wealth going to those who create more being a bad thing is kind of a misnomer. Because what it is also saying is that you don’t like nice things. Because the machine that creates nice things also makes people who create them rich.

Neil Gaiman in a talk reiterates constantly “Make Good Art!” He’s not just talking about art for arts sake. He’s talking about people who make good art will be rewarded greater than those who don’t. So the best way to become notable and by extension wealthy is just to create something good.

Market based meritocracies pretty much only value money. It is like the internal points mechanism, like a scoring card to see how well everyone is doing. It measures money, it creates money and it recirculates money. It’s all part of the machine.

I think it’s important though not to become too obsessed with this because you will start to lose a part of your humanity. I wouldn’t really want to live in a world where everyone thought like that. Because for a lot of people, they don’t care about money. It’s just this thing you need to get in order to survive. And if you can get it doing something you love, all the better. How much of it you can get isn’t the important part but how much of something you love doing can you do while earning the maximum amount of this money thing.

The reason startups are important is because they are one of the primary engines that bring new technology into existence. New technology makes life better for lots of people. So any vehicle which makes life better for a lot of people will be very valuable. How valuable? As valuable as the number of lives it improves. So if a product improves the lives of a lot of people, the company will inevitably grow into a big company to satisfy this demand.

A company that does not have the potential to grow into a large company is not typically valuable, except to the founders of it and its customers. It doesn’t have the ability to meaningfully impact the economy or an industry. The side effect of impact is growth. So the measurement for how impactful a company is being is how big it is.

These are lifestyle or non-high growth businesses and they are like a species who simply do not want to take part in the evolutionary process or the often binary world that comprises startup outcomes. They are optimised not for wealth creation but for happiness. This is perfectly ok. Inksimple is one and I’m happy with it. It doesn’t need to grow into a huge company to make me happy.

But if everyone started a company with the goal for the company to be around in 10 years, I think there would be a lot more successful companies. Because if you were going to work on a startup for 10 years, how big could it get? That alone would stop a lot of startups that seem silly from getting started in the first place.

I think most companies should be started with the aim to go public. If every founder planned and tried to take their company to IPO. There would be a lot more success stories. That’s not to say every founder wants an IPO. But it is the right trajectory to aim for because an IPO takes a decade to accomplish in most instances and creates a lot of wealth, not just for the founders and investors, but for society.

 

The way a startup succeeds is actually Darwinian. They win by a form of natural selection. A company does something very specific in a small niche to avoid predators until it grows, specialises and dominates that niche then expands into other niches. This is the same thing most animals do to ensure survival.

I’ve always found the business school theory for explaining this phenomenon to be a little outdated. I think Darwin’s Theory of Evolution and Natural Selection to be a better one. Since built into the theory is the missing randomised luck element.

For the most part the entrepreneur is glorified as having captained their ship through stormy seas. But luck has a lot to do with it. Much of a company’s success is just pure luck. That’s not to say luck is the cause but rather the catalyst. It’s worth defining, what is luck and how do you make it?

You can’t. Almost by definition you can’t make luck. But you can optimise it and if you are optimising luck, it means you are becoming luckier. The easiest way of doing that is to always be prepared to try new things, meet new people and take leaps of faith. When you take a leap of faith and land safely, to the outside it seems like you just got lucky but really it was manufactured. And if luck has such significance in the variance of startup outcomes then it is of course worth doing.

One way to make luck is to keep yourself alive for longer, so more things happen to you. Just being alive for longer will act as force multiplication on the outcome when something lucky does happen. What that equates to is not to spend much money or as little as you can because the amount of money you have is an indication of the health of a company. It is like the health bars in old video games. Having a lot of money is like having a lot of health bars so it’s hard for enemies to hurt you. Conversely, not having much money is like having very little health so everything can hurt you.

 

How do investors fit in all of this? They act as catalysts. They facilitate the reaction of a startup germinating and reaching a market. The means by which they do this is by adding capital. I think the best analogy, if you remove the negative connotations of it, is they add fuel to the fire. A car can drive a lot further distance if it already has a full tank of petrol.

It’s commonly said that startups follow a power law distribution. This is always said matter-of-factly like it is implicit or assumed knowledge. But I’d be curious if anyone really knows why. For the smash hits, I think it is more they seem to catch lightning in a bottle and build a company around not letting it escape. A task not easily replicable.

But in the aggregate are broad lessons. Think about your own life. Who are the smartest, most productive, determined and ambitious people you know. I’d bet they seemed much smarter and more productive than everyone else by a long shot. That is why startups follow a power law distribution. Smart people who have a high output usually create important things. So if you don’t see yourself as smart and productive, you probably won’t make a very good founder.

So when a smart, ambitious and determined founder also catches lightning in a bottle you have the power law distribution at work and something very special is about to take place.

