Farming tacit knowledge in a remote-first, asynchronous setup

Summary
Team and company handbooks are an efficient way to document your explicit knowledge. However, most of your organisation's knowledge about emerging practices and technologies is tacit. Leaders need to invest in systems that provide the capabilities to farm this tacit knowledge.

1. Your knowledge stack should take inspiration from the consumer internet. Think freeform, frictionless and emergent.
2. Point to point collaboration tools make teams efficient, but they also lead to locally optimised silos.
3. Over and above team collaboration tools, invest in systems that offer most of the following capabilities.
• Serendipitous discovery.
• AI assistance.
• Reputation patterns.
• Means for expression.
• Structure, after the fact.

Remote work would have renewed your organisation’s interest in knowledge management. Considering one can’t walk up to co-workers for a quick clarification, could we instead ask the system for an answer? In an asynchronous, remote-first culture, a solid knowledge strategy can be a productivity power up. The converse is also true. Without a proper knowledge strategy, your people will frustrate themselves, searching for answers to the problems they face at work.

Depending on where you sit in the organisation, you may or may not have an influence on the way you surface and share knowledge in the company. Understanding the levers for knowledge sharing may, however, help you make an effective case for influential leaders. And while the organisation takes its time to make change, you can use these levers to build a locally optimised knowledge strategy for the people you lead. 

We’ve discussed the value of creating a team handbook. GitLab and other organisations have public handbooks to run their company! These are the kinds of knowledge that I refer to as “stocks”. With complex knowledge work, the approach of creating stocks of well structured organisational knowledge has its limitations. 

We often solve problems where there is no established practice to follow. I argue that much of software development is just that. We start with a hypothesis and then construct experiments to validate that hypothesis. The result of the experiment tells us what to do next. In a fast changing domain such as this, a majority of knowledge remains tacit. While techniques such as lean startup and design thinking codify the process of creative discovery, learning remains rather chaotic per domain and per emerging technology stack. Organisations try their best to make knowledge explicit, but frankly patterns emerge faster than we can document them.

In the 2020s, knowledge can take many forms. Some of these are traditional; others are edgy. Traditional forms of knowledge - process docs, templates, toolkits, learning materials - are ideal for stable practices and information. Modern forms of knowledge - questions and answers, videos, podcasts - work best for groundbreaking, emerging practices. A lot of know-how about such practices is tacit. 

Lew Platt, former CEO, HP

“If only HP knew what HP knows, we would be three times more productive.”

So how should your organisation farm all this tacit knowledge? In this article and a few subsequent ones, I want to share my thoughts about creating a knowledge ecosystem that keeps pace with your people’s know-how. I argue that we have much to learn from the consumer internet. Allow me to explain.

Take inspiration from old wisdom

About a decade back, Andrew McAfee was capturing the imagination of big corporations with an insinuation of what their collaboration systems should look like. In his landmark book Enterprise 2.0, McAfee advocated for the benefits of adopting consumer internet patterns as part of collaboration in the enterprise. 10 years down the road, his advice has indeed come alive. Look around you. Workspace by Google, Basecamp, Slack, Confluence, Jive, LumApps, MS Teams, Workplace by Facebook, HCL