Startup School Y Combinator: Nontrivial Aspects of Teamwork [Part 2]

Stanford course CS183B: How to start a startup . Started in 2012 under the leadership of Peter Thiel. In the fall of 2014, a new series of lectures by leading entrepreneurs and Y Combinator experts took place:



Sam Altman: And also this kind of communication is a great opportunity to find many new employees. And after you hired these first people, and they started working in your company, how did you help them quickly begin to work qualitatively and take a suitable place in the corporate culture? Finding new employees is often not easy, but still not as problematic as making it so that work brings them joy and is effective. What have you done to ensure all this?
')
[ The first part of the lecture ]

Ben Silberman: Well, the methods we use have changed since we became a larger company (Pinterest). When we first started our activities, we "shook" over each candidate. And the whole procedure of introducing new employees into the position was that we told them: "Here is your computer, we have already organized your workplace, and do not worry, we will solve this problem together."

This is typical for start-ups: we all worked in a three-room apartment, and such things as everyday communication and joint pastime arose by themselves. It was not necessary to do something specifically for this.

The only thing I want to add is that we have always reminded the staff that we want to ultimately achieve. Because it is very easy to leave a person alone with the solution of a problem, and it will seem to him that the whole world has narrowed down to this tiny problem. We always said that one day we’ll do the same thing for Google that they did to search for information in general. And we are working on doing this.

Now that our company is constantly expanding, the solution to this problem is becoming more formal. We spend a lot of time trying to remember what an employee was like from the day he first appeared in the company and the interview until the 30th day of work in the company.

Do they know anyone from the company? Do they know their manager's name? Do they get along with people from their team? Do they know the basic structure of the company, and what are its main objectives?

We have a program to solve these problems. It is designed for a week, and within its framework, employees perform certain tasks in order to better adapt to the situation of the company. And this is what we have always tried to develop. The main information that we seek to get through this program is what they thought about the company immediately after they started working, then 30 days after that.

Then we also ask their colleagues and managers if this person knows about the company. Did we manage to help him become an effective worker? If it turns out that it was not possible to do this, then it is very important: a) to stop hiring new employees, since the arrival of new people does not improve the efficiency of the company; b) reorganize work with new employees.



John Collison: I believe that there are two approaches that will be useful at any stage, and the only way to use them will be different. The first approach is that you load employees with specific work as quickly as possible. So you determine what problems they have, and will be able to assess the progress of the employee on how much real work he manages to do.

Therefore, when we hire engineers, on the first day we try to occupy them with developing something. When we hire people for business-related positions, on the first day we take them with us to real business meetings, the holding of which is their job. Sometimes it is easier to act carefully and gradually introduce a person into the workflow.

We usually prefer to give people serious work right away.

The second approach is that from the very beginning we make comments and give comments on their work, and, of course, tips on how to adapt to the corporate culture of the company. And, in turn, if your organization has a well-thought-out culture, which is what business representatives are trying to organize in this audience, then a new employee will need some time to adapt to it, and not the fact that it will be easy.

In Stripe, our corporate culture is largely organized around working with textual information, so all employees sit side by side in headphones and communicate with each other through an instant messaging system. And for many people who come to us and start working, this situation is very unusual.

Patrick Collison: For normal people.

John Collison: Yes, exactly. Therefore, regarding everything, starting from the ways of the most effective work and ending with minor details of the corporate culture - the more you explain everything to employees, the better they will work. And at the same time, it is not quite the usual thing to explain to people whether they are doing their job well or badly. In ordinary life, you do not do this - I think you are quite restrained. But when working with subordinates, you have to do this so that they do their job correctly.

Sam Altman: I think, after the foregoing, it is appropriate to proceed to discuss the situation when your companies have already become large. What are the most significant changes you made to the recruitment policy and management teams of employees after your company did not work two, but ten or a thousand people?



Ben Silberman: There were a lot of changes. As for the management of teams of employees - we try to make them extremely independent and give maximum freedom of action within the company’s rules.

Thus, over time, we wanted to organize everything in such a way that our company was a kind of startup consisting of many startups, instead of a single organization with a common policy for all of its segments. But it is easier said than done. I do not think that we succeeded in accomplishing this until the end, but, nevertheless, now each team of employees is managed independently and has the ability to independently organize work with the necessary resources to accomplish the tasks set.