What investors do is buy risk. The founders can raise investor money and pay themselves a salary thus removing most of the financial risk of starting a company. Without investors, the founders aren’t able to pay themselves a salary so take on all of the risk of the venture personally. But then they’re rewarded with a higher stock ownership.

If a founder really thought their company was going to succeed big, they would try to sell as little stock as they could by raising as little money as they needed. They’d try to succeed big without raising a huge amount of money. Because it means they think that stock will be more valuable one day. Counter-intuitively, that is probably the stock investors most want to buy and the companies they most want to invest in. Or the company can grow slowly and what they buy then is time.

The three dimensions are money, risk and time. Each impacts the other. You typically have a lot of two of them and very little of the third. If you have a lot of time and money then there isn’t much risk. If you have a lot of time and risk then you won’t have much money. If you have a lot of risk and money you won’t have a lot of time.

An investor funds startups knowing most will fail. But the ones that succeed do so really big, displacing large companies and becoming large companies themselves. Then new startups come along and displace them again. The cycle continues ad infinitum. This is the commercial process by which most new technologies are created and the world becomes a better place.

When you don’t fund startups and create new technologies and industries, it is a akin to saying that progress and innovation have stopped because those things are so heavily intertwined it’s causative. Lots of startups also act as Cicero once said “Hannibal ante portas” meaning the enemy is at the gate, because just their presence forces big companies to not be complacent, invest and innovate for fear of being disrupted.

There is a common line of thinking that if something makes a lot of money there will be no incentive to disrupt it. This is true with those already established. If a company is making a lot of money doing something, there is no reason they will want to change or invest in things that will disrupt their core business. But there is all the incentive in the world for a new startup to want to disrupt it. So built into the mechanism is the means of displacing it.

But that means allowing big establishments to fail when they are disrupted by new companies. When they aren’t allowed to fail, as occurs in certain industries such as fossil fuels, healthcare, banking, finance etc where sometimes a government will give grants or bail out a company, it prevents an industry from really innovating. Because innovation then becomes top down instead of bottom up. The innovation is protectionist, with the goal to protect and expand what already exists instead of to create something entirely new, which may displace everything but unlock entirely new industries.

This is also a self selecting process because it is how big evil companies are put out of business. A lot of people believe that corporations are just bad. You see on the news words like business, profit, capitalism and corporation used as if they were dirty words. What people see are big evil companies who create negative externalities and make the cognitive leap that all corporations therefore must be bad. But the best way to kill a big evil corporation is to create new corporations which then disrupt and displace the old ones making them irrelevant. So when people dislike Monsanto, they should go found and fund more biotech companies.

Startups are a market course correcting itself. Big companies become large and then they get disrupted if they do not innovate enough. In every story of a big company going under, they would have foreseen and been able to prevent their own doom had they acted fast enough. It’s a type of cognitive bias where people like to believe their success will be everlasting and they don’t need to change. But this is akin to wishful thinking. The easiest way for big companies to not be disrupted is to buy the tiny startups disrupting them.

 

When you’re the founder or a shareholder of a company you have this thing called fiduciary responsibility. What that means is you have to make as much money for the company as you can. And not do anything to get in the way of the company making as much money as it can. This is actually a good thing because startups are an engine of growth for the economy the same way a business model is an engine of growth for the startup.

What is the economy? But a network of how companies interact with people. What are companies? They are clusters of people. So really it’s how people interact with each other with money as a value medium. So startups are in a sense points of highest yield. A thing which requires the least amount of effort but yields the highest returns or results. Startups are like that for the economy because they can be started for not a lot of money and create a lot of jobs and innovation. Killing N birds with one stone.

It is always a good thing to create big companies. Because they unlock new industries and new types of jobs. Entire new professions spring out of nowhere due to the advancement of new technology. But it also destroys jobs by making certain types of industries obsolete. Agriculture was the bulk of most nations economic growth in the early 1900s, now it represents nowhere near that.

There is a degree of recoil that new technologies also destroy industries so they destroy more jobs than they create. If technology automates or destroys most jobs, what will people do? It’s something that troubled me with the theory because at first glance it seems entirely true. Only after thinking about it more deeply does that assumption return false.

In the short term it may destroy them, but in the long term technology always creates jobs. The problem is the long term tends to be a while away. Technology has eroded the safety that big corporations used to provide because they risk becoming displaced. When a ship is sinking it needs to trim excess weight to stop it from sinking faster to give them a chance to repair the ship. Often the excess weight is actually people. The most expensive resource is always people, so all things being equal big companies try to avoid hiring unnecessary people or by paying them less.

If you can buy something for X, why would you voluntarily pay more than X to get it. If a coffee costs $3, you’re not going to want to spend $6 to buy it. People make these micro decisions every day. But that’s also the kind of decision corporations make when hiring people. That analogy angers people when stretched to humans.

