How Can Start-Ups Scale Up?

What are the toughest challenges facing start-ups that grow beyond 50 employees?

Alter: It’s not one size fits all, but there are four general areas. First, your business changes completely from a product business to a distribution business, whether that’s sales or other aspects of making your product and getting it out the door. Secondly, do you have the right people? Are you changing out the people that you have? The role of the CEO as the leader has to change as well. Thirdly, cash. It’s expensive to scale, and if you run out of cash, it’s a real problem. Finally, many entrepreneurs struggle with the need for process, structure, and rules. It’s like asking an athlete to show up on the field, but not telling her the rules. You need a set of rules that everybody understands so they know how to play the game as you scale.

Lannert: There’s a fundamental shift beyond 40–50 employees. You usually get to 40 or 50 because you’re very good at doing, and suddenly you need to become good at leading, which is a different skill set. You can’t run in lockstep anymore. You need to start building systems and processes for communication.

Shekhawat: Most technology firms these days have offshore strategies to offset costs. When we had 4 or 5 employees in the United States, we already had about 30 people overseas. That adds a coordination problem: you’re separated by 11 hours across time zones. There’s also culture. We had our overseas group in India, and they marched to a different drummer. It took a while to get the cultures coordinated.

What kind of entrepreneurs make it through that transition successfully?

Alter: They’re adaptive, and able to learn from mistakes. They’re comfortable enough to bring in people who have more experience than them, who might be smarter than them. They realize that their job as CEO is to make the company successful, not to do everything themselves.

Lannert: One mistake I made is that when you realize you need help in managing, communicating, and keeping everyone in lockstep, do not go into every department and promote your best performer into management. There is a difference between doing and managing. Professional management is a skill set. We took our best doers and put them in positions of management. It took a couple of years to unwind that.

Shekhawat: In the early days, I always thought of a good employee as being a Swiss Army knife: a generalist with all these different tools. I learned that you need specialization. Before we hired our first head of quality assurance, QA used to consist of beer-and-pizza parties. We’d try out the product on a Friday night and catch as many bugs as we could. This QA expert came in and said, “This is idiotic. You cannot treat QA as a screen at the end of the development process. You’ve got to treat it as something you build in at every step.” I looked at him and thought, where have you been all my life? It was my transition moment from generalist to specialist.

Alter: When we tell the story of a business, it’s often about increasing success; but the reality is, it’s much more like a heartbeat. In those points when you’re down, and you need somebody to plug into a particular role, you may find a person with skills and experience, but he might not be the right cultural fit. Another challenge is turning focus from a set of customers that got you to this point to a new set of customers. Initially, most in that first set are probably early adopters. They found you and they put up with some of your bugs, and now it’s time to turn to people who are expecting more of a finished product. What works for them is usually very different from what got you there.

Lannert: When you’re hiring, it’s not what people have done; it’s what they can do. Jobs, products, and the economy will change. There’s no “rinse and repeat.” It’s so tempting for a start-up to say, “Please bring in your playbook and let’s hope and assume it’ll work.” It doesn’t. Don’t become enamored by a résumé. If I’m hiring someone mostly because I’m impressed with a résumé, it’s not the right hire. I need to be most impressed by how we start to brainstorm in an interview, by her curiosity, by how she builds relationships.

Shekhawat: I look for attitude, and then a core set of skills, and then a demonstration of how you contributed those skills at the last few places you worked. The other thing to bear in mind is that when you pull a group of talented people together, it doesn’t tell you how they will work with each other. You have to focus everybody on the thing you’re building, outside of all of the individuals, and make everyone understand we are coauthors and no longer individual contributors.

Alter: As the CEO, you have to choose the things you’re going to do and the things you’re not going to do anymore. That sets the tone for everybody else.

How can start-ups maintain their entrepreneurial spirit as they grow? Are formal processes the enemy of a start-up culture?

Shekhawat: Process is not the enemy. Too much of it will kill you, but the CEO needs to determine how much process to have. There are a lot of fire drills in an early-stage start-up. It’s the heroic save, the diving catch—looking good because you saved a client. That does not scale. You have to cut out those things, and stop rewarding diving-catch behavior. It’s a balance between the diving and the team that got you to be in a good position to make that catch.

Lannert: Process isn’t bad; bad process is bad. These are helpful tools for a group of people to be productive at scale.

