Terminations and compensation changes at Holacracy One

Ever since I read Brian Robertson’s book five years ago, one of my deepest qualms with Holacracy has been the gap it created around People practices such as performance management, compensation changes, and terminations.

Five years ago, I wrote: 

Dealing with “human spaces” problems such as hiring/firing, compensation, growth, etc. is not part of the Holacracy “operating systems” but mere “add-ons/apps” that each business should figure out on its own, once they adopt Holacracy. Yet at the same time, it’s been acknowledged that Holacracy is at odds with traditional “human spaces” solutions, and the friction between the two is a key cause for some companies attempting to adopt Holacracy but failing to successfully do so.

To use the “Holacracy as an Operating System” metaphor which Robertson introduced in his book: he created a new OS that’s incompatible with the existing apps (manager-led people practices) and launched it without an ecosystem of compatible apps. Only radical early adopters would switch to a new OS without an ecosystem of compatible apps already in place, which sadly, not only slowed down Holacracy’s adoption but also made it the target of very valid critique. 

Since then, I’ve learned to look at Holacracy less as a monolithic dogma and more as a source of inspiration for several potent building blocks for designing more human organizations. Top of which are the “OS Kernel”, the Holacracy Constitution, a compelling alternative to CEO/founder dictatorship for governing organizational power; and variations of the Integrative Decision Making process for more effective decision-making. 

While those are important pieces of the larger puzzle, it is still missing the key pieces of people practices and will not be complete without them.  

Robertson has softened up over the years too. When I heard him speak last year, he had a growing appreciation for the human aspect of the organization and for the need to cultivate not just “role” but also “sole”. 

In a series of blog posts published earlier this summer, he outlined how his Holacracy-run company, Holacracy One, handles two of the most challenging people practices in any organization —  terminations and compensation changes: 

Getting fired without a boss

A self-directed compensation system

The processes are also covered in detail in the company’s governance system, GlassFrog. 

Termination process (Partnership Review)

Roles: 

• Partner — the person whose affiliation with the organization is being reviewed. 
• Partnership Assessor — responsible for assessing Partners for continued partnership during scheduled Partnership Reviews.

Process: 

1. New Partners automatically go through a Partnership Review after three, six, and twelve months of tenure. Any Partner can request a Partnership Review for another Partner. 
2. All Partnership Assessors must participate in the review or waive their right to attend. They can also invite additional Partners to attend, as long as no other Partnership Assessor objects.
3. After sharing the context for the review (the reason it was called) participants vote (advocate for / neutral / against) on the following prompt: “knowing what I know now, if this partner weren’t already in the partnership, I would invite the reviewed partner into the partnership again now with their current relationship terms”. If the vote is not unanimous the Partnership Assessors explain their votes, discuss the matter and vote again. 
4. To remain in the company, a Partner must have more “advocate” than “against” votes and at least one “advocate” vote. 
5. If the decision led to a negative outcome, the Partner is notified and has three days to decide whether they want to accept the outcome or discuss it with the Partnership Assessors. If they opt for the latter, they have two weeks to have those discussions and decide whether to request a do-over. 
6. The outcome of a do-over Partnership Review is immediate and final. A negative outcome kicks-off the Partner offboarding process. 

Compensation change process (Compensation Review)

Roles: 

• Partner — the person requesting the compensation change. 
• Compensation Architect — responsible for designing, implementing, and evolving the organization’s overall systems and related processes for determining Partner compensation. As well as defining and publishing the organization’s possible compensation tiers, along with general criteria to guide others in assessing placements within it.
• Compensation Guardians– responsible for assessing proposals for compensation placements against other existing placements, and against the general guidance published by the Compensation Architect.

Process: 

1. Impact write up: The Partner submits a write-up of (a) the skills and other capacities with which they contribute most of their value to the organization, and (b) any constraints or weaknesses that limit the value those skills and capacities would otherwise be able to add. Part (a) can be written by someone else (after obtaining the Partner’s permission), but part (b) must be written by the Partner. All Compensation Guardians with relevant focus must confirm that the write-up is materially complete and accurate.
2. New compensation level proposal: The Partner submits a proposal for a new compensation level, using one of the allowed levels defined by the Compensation Architect. The proposal outlines arguments for and against the change, and an assertion that the author thinks that granting the new level is the right decision for the organization all things considered. The proposal can also be authored and submitted on behalf of a different Partner. 
3. Compensation Review: The proposer schedules a “Compensation Review” meeting with all relevant Compensation Guardians. Compensation Guardians can waive their right to attend, as well as invite additional Partners to attend, as long as no other Compensation Guardian objects. The Compensation Review follows the same structure as the Partnership Review (above) with a modified prompt: “If this person didn’t already work here, would you advocate for hiring them again at their proposed new compensation level?”. To be approved, a proposal needs more people advocating for it than against it, and at least one positive advocate.