In an economic sense what people are selling in a job is their time in exchange for money. The more talented a person is, or the more value they create, the more that time is worth and thus the more they get paid for it. But if a corporation can get away with paying people X, even if X isn’t very much, they will.

So when someone isn’t earning very much people infer they have less talent. What they’re actually saying is the market doesn’t appreciate their talent. Economically, this is true. Few things price something as efficiently as a market. As their talent grows, for example by learning and mastering skills, the value they create grows proportionally. The mistake is when people derive their self worth from their income. Because at any point technology and society can change and their value will be made non-existent.

There is a very subtle inflection that happens where people see those not earning much and make the cognitive leap that they therefore deserve not to earn very much. Because the market does not value their talents. But people should not be judged by a market. That is decisively a losing battle. When society stops valuing people by how much they earn I think will be a huge leap forward for humanity.

The entire notion of a job is a societal construct and a recent one at that. A job is not a real thing, it doesn’t define a person. It is just a series of tasks people carry out in exchange for money. If hypothetically we have created perfect technology and all jobs are automated or destroyed, society will have to reconstruct itself in such a way that having a job is not the only means for people to provide for themselves.

The problem is this is a governmental problem to solve not a corporate one. Corporations are not going to suddenly create more jobs than they have to. Given a choice a company will choose a hundred robots over a hundred people every time if it turned out to be cheaper. Wanting corporations to act morally of their own accord is a losing battle. It’s like wanting a hungry tiger not to eat you. The corporation is designed to extract as much value as it can. The government is designed to disperse that value throughout a populace. These two interests should be decidedly separate.

A good example is minimum wage. If a government enforces a minimum wage, then it is impossible for companies not to follow it. A government is supposed to reflect the interest of its citizens so if they’re not doing that then it won’t happen and means the government is failing its people. Companies will always have to listen to the government because they create laws. Companies will always follow the law. But if you make laws that hurt people then you are just making sucky laws.

The government will have to keep up with technology to fix this but it will happen. I have hope in the human race that it will get over such a silly limitation as job dependency. There are already promising initiatives such as a universal basic income which fixes that very problem.

During the macro trend from horse driven vehicles to engine powered vehicles, nearly all of the stables went out of business, destroying all of the jobs associated with raising horses. But all of the progress that came with cars would not have been possible if you tried to protect the jobs of the stable owners. In fact it’s hard to imagine a world today where everyone still rode on horses.

That one analogy reassures me that technology is a good thing and I’m just not able to see the benefits of it yet. It stems from the inability of people to see into the future and think back to the present. Project yourself 10 years into the future and then reverse engineer.

If you imagine the future and envision ways people living in it will see us as being primitive today. Then you have a similar idea of how new things will develop – it will develop in the direction to remove any inefficiencies that are the reason things seem primitive today. If you know where you are going, you’ll figure out how to get there. Future generations will see incurable diseases today like Cancer the way we see Influenza of the 1920s because technology will have wiped it out.

It’s very easy to think that society is this big unchanging cluster. But in reality, it is in a constant flux of change. It wasn’t something I grasped until recently. Because when a change does occur, society subsumes it and it seems like it has always been there. I always imagined it like a glass of water. If you put a pebble in it, the water will rise. Removing the pebble, the water will fall. But from the vantage point of the water, it is almost as if no change has occurred at all.

The same notion is applied to school, university and jobs. To us it seems like everyone goes to school, university and has a job. That’s because they do. But it wasn’t always the case and this is in fact a recent trend. Less than 200 hundred years old. Before that everyone had apprenticeships until they mastered a craft, at which point they would practice their craft for the remainder of their lives.

Everything that makes up society is created by other people. Everything. Think about that for a second. Everything you have ever experienced or witnessed has been created by another person. As N people create T things, you have built up a society and everything you see around you. So the barrier to entry of creation is actually very low. However the barrier to adoption is high, precisely as high as peoples need for new things and is synonymous with the commercialization process.

The progression and sophistication of humanity can be tied to the creation of new ways of doing things. Nearly every improvement in society is based on some core technological breakthrough. Many of which go on to revolutionise everything – as happened during the Industrial Revolution. Today people have a lot more freedom and choice than they used to. A lot of that freedom and choice is a direct result of technologies that came from startups. And so cultivating startups proves to be one of the easiest ways of creating new things and making life better for lots of people.

Commercializing new things helps accelerate the creation of them. New things happen faster. The net result is a bit like learning how to learn. If you learn how to learn, everything you learn subsequently becomes easier. I once read a line from YCombinator that pretty much summed up how to succeed in a startup. It comes down to: Make something people want. Spend as little as possible. If you do those two things, you win.

Simple, right?