Sun Feb 26 2023

Why does OKRs fail?

As I read more about OKRs, I have started to wonder how so many companies that saw great success from adopting OKRs later suffered serious failures or structural issues. This post is thoughts and musings about this problem.

How do OKRs go wrong?

I have been reading John Doerr’s Measure What Matters, which I gather is an extremely influential book in describing how the OKRs approach to goal setting works and how it helps teams accomplish better outcomes. John Doerr cites and praises the success of Google, Intel, AOL, and other companies that have successfully adopted OKRs. While these companies did achieve great growth and market share over the time period the book covers, over the last two or three years many of these same companies have struggled significantly despite having made OKRs a core part of their company’s culture.

What happened to these companies? How did they go from highly focused, highly disciplined producers to near-static, highly risk-averse rent-seekers?

Ask Google about how their OKRs failed them

In 2023, Praveen Seshadri, co-founder of productivity software AppSheet, posted an essay on Medium in which he strongly criticized Google’s culture. In this essay, he highlighted four main problems facing Google’s culture:

No mission
No urgency
Delusions of exceptionalism
Mismanagement

He also talks at length about how Google’s culture has become consumed with process, control, and minimizing risk to the point that nothing of any consequence gets done if anything gets done at all.

In similar fashion, in 2021, Noam Bardin, co-founder of mapping and GPS application Waze, wrote his own essay about his experience working in Google for 7 years after Google acquired Waze. He hightlighted many of the same struggles as Pradeep Seshadri (who actually cites Noam Bardin’s essay as another example of what he experienced). Noam Bardin talks about several issues he found with Google’s culture. Those issues include but were not limited to a too heavy emphasis on corporate policy over helping users, an incentive structure that rewards non-user-benefitting work, and too strong of a focus inward on what benefits Google and individual Googlers.

How did Google come to this point if they previously had such great clarity, focus, and drive? Where were their OKRs to save them? It is possible that they lost the people who made the OKRs framework function. John Doerr credits Larry Page, Sergey Brin, Eric Schmidt, and himself as critical parts of implementing and succeeding with OKRs in Measuring What Matters. These four men may have been the heart of OKRs. More broadly, Pradeep Seshadri in his essay points out how after three years he had been at Google for longer than half of all of Google’s employees; he highlights this as a sign that Google has probably not been able to maintain a coherent positive company culture. I agree with Pradeep Seshadri that such churn, growth, and general change in the company’s makeup would also remake the company’s culture. I suspect many of those people do not have a good handle on OKRs or how to apply them successfully since they are so far removed from the generation of Google employees who implemented and succeeded with OKRs.

Google may also have grown complacent and content thanks to its overwhelming monopoly in search advertising. Both Pradeep Seshadri and Noam Bardin discuss observing feelings of entitlement or unreasonably high self-esteem among Google employees. I think it is very well possible that Google has stopped asking how to be better because they think they already are the best. This feeling could significantly undermine the use of OKRs because one of the core components of the OKRs framework is frequently asking yourself “how can I do better?“.

I also agree with Pradeep Seshadri’s assertion about how the dominant success of search advertising strongly disincentivizes Google from doing anything that could potentially threaten that success. This also to me feels counter to OKRs, which include a high amount of risk-taking in order to push you to achieve and to learn more.

Outside of these two essays, Google has more widely shown a distinct inability to focus or to deliver lasting positive outcomes for users. One merely has to stroll through the Killed By Google cemetery to find hundreds of services that Google has developed and then killed in its lifetime, regardless of the success or value of those services.

I think Google has stopped being the kind of high performing culture that would benefit from OKRs. I agree with Pradeep Seshadri that Google lacks any kind of mission worth really working towards. In addition, based on the experiences Pradeep Seshadri and Noam Bartin share, it seems to me that Google no longer truly rewards or empower individuals to make a difference in the lives of their users.

