All over China, leaders of fast-growing companies are asking themselves: how can we grow as fast as we need to without destroying our company? It’s a question Verne Harnish has been answering for years. Harnish is the CEO of Gazelles, a Virginia-headquartered executive education and coaching firm focused on advising fast-growing companies that hosts seminars and events for business leaders all over the world. He is author of Mastering the Rockefeller Habits: What You Must Do to Increase the Value of Your Growing Firm, a book that suggests using John D. Rockefeller’s foundation of Standard Oil as a guide to growing a company. A second book, The Greatest Business Decisions of All Time, co-written with the editors of Fortune, was published recently.

In this interview, Harnish shares his thoughts on managing fast growth.

Q. Why do you see John D. Rockefeller, the founder of Standard Oil, as a good example for our time? After all, wasn’t that more than 100 years ago?

A. He managed to emerge from a very crazy and tumultuous new industry, not unlike what we saw with the dotcoms. There were a thousand refineries that popped up and died the way dotcoms in the late 90s, and it was kind of the second wave that he caught, where a handful of companies emerged as the real players, as again we’ve seen here this past decade. So, his era pretty well mirrors the era that we’re in right now with these new technologies, and I thought there would be some great lessons for us to learn from him.

Q. Of the lessons that you draw from his example in building Standard Oil, which one do you think companies could profit from most?

A. The thing that enamored me the most with John D.–and he talks about it in the book Titan–he thinks that the very heart of his company’s success in this crazy, fast-growing industry is he had structured a daily huddle with his directors.

When they were in Cleveland, he and the four other guys he launched the company with would walk to work and walk home together each day, and he realized that it was on these walks that he really was able to get things figured out. So when they moved to New York City, he made sure that the leadership lived near enough to each other and close enough to Standard headquarters, they could still do these walks and talks, and he structured a daily lunch with his nine directors.

Let’s speed forward a hundred years later: right now, the highest market cap, most valuable company on the planet today is Apple. One of the keys that Walter Isaacson talks about in Steve [Jobs]’s biography and more importantly Adam Lashinsky in Inside Apple, is whenever Steve needed to figure something out, he’d go on these walks and talks. And, in parallel to Rockefeller, Steve had lunch almost every day with Jonathan Ive (Apple’s lead designer).

Q. Is there a particular size that works best for that huddle?

A. The strategy part of strategic planning is really a very, very small team. We recommend it’s two, three (or) four folks who have breakfast or lunch two or three times a week, but at least once. The idea is to really get some talk time around the big strategy questions.

Now, you don’t do this in isolation. Those three or four need to be deeply immersed in the industry and talking with employees and interacting with customers–they talk about how Steve Jobs didn’t like focus groups and I don’t either, but he spent an inordinate amount of time interacting directly with customers. He’d call them up and deal with their issues and he stayed very much connected to the real world, which is important because you can quickly get isolated once you’ve got a few employees between you and reality.

For execution planning, the more people you can get on board the better. We like to see at minimum, everybody in management. A typical sized company we work with his 400 employees, which means there will be 40 or 50 manager/supervisors on up to senior executives. And in fact, the work that Bill Isaacs has done out of MIT with dialogue is that a real dialogue requires 40-60 people so that you’ve got enough diversity of thought in the room to really nail down what has to get accomplished.

Q. Stepping back now, how do you find the right people for those teams in the first place?

A. One of the keys is to recruit for cultural fit versus skill. That is the fundamental mistake most companies make. They hire somebody based on what looks like a match between the things they’ve been doing on their resume and what the company’s looking for, and they get them in, they find out there’s no fit to the culture, and that’s when everything blows up. If you get somebody who’s sharp and motivated, they’ll acquire the skills.

The second real weakness, at least in growth companies: in a lot of companies, HR and the managers required to do the hiring don’t have a marketing bone in their body, and you really need an effective marketing function to recruit talent.

