Big companies frequently have large bureaucracies and detailed processes in place for doing just about anything. Do you want to change the user interface for the Sign Up page? You'll need to run a focus group to see what your users think about the new look. Do you think your company should take on a new project? Write a 10-page design spec and get the VP of Product, the head of marketing, and the lead security engineer to sign off on it. Are you releasing a product upgrade soon? Follow this 63-item checklist. These rules and procedures exist to make sure everything a company does is polished and consistent. These rules also make it next to impossible to be nimble and to experiment since every significant action has a high activation energy.
In chemistry, the activation energy of a chemical reaction is the minimum amount of energy that needs to be applied to make the reaction proceed. Activation energies in business are the same: depending on the company, it can take a lot of energy to make something happen. One of the wonderful things about startups and small businesses is that activation energies are much, much lower than at big companies. Do you want to change the user interface for the Sign Up page? Make the change and push it to your production website. Do you want to work on a new product? Get the CEO's approval and go ahead -- or just go ahead if you happen to be the CEO. Are you releasing a product upgrade soon? Follow this 4-item checklist -- and if something goes wrong, email your five existing customers to apologize, then move on.
When you don't have rigorous processes or thorough checks and balances, the initial barrier to taking action is low and it's easier to make mistakes. However, your can also move more quickly, change directions more easily, and experiment more frequently. For tech startups whose priority is typically learning rather than revenue growth or feature optimization , being able to experiment is an invaluable asset. Unless you're making medical devices or dealing with financial regulations, you should be taking advantage of the utter lack of bureaucracy at a small company and trying out as many ideas as you can in order to learn what resonates with your target customer.
Because their primary goal is learning, most startups should have only the most minimal processes and approval chains. Every time you add a layer of checklists or forms or people who have to sign off on something, you're stifling your company's ability to learn quickly. While big companies have a lot to lose if they makes mistakes (e.g. a 10% drop in user registrations might result in millions of dollars in lost revenue), startups have much less at risk (i.e. that same 10% drop only costs the company hundreds of dollars in lost revenue). Not having as much at risk is a huge advantage. You can make multiple tweaks to your website's search results every day for a month while Google, a company with far more resources than you, might try just one change per week because every tweak requires triple-checking and approval by a small army of managers and directors .
I've worked with companies that neglected to embrace this principle and wound up handicapping themselves because of it. One such startup had a few dozen employees, but its bureaucracy would have been better suited for a few hundred. Between their overly formal quarterly performance reviews, committee approvals for minor decisions, and territorial VPs, this company rolled out new features more slowly than its competitors and experienced low employee morale. It was sad to see them hamstringing themselves. I think they assumed that one needs to copy big, successful companies to be "legitimate," but the assumption is faulty because not everything that big companies do is worth emulating for a small company. Having 3 layers of management in charge of a 4-person team doesn't make your company legitimate, it just makes it a crappy, inefficient place to work.
As your company grows, consider what you can do to maintain a lower activation energy, as well as what you can do in that state that may be harder to do in the future. Innovation happens much faster when the barrier to getting things done is minimal, and being part of a startup can offer a rare opportunity to see what people can accomplish when they're not impeded by management and processes.
 In its early days, the highest priority for a technology startup is to find product/market fit. Roughly speaking, this means you have a market in need of a product and a product that satisfies that market. If you have a great target market but don't have a product that appeals to it, or if you have a great product that nobody wants, then you don't have product/market fit. A typical young startup spends a lot of its time experimenting with different features and customer bases until there's a good match between what the company makes and what its potential customers desire. Only after this stage is reached does a company start to think about growing its customer base and its revenues.
 Note that Google tries to act smaller than it actually is in several ways. Externally, many products are perpetually in "Beta" mode, which lowers users' expectations and makes it more acceptable to change or occasionally break product features. Internally, the company has a fairly flat hierarchy relative to other companies of its size. I spent a few months at Microsoft after college and discovered that Steve Ballmer was about a dozen levels above me on the corporate ladder; at Google, I was only 5 layers removed from Larry Page and Sergey Brin. When I worked at Microsoft, it was a 50,000-person company that felt like a big city. Google had 5,000 people when I joined, but in many ways it felt like a 200-person startup, and the speed at which it launched and iterated on new products was a reflection of that.