They know what is most important in a given situation and how to determine it. So the problem of company management becomes more solvable. On the other hand, if you can not divide the company into independent segments, the solution to this problem seems absolutely impossible. Faced with it, you just think: “My God, the complexity of managing a company increases exponentially - we will fail.”

The formation of such independent segments is necessary. In any case, this is what we are trying to implement.

But, for example, with us on Pinterest this is complicated by the fact that we strive to shape these segments around the most talented designers, the most experienced engineers, authors of articles and people who represent the company to the public.

We want them to be independent and independent. All of these conditions make it difficult to achieve our goal, but they are key to our approach to product creation. We bring together people who are interested in the various activities described above, assign them to a specific project, and then try to eliminate everything that might hinder them so that they can act as efficiently as possible.

If new obstacles are discovered, we try to determine how we can still speed up the work of our employees. I think the recruitment process has also changed a bit. When more and more people start working in your company, the most significant change is that recommendations on the choice of candidates to a greater and greater degree begin to come from the team of employees that you have already hired.

And one of our most successful and useful decisions is that, as our 14th or 15th employee, we hired a professional recruiter. She worked with startups as well as with large companies such as Apple. She could see in advance what an employee would be like and could determine signs of the presence or absence of talent. And she taught us not only how to determine whether a person is talented, but also whether he “fits” well with the company's corporate culture. And now, remembering how everything turned out, I personally really appreciate the fact that she began to work with us.



Patrick Collison: A huge number of problems associated with the development of the company's management system. And either your business becomes unsuccessful very quickly, or all your problems are reduced to the issues of management improvement.

What usually takes people by surprise and also took me by surprise - this is how rapidly the magnitude of the time interval for which you plan your activities is changing.

After a month of existence of the company, you most likely plan your business for a month in advance, aren't you? At this time, you are drawing up a development plan based on who you work with, perhaps there are informal relationships in which employees have not yet finally decided whether they will work with you on a permanent basis or not. And the period for which you are planning your business depends directly on how long the company exists.

Also, you should consider the process of finding new employees - at the beginning of your activity you need to hire people who will work efficiently from the very first days. You cannot afford to hire employees who, as you think, have great potential, but they will be able to realize it only in a year or two. However, a year or two years after the creation of the company, it makes sense to make such decisions.

Moreover, if you are not doing this yet, then perhaps you are planning a business for too short a period of time. That's what matters.

On the one hand, such problems are easily solved. For example, how to establish reliable relationships with people? But we all do it every day. How to do it systematically and at the same time effectively within a large company?

The solution will always be to use approximately the same methods that you would ideally use for a small company - you only need to determine what needs to be improved so that everything works as efficiently as possible in a large company. A rapidly expanding company, whose number of employees is increasing by one or two people per year is extremely atypical.

What would be the best way to manage this level of employee growth? I think it would be right to use a systematic approach to solving this issue and finding ways to manage it. It is necessary to understand that the only thing you can do is to correct the methods used in the initial stages of the growth of the company. In Stripe, we sit down three times a day and eat at long tables. And it is quite obvious how active the communication takes place during such meetings. And the ways of organizing such communication can be a huge amount. I believe that the work of the company largely depends on this.



Ben Silberman: I was very interested in one moment. You guys are serious about informational openness, and I wanted to ask, have you improved it over time? We, for example, are constantly thinking about how to do this. I'm just interested in your opinion.

Patrick Collison: I can't remember who exactly gave this definition, but a start-up is an organization that has not yet encountered the problem of the relationship between the principal and the agent. And in the case of large companies, what is acceptable specifically for you is often generally unacceptable for the entire company. And as a result, in a large company you cannot do many things that are useful to you, absolutely openly, as employees simply will be disappointed in you. But in the case of a startup, everyone is moving in the same direction, and therefore you can make all the information publicly available.

As I said earlier, a copy of each outgoing letter came to Stripe to all employees, except those who refused to send. We thought that this would be more efficient, and no additional meetings would be required if everyone was aware of what was happening. Over time, we have created a complex system of mailing lists. We are currently using a program that generates Gmail email filters. But a few days after we had 50 people in the company, we asked the staff how they were doing their job. And they replied that everything was just awful, since they could not find all the letters that were supposed to come. So they missed some information.

John Collison: At some point, Gmail crashed.

