Structural transparency vs content-level transparency
5 min read

Structural transparency vs content-level transparency

Many companies I work with consider transparency a virtue. But when I started to really dig in, it became very clear very quickly that "transparency" is one of those words Molly Graham would call a black hole word– (paraphrasing here) words that everyone uses freely but lack a concrete definition. And I find that transparency is a particularly tricky virtue because it's almost impossible to disprove because no one knows what they don't know.

Let's say a company preparing for a board meeting.

  • Who knows the meeting is happening: how much do they know about the meeting? timing? attendees? content? do they learn about it before the meeting happens or after it's over?
  • Who has write access to the prep documents: Is it just the board deck working group? Are subject-matter experts invited to contribute directly on the document or do members of the working group translate the information? Do they get access to sections or the entire thing?
  • Who gets reads access (and when): Do you share it with leaders around the organization? The finance team? Everyone? And when are those invitations extended? Before or after the meeting?

What most people are talking about when they talk about ‘transparency’ is what I would call structural transparency. This refers to what information is shared freely and openly across the organization and with whom.

But, another critical part of transparency norms comes down to what I would call content-level transparency, insight into the raw work as it’s happening.

It's also important to acknowledge that there is such a thing as good or bad transparency. Good transparency answers questions before they come up, brings people along, and encourages consistent and principled decision-making across the organization. Bad transparency distracts people, invites the wrong questions, treats information as currency, embraces gossip as truth, and results in politicking. In the board prep example, you could imagine that some information could be sensitive (ex- personnel updates) or distracting (ex-discussion of upcoming layoffs).

By setting the right norms, and consistently applying them through the right systems, the company can stay on the good side of transparency. And, not just the leaders.

The case for shifting towards more content-level transparency

That is, creating systems and norms that offer more insight into the raw work, as it’s happening.  

I'll start off by saying that my thinking on this topic was heavily influenced by Stripe’s email transparency system. I don’t know a single ex-Stripe that has not implemented (or tried desperately to implement) this at their next company. It is an incredibly powerful force for both efficiency and culture and I think more companies should adopt a version of it.

At most companies, if A and B are working on x, no one else would know unless A or B tells them and no one else would have access to details unless A or B invites them. Every company has their own norms for who to extend these invitations to and when. Or, conversely, what filters to apply on information. The first filters that usually come to mind are the tops-down filters; what information leaders decide not to share with their employees. But, the filters that have a much greater impact on the overall transparency of the organization are the ones that employees set up for each other. There are the pernicious filters, which primarily come in the form of hoarding information with the intention of maintaining power (very bad), and the benevolent-intentioned filters, which mostly come in the form of holding information back for fear it will upset or distract their team (less bad, but still not optimal). The norms around this dictate a company’s structural transparency.

The challenge with this system is that, even in the most benevolent environments, the filters are always going to be a little off because they require an individual to make a judgment call about every atom of work. Plus, the information funnel is a leaky, unreliable one, especially when it relies on word-of-mouth. Ideas can get taken out of context, delivered at the wrong time, misinterpreted, or misconstrued, even in the most benevolent environments. And when this inevitably happens, a teeny little crack in the foundation of trust gets created. And once a crack exists, the laws of physics dictate that it’s probably only going to get deeper.

No matter what those norms are, and no matter how open a company aspires to be, colleagues are required to put an **enormous** amount of trust in each other that the right information is made available to them at the right time. They must believe that the correct information either comes to them when it needs to, or that they will know enough to know what to go asking for, and then get it by request.

The innovation of Stripe’s email transparency system was to bolster the structural transparency norms with content-level transparency.

At Stripe, all of A and B’s work on X was open for the entire company by default. A and B might still extend proactive invitations to others to join in on or take a look at the work in depth, but C, M, W, and everyone in between could also pop in and check out what was going on for themselves at any time.

Perhaps counter-intuitively content-level transparency, doesn’t require trust by default because anyone can see the primary sources of information. As a result a much more genuine trust emerges because they can come to conclusions on their own. And, with filters relying less on human judgements, there is a lower likelihood of those crack-creating misinterpretations (and, people are on better, more consistent, behavior when sharing with a broader audience).

This system trades the broken filter dragon for the information firehose dragon. I’d take that trade any day (there are simple tactics one can implement to help). I know it also sounds hard and scary to think about making all of the information produced at your company available to anyone at any time (I'll share more about this soon).

There are very few things that you can do at a company that will reliably result in more honesty across your organization and deeper trust amongst colleagues. Shifting your organization towards more content-level transparency is one of them.

A non-exhaustive list of benefits of a content-level transparent system

  • Everyone can spend less time routing information when the goings-on of the company are available, searchable, and referenceable.
  • ~Primary sources are available on all work at all levels and help anyone go deep on what’s actually happening, not just what people say is happening. (Look for the story about the time all of Google's Eng Leaders were fired because their reports of what was going on didn't match what the IC's were reporting in their weekly snippets).
  • Collaboration time can be reserved for the more interesting (and satisfying work of) exchanging ideas and perspectives.
  • No one has to reinvent work that’s already done. Instead, they can focus on moving the entire system forward.
  • Publishing to a broad audience keeps everyone more honest.
  • When everyone relies on the same, true, information, your organization is naturally less political.
  • The potential for anyone at the company to read what you published is a forcing function for thinking more globally.
  • Everyone will feel more connected to each other because they have the opportunity to see their colleagues in action and see the context leading up to the work.
  • Read access is the same for everyone regardless of your function, level, or tenure. High trust yields high trust. (And, trust is more easily validated through primary sources).
  • You avoid the inefficiencies (and the accompanying hierarchical bullshit associated with) "managers talk to managers who talk to IC's."
  • There is a high-upside opportunity here to trade “I’m-taking-a-quick-break-and-checking-TikTok” time with “huh-this-company-document-looks-interesting-I-wonder-what-it’s-all-about” reading time.