Standardizing procedures when business scales
By Tyler Weitzman
It amazed me to hear that the executive team at Google — Larry Page, Sergey Brin, and Eric Schmidt — continued reading and deciding on every single resume and hiring application at Google for such a long time. At that size, there are not only so many more hiring applications but also so many other things that require attention. Moreover, recruitment continues to be among the most significant tasks that need to be done right — can three people do “the right job” when there are so many folders to go through? What’s the secret here? Probably the same secret that allows about two dozen admission officers at Stanford to cover 40,000 yearly applications — standardization through data and rules.
Google followed very strict, simple rules for hiring: (1) Don’t hire your friends (2) Look for smart candidates with high GPAs from strong universities (3) Look for people who have done something exceptional (related or unrelated to work). The idea reminds me of GSB Professor Kathleen Eisenhardt’s research and book — “Simple Rules: How to Thrive in a Complex World.” The standardized procedure of adhering to a few simple rules guarantees doing “the right job.” In her book, Eisenhardt mentions the idea that often times the simple rules can do such a great job because they proxy more complicated information. This phenomena is most certainly the case here; for example, Eric Schmidt pointed out in Stanford’s CS183C Q&A that rule #3 tends to bring people who are more interesting and exceptional and who bring with them strong energy and culture.
In addition to adopting simple rules to standardize procedures for large volumes, Google also adopted quantitative analytics for every part of the company: Interviews, facilities, sales, etc. Eventually, Google has even optimized for the fact that 4–5 job interviews are optimal before diminishing returns kick-in.
The tribal stage likely offers the best time to transition all the processes of the company over to standardized procedures and rules with quantitative analytics. This stage offers the best time because the company is not so small anymore that standardization is frivolous, and not so big yet that standardization requires huge resources. Moreover, the quantitative analytics very likely holds data that can prove useful in the future. Save everything.
Perhaps most interesting to me though is the standardization of culture and vision. These two components are hard enough to define and enforce in the family stage, so how can they be enforced at the tribal stage? This question is critical because if it is not addressed then as the company continues growing in size culture could be degrading and people could be misaligning, leading to devastating results.
I found the most satisfying answer to this question from John Lilly’s experience working as CEO of Mozilla. His learned solution that he mentioned in class is to repeat the same vision and goals over and over to everybody, with the key lesson being that it is necessary to continue repeating the same thing over and over (even when you no longer believe it!).
In particular, Lilly commented on the ever-changing subtleties in culture, vision, and goals, explaining that only once goals have changed significantly enough would he announce, “I am now changing what I’ve been saying for the last four months.”
This solution works brilliantly off the fact that people need continuous reinforcement and repetition of the same idea to align with it. Using this method to standardize culture, and using quantitative analytics and simple rules to standardize procedures, a company in its tribal stage is set up well to scale its product market fit far and wide.