Patrick Collison: Right. At some point, Gmail crashed due to the fact that we sent too many emails. Thus, it is difficult to raise the level of informational openness, because, for example, you can share with someone from the company some excellent idea, and at the same time, the person sitting opposite will consider it the stupidest idea he has ever heard.

All information you share is thoroughly assessed by the entire company. All this is a source of problems, although on the other hand the fact that employees are sufficiently aware of what is happening in the company is a positive thing. I can not bring any more convincing arguments in favor of this, except that everything has worked so far. And I am really interested in how this will work, when there will be five thousand people in our company, if we ever expand to this level.

John Collison: I think that two things helped us increase the level of informational openness: we changed the applied tools and transformed the corporate culture regarding this. Regarding tools: in the past, employees could learn about everything that happens in a company by reading the mail that comes to us. Now, every week we carry out emergency work on notifying employees about what is happening in the company - the amount of information is very large.

Then, with regard to issues of culture: now a huge amount of information is available to employees within the company, and therefore it is necessary to create rules for its use. For example, it is obvious that some of the information may be confidential and intended only for Stripe employees. There are also less obvious situations, for example, when you exchange letters or communicate on Slack or IRC, which are currently being watched by 170 people [employees of Stripe] - it is easy to start worrying about the attention of such a large audience.

And it’s also quite easy by sharing with someone an idea that you thought was quite adequate, and having received in return a stream of criticism, to lose the desire to share your suggestions in the future. And it is necessary to introduce rules regarding when to enter into a discussion and how to behave during such communication, as employees now have to deal with a much larger audience.



Patrick Collison: I know that it's not good to put a girl in an uncomfortable position, but Emily had an internship at Stripe this summer. I wonder: as an intern, what do you think about our company?

Emily: Overall, everything is very cool. In the first week, I spent most of my time reading information about the company through Hackpad and trying to understand what you were doing. Often, much distracted from the information about my main job, as there were many areas of the company's activities that were also interesting to me.

Patrick Collison: By the way, the Hackpad is something like Google Docs, only with a news feed, where you can see all the documents.

Emily: And you strive to make publicly available information about everything that you work on. But overall, this allows you to expand your company fairly quickly. In addition, the company held speeches by team leaders, during which they talked for half an hour about what they were doing at the time and how we would be able to assist them if we were interested.

Patrick Collison: Do you think that informational openness benefits the company?

Emily: Yes. I remember how I could not remember which mailings I should have, and which ones should not be subscribed. In the first week about two thousand letters came to my inbox. But then I learned that there were only three or four teams, the information about which I really needed.

Sam Altman: Question to Patrick and John. Was it so that the people you first recruited were able to subsequently take up management positions?

John Collison: We in Stripe - yes. Many of those first ten people now occupy management positions. I think that many companies need to be well understood - people are not necessarily born with leadership skills or management skills.

You must be able to develop these qualities in employees and help people progress if they have been with the company for several years.

This requires a lot of patience, especially if employees are fussing and are in full arousal. But at the same time for the company is destructive, if its leadership is unable to develop similar skills in their employees.

Ben Silberman: I think that in our case, this question can be answered with both “yes” and “no”.

One of the advantages of working in a startup is that you can be entrusted with the task that no one in your right mind would have entrusted you with in another situation.

These may be management tasks or work with a project. And at the same time, if you ask someone to take such a risk, and the employee does not cope with the task, it should not threaten unconditional dismissal. Otherwise, the person will not dare to take on this responsibility. And we have employees who started work as simple programmers or engineers, and now they manage large teams. At first they said that they would like to start a separate project, take over the management of the group, and then the responsibilities of the manager.

There are also people who are very happy that they used this opportunity, because they know that they will never again want to take up such work again.

And we just said to such people that they don’t have to be in charge of management, and it’s enough to do everything possible for the company just by doing their job as an engineer or designer. But it is very difficult to predict how things will turn out, until you give a person a chance to try. Therefore, I prefer to give this opportunity to as many employees as possible. And only in some cases, when work requires a great deal of experience in business development, you are looking for someone who is likely to be able to successfully cope with this task.

I will also answer the question of how our initial goals differ from what we are doing now. When we first started recruiting employees, we wanted to create a really useful service that users would like. I like to collect different things, and I think other people are also interested. But what we did not expect, and what turned out in the early stages of our activity, is that viewing other people's collections is an excellent way to find what you were unconsciously looking for. This helps solve a problem for which many other services do not have a solution.