Let’s bring it back to culture. Culture isn’t an ethos or a feeling. It’s behavior, how you get stuff done. What do you tolerate? What makes you laugh? How do you celebrate? It’s rituals around behavior. At our company, we have the schmutz pact. Instead of saying, “You should give feedback,” we say, “Have you ever been to lunch with someone and noticed he (or she) has food in his teeth? If you tell him, he’ll thank you, or say, ‘Why didn’t you tell me earlier?’” Nobody wants schmutz on their face. Giving feedback in the moment is kind. So instead of saying, “You have a moral imperative to speak up,” we just say, “Hey, schmutz pact. Let’s have a conversation.”

Shekhawat: I think of process as a behavioral algorithm. It’s a set of steps, and, like with any algorithm, it runs its course, and then it has to be shut down or killed off. Your processes—your schedule of meetings, how long they run, who’s in them—should all be considered living things that don’t deserve to exist a moment more than they are necessary.

What exactly are the distribution challenges for fast-growing start-ups?

Alter: When you’re starting a business, you’re trying to figure out product-market fit: Will the dogs eat the dog food? The next challenge is: How do you get people to buy more? How do you get distribution? How do you create a reputable, scalable, repeatable process? People struggle with that. It’s so different from the initial start-up phase. It requires you to develop a process where people have to do the same thing over and over again. And often, the people who like to figure stuff out don’t like to do the same thing over and over again. It’s hard to build a scalable distribution channel quickly. Not only is it a real challenge to think through, but it can be expensive, especially if you scale too quickly.

Lannert: You go from having some of the founders or the CEO selling, to having someone else in the organization selling, to having a repeatable process where you can sell the same thing to satisfy customers profitably. Then you start to get to scale. One challenge is that the first sort of enterprising, really creative, iterative salesperson may not be a great sales leader, or cultivator of people, or establisher of process. One of our early salespeople is a magical salesperson, but he’s not good at setting up a methodical process. He wants to change everything every day and constantly iterate. Works great when you’re an independent contributor. It’s havoc when you have 20 direct reports.

Shekhawat: There’s a tipping point where you’re not having to push the product as hard to the customer; it’s starting to get pulled through. That’s when you know that what you designed, as a solution, is a good fit to a problem on the other side. And that has to be an important problem. It has to be owned by someone who has a budget, and then you have to be viewed as a successful company, because as a young tech company, it’s hard to get into large organizations.

How important is flexibility and the ability to pivot?

Lannert: We were going to die if we didn’t change. Jellyvision originally made games on CD-ROMs. When those died, so did our business model. And so we had to change, from a B2C gaming company to a B2B enterprise software company. We more or less shut down most of the company. We went from 70 people to 10 people so we could retrench and build all-new software. We rebooted back to start-up days, where you’re not only an incredible designer, you’re also a project manager and you take out the trash. It was almost a restart, more a rebirth than a pivot.

Shekhawat: Our story was less dramatic. I think of it more as a sailboat tacking, constantly adjusting to the wind, but generally heading in the right direction. There were a lot of dead ends, a lot of lost money and time. But I never wanted to end up saying, “I’m really far in the wrong direction and I can’t make my way back. I don’t have the resources. I don’t have the credibility.” It was about taking calculated risks, trying to solve anticipated customer problems, to become more relevant to the customer. Again, the more of the problem you solve on the customer side, the more likely your chance of being pulled through. Sometimes you get it wrong. We thought our customers had certain problems, but they really didn’t want to solve them at that stage. That led to development cycles that went sideways, but then we just became much better at solving the central issue.

Alter: The biggest challenges are the pivots you didn’t take. One of the toughest challenges of a CEO is to say no. As you’re starting to scale up, the world is your oyster, and you can do all of these things, but the most successful people are narrow and focused as they scale.

Shekhawat: There are a lot of shiny objects that look like quick revenue. Someone will say, “Why don’t you just build this for me, and I’ll pay you?” You do it, and before you know it, you’ve ended up with essentially spaghetti code: your code doesn’t make sense for anyone, and that doesn’t scale. We talked about scaling people; you have to scale the product too. In our case, we serve 600 global companies off a single code base, one code line. What took down most of our early competitors was allowing the code to branch off, and at some point you’re looking at a massive rewrite of the whole thing, and that’s problematic.

Lannert: To pivot successfully, you have to think, What’s at the fulcrum? What is your company uniquely great at? For us, it was an interface. We completely changed industries, but the interface was still at the core. A successful pivot is, at its root, about honoring what you uniquely do well as a company, as a team, as a brand, as a product, and staying true to that.