Ask Intel about how their OKRs failed them

John Doerr, in Measure What Matters, talks about how OKRs were pretty much invented at Intel by Andy Grove during Andy Grove’s time as President and then CEO of Intel. According to John Doerr, Andy Grove successfully navigated a massive pivot in Intel’s core business from memory to microprocessors and laid the foundation for Intel’s dominance of the microprocessor market.

However, in 2022, according to a CNBC article about AMD and Intel, AMD surpassed Intel by market capitalization for the first time ever. I grant market capitalization is likely an unreliable measure for comparing companies, but AMD has been in Intel’s shadow for decades. Similarly, in 2015 NVIDIA became the most valuable chip company in the United States instead of Intel.

I have been reading about Intel’s troubles over the last few years. In 2021, tech news outlet Ars Technica cross-posted a Financial Times article about Intel ousting their CEO because of their troubles. Some of the biggest setbacks were Intel suffering multiple delays for its new generations of chips. It also seems that during this time Intel’s primary competitors Samsung and TSMC overtook Intel in chip manufacturing.

How did Intel go from being a highly-focused, ambitious organization with dominant market position to being overtaken by new and old comptetitors across its business? As I reread the stories about Intel from the 2010s to 2020s, it feels like Intel also becamse complacent. They achieved their dominant market position and stopped feeling the same kind of urgency that drove Intel to abandon its memory business and commit 100% to its microprocessor and peripherals businesses.

I also think the culture changed and lost its way after Andy Grove stepped back from his involvement in Intel by 2004. John Doerr credits him as a critical driver of the OKRs-style of management, and replacing Andy Grove with someone else would change the company culture. In addition, I sometimes wonder if part of the problem was placing a person from a finance background in charge of the company. After having finance people instead of engineering people as CEO, Intel and Boeing both seemed to have suffered losses in their core competencies that translated into terrible outcomes for their businesses. In fact, a brief 2021 post on the Reuters website reports that Intel CEO Pat Gelsinger, who is an engineer and who replaced the finance CEO, said Intel will not be focused anywhere near as much on stock buybacks as it previously was. According to a profile of Intel by investment analysis site Seeking Alpha, Intel spent $44.6 billion in stock buybacks from 2017 to 2021, much of which lines up with the previous CEO’s tenure. I mention this because it seems to me that CEOs from a finance background focus on artifically juicing the company’s stock value through stock buybacks, cost cutting, and focusing on short-term gain instead of investing in the company’s core competencies, opening up new markets, and working towards sustainable long-term growth.

While I cannot cite first-person accounts of the company’s culture, I wonder if the changes in the CEO and executive leadership of Intel undermined the foundations that made OKRs so successful for Intel during Andy Grove’s time with Intel. Measure What Matters talks at length — I believe righty so — how executives must embody the values and behaviors they want their team members to have. If your executive leadership no longer values engineering, mastering your craft, and creating truly imaginative, revolutionary technology the company will eventually stop valuing them as well.

In this case, I feel that Intel also lost the culture that made OKRs into an accelerator. They stopped being the kind of high-performing organization that would benefit from the focus that OKRs offer.

OKRs are for successful teams

I grant that Google and Intel are only two examples of companies that adopted OKRs, were very successful, and then were less successful. There are other examples like this, such as AOL and Yahoo, who were very successful but have gradually disappeared.

It seems to me, however, that the teams that saw great success from adopting OKRs were already likely to be successful to some degree, and OKRs acted as an accelerator or multipler for their success. The teams were already recording wins even if they were threatened or needed to pivot their businesses.

From Intel and Google’s cases, I believe they lost their cultures that primed them to be successful from OKRs. They also lost the key people who were responsible for developing and carrying out their implementations of OKRs. As their culture degraded, OKRs lost their accelerating power and became another way to micromanage team members’ work. In addition, the companies lost their guiding missions or overall vision to which all OKRs could be aligned.

Overall, I think the most important part of OKRs are not the framework itself but the psychological safety, the empowerment, and the foundation of success on which OKRs are layered.