Netflix’s No Rules Rules — Part 2

Every good plan changes when it meets reality, and my plan here is no different. While this was intended to be part 3 of the series, I decided to cover it now and leave the specific Netflix practices for later. 

Chapter 10 of No Rules Rules discusses the work that Netflix has done to adapt its unique culture to its multi-national footprint. Since Netflix started off in the US, many aspects of its culture aligned relatively well with general US culture. In places where intentional differences were introduced, the degree of difference was small enough that behavior change and convergence could be driven through standard means: role modeling, training, education, and processes/incentive systems. However, as Netflix expanded internationally, some aspects of its culture were at more extreme odds with local cultures. For example, its high candor culture, which celebrates direct negative feedback, was at odds with Japanese culture, which favors a much more indirect approach to negative feedback. 

The need to more deliberately explore and reconcile those cultural differences led to the initial collaboration between Reed Hastings and Erin Meyer, the book’s co-authors. Meyer is a leading researcher in the field of cultural differences, and her capstone research is covered in detail in her 2014 book The Culture Map. A more digestible summary of it is available in an HBR article titled: 

Navigating the Cultural Minefield

Meyer’s research “stands on the shoulders of giants,” synthesizing and integrating previous work by leading researchers such as Edward Hall, Geert Hofstede, Robert House, Mansour Javidan, Roy Chua, Michael Morris, and Richard Nisbett.

Meyer created a framework for analyzing and comparing cultures across 8 scales: 

1. Communicating (low context <> high context)
2. Evaluating (direct negative feedback <> indirect negative feedback)
3. Leading (egalitarian <> hierarchical)
4. Deciding (consensual <> top-down)
5. Trusting (task-based <> relationship-based)
6. Disagreeing (confrontational <> avoids confrontation)
7. Scheduling (linear time <> flexible time)
8. Persuading (principles first <> application first)

Communicating —  This scale measures the degree to which cultures are high- or low-context. In low-context cultures, good communication is precise, simple, explicit, and clear. Messages are expressed and understood at face value. Repetition and putting messages in writing are appreciated to clarify communication. In high-context cultures, communication is sophisticated, nuanced, and layered. Messages are both spoken and read between the lines. Less is put in writing, and more is left for verbal interpretation.

Evaluating —  All cultures believe that criticism should be given constructively, but the definition of “constructive” varies greatly. This scale measures a preference for frank versus diplomatic criticism. While often confused with communicating, many countries have different positions on the two scales.

Leading — This scale measures the degree of respect and deference shown to authority figures, placing people on a spectrum between the egalitarian and the hierarchical. 

Deciding —  This scale explores the differences between building group agreement and relying on an individual(usually the boss) to make decisions. We tend to assume that it’ll be strongly correlated with the “leading” scale, but this is not the case in all cultures. 

Trusting —  Cognitive trust (from the head) can be contrasted with affective trust (from the heart). In task-based cultures, trust is built cognitively through work. If we collaborate well, prove ourselves reliable, and respect one another’s contributions, we come to feel mutual trust. In a relationship-based society, trust is a result of weaving a strong affective connection. If we spend time laughing and relaxing together, getting to know one another on a personal level, and feeling a mutual liking, we establish trust.

Disagreeing — This scale measures how cultures view open disagreement — whether they feel it is likely to improve group dynamics or negatively impact team relationships. Some counties view the public airing of disagreement very dimly, while others are quite comfortable having spirited, confrontational meetings.

Scheduling — This scale measures how much value cultures place on being structured versus reactive. In some cultures, people treat the schedule as a suggestion, while others stick to the agenda. 

Persuading —  This scale measures a preference for inductive versus deductive arguments. Some cultures tend to find deductive reasoning more persuasive, whereas other cultures typically prefer an inductive style.

---

Netflix used Meyer’s research to identify the countries where the Netflix culture is most at odds with local cultures to:

•  Support team members in those countries better, knowing that they’ll likely take a little longer to adapt.
• Look at elements of its own culture that are most at odds and mitigate some of those tensions, such as increasing formal feedback in cultures that are less direct on the “evaluating” scale. 

On Meyer’s website, individuals and groups can take Meyer’s 24-question assessment, see where they fall on each of the eight scales, and compare their culture to the cultures of the countries in which they’re doing business. 

In my mind, the intra-cultural comparisons are more interesting than the inter-cultural ones. As my assessment suggests, while on some scales, my disposition was a blend of the two cultures of the countries I’ve spent the most time in (US and Israel), on others, I was more extreme than both in one direction or the others. Suggesting, perhaps obviously, that country culture is not the only contributing factor to my culture, and the two can differ quite significantly.  

Therefore, I see the tool’s key value as a fascinating diagnostic for organizations in codifying their organizational cultures across the eight scales, by aggregating all team members’ individual results. And it’s not just the medians that matter; it’s also the variabilities. Understanding on which scales there’s strong cultural coherence (low variability) and on which scales there’s weak cultural coherence (high variability) can help an organization focus its efforts on the latter if it wants to create a more coherent culture.