Over the past year, we have spent a lot of effort on creating a system of recommendations for products, a search system and displayed product lists, as well as optimization of work with unique ideas - “pins” that users themselves find and sort.

That is, there is always a gap between where we are now and what we should achieve. Now we exist for a long time, but it seems to me that this gap has become even greater. But I think this is typical of people who set up companies.

Sam Altman: The question is: most startups are different from Apple since the creation of the first iPhone, and you cannot guarantee that the grandchildren of your employees will remember their feat, since most of the start-up firms prove to be unsuccessful. How to convince people to work your start-up company?



Patrick Collison: It seems to me that in many ways people are attracted to work in such firms, because success is just not guaranteed. Otherwise, people would not be so interesting. In such a situation, there is only a chance of success, and nothing but this potential opportunity. Regarding the fact that employees do not see their families and children - work in start-ups really involves spending more time in the workplace, but I think that in this particular example, everything is a bit exaggerated.

In general, employees of start-ups themselves, where they have to work more in the early stages, tend to dramatize.

Similarly, people argue on fishing. In each startup, employees think that they have to work a lot more and longer than employees in start-up companies that were created before. It seems to them that they have not slept for two years already. I think, in fact, people do not have to give so much energy. On average, you have to work only for two hours longer than in a normal company. This is the price, however, it does not hurt to enjoy work in the next five years.

Ben Silberman: Even Apple was not truly an Apple company before everything was done. No sane person you hire will think that you are capable of predicting the future, and that applying for your job guarantees success. In fact, if you promise them all this, and they come to work for you, then perhaps these are not the people who should be hired, as they will not be completely reliable in the future. But I think it will be fair enough to tell you the advantages of your work and what your company seeks. You need to say what difficulties you are waiting for, and describe the best plan for your actions, as well as explain that their role in the company's activities is extremely important.

I liked how you said that if you are going to fly to Mars, it will interest the best people, and thus, you are already getting closer to achieving your goal and the new employees in turn understand this. What I do not support is the concealment of difficulties that will arise in the implementation of such goals. And if people come to the company for the sake of reliability, like Google’s, and at the same time want to work in the conditions of greater informational openness typical of small startups, then this is a very bad sign.

For example, when I conduct interviews, people often say that they are simply delighted with the activities of our company. In such cases, I usually ask where they have been interviewed. And if they call, for example, seven companies that are not related to each other, except that they are at the same growth stage, they say that they like to study and find things, and therefore they had an interview in Stripe, Jawbone and where something else - all this may be a sign that they speak insincerely. And honesty on their part is of great importance to you. And when it becomes difficult, such people do not try to accomplish the task at any cost, because they come to the company only to work in comfortable conditions, and not to create something useful.

Patrick Collison: I think that another factor that motivates people very well is getting a certain benefit for themselves, namely, the prospect of personal development of a person. In a startup, the requirements for employees are higher, since the working team is very few. And no matter if you are the best or worst employee in the world, you most likely will not be able to influence the success of, for example, Google. Therefore, if you really want to test your skills and find out how much useful you can do, then any startup will be the best place to do it.

Sam Altman: How does your audience influence the approach to recruiting new employees?

Ben Silberman: Traditionally, it is recommended to hire only those people who use your service constantly. And maybe this works great if you create an API. We are looking for people who have vision and are well able to find information on the Internet. They also need to know how our service works, and be able to use it. But they do not have to use our service by this time. And it suits us - we will try to find out what prevents them from using it.

Let them go to work for us, and we will eliminate this barrier, and also we will be able to better understand the desires of users. If you read the literature on creating a startup, you can find a lot of tips that look very relevant, but will be useful only if they are suitable specifically for your company. And so we had to expand the circle of searches and attract people who are interested in our goals, who are important, what products we create and how we do it, even if they were not among our first users.

John Collison: I would like to add a few words about how difficult it is to find the first employees: applicants have many other options, and your company is a kind of “ugly duckling” who has not yet turned into a swan. Hiring people who are excited about your business is the right way to find employees. This gives you an edge over other companies. In the case of Stripe, we hired four Stripe users at the beginning of our business, and these were people we would not have found in some other way. I'm sure something like this was the case with Pinterest, and people came to your company, realizing the benefits it would bring.

Sam Altman: Thank you so much for coming today.