You need roughly 40-plus applicants for every position. It’s almost like this dialogue number: you need 40-60 to get enough diversity of applicants to be sure that you’re going to pick the right one. Most growth companies, because they’re short of time and they need somebody today, they’ll pick up anybody who passes the breath test–can he fog the mirror? Man, you make the wrong decision, you’re going to pay dearly. Three times annual salary is the standard cost of making a bad hire.

We see three steps to hiring. First, using guerilla marketing techniques to attract a large pool of candidates. Atlassian did this innovative bus tour of Europe… they got a thousand folks interested, from which they were able to choose the 15 that they were looking for.

Once you get that large pool of applicants, we are avid fans of the Topgrading methodology of putting together the scorecard and then doing what’s called the pre-interview and the chronological in-depth structured (CIDS) interview (an in-depth interview that reviews key moments of the candidate’s entire work career), a methodology that delivers a 90% success rate in choosing the right person if done correctly. If somebody does the feel-good interview, you’re right about the third of the time, so you get a lot of turnover. If you can do any kind of testing, that’ll increase your odds by 50%. The standard behavior-based interview process gets you to about 60%, but you’re still going to mess up 40% of the time. As there’s just no reason to mess around, we teach all of our clients this Topgrading methodology.

Once you’ve got the right person selected, then what you have to do is retain them. I don’t think enough companies focus on retention. The key to retention is something Jim Collins talked about: is management focusing on the right question, which is not, what do we have to do to motivate people? If you’ve hired right, you’ve already got motivated folks. The thing you’ve got to focus on is the question, what do we have to stop doing that’s demotivating everybody? We have a tendency to bring talent into our company and then all of a sudden they hit a stupid policy and you turn A players into B and C players. You can take a highly motivated person and just suck every ounce of energy right out of them.

Q. Is it hard for a company to keep the culture consistent as it scales up?

A. It is, I think, one of the sole reasons why–at least in the United States–only 4% of companies get to a million in revenue, which is about 10 employees, and only 4/10s of 1% make it to about 50 employees. And then the air gets even rarer above that.

The real sticky point, the point where you start having this cultural breakdown, is about 60-70 employees. You know you’ve hit it because it’s the day you walk in and there’s this person working in the company and you’re like, ‘Who’s that?’ And they’re like, ‘Well, that’s our new vice-president or assistant of or supervisor for…’, and you’re like, ‘Whoa! Man, I don’t know everybody any more’.

That’s when the culture starts to leak. That’s when you start to have to formally focus on passing your DNA as a leadership team down to this middle management that’s now between you and the rest of these people.

Q. What kind of processes work for passing that DNA along?

A. One is a well-structured monthly management meeting where you really engage the middle managers in solving problems together with the senior team, because it’s really in the process of doing real work together that you can observe and see who’s getting it, who’s not getting it. You can lead by example, you can teach, they can see how you respond. As we say in parenting, more is caught than taught, and so these aren’t one-day workshops, these are one-day working sessions.

The second thing that’s critical–because the real keepers of the culture are your core values and your core purpose, and you need to align all of your HR systems around this one list.

HR is usually the last function to really develop and it develops very randomly. One day somebody says, you know, I think we ought to think of some awards to give out at our holiday party, and so you come up with a list. Then one day somebody says, man, we really need to get some more formal recruiting guidelines, and so somebody comes up with a list, and then one day somebody says, well, we have to have this employee handbook, and so somebody pulls one off the Internet and changes a few words and says, now we have our employee handbook.

We’re adamant you take your one list of core values and those become the headings of your handbook, they become the names of the awards that you give out, and as we just already discussed, you want to recruit folks who match your core values and interview for those. Also, every time you praise someone, you tie it back to one of your core values: I’m really excited to see this, this is exactly what we mean by first class for less or having ecstatic customers or honoring intellectual capital. Or, I’m very, very upset because this is the opposite of what we mean by…. If you always tie back your carrots and sticks to one of those core values, then folks are going to know them.

There’s a dozen HR activities and systems, and if they all align around one list, people don’t have to carry plastic cards. Everyone’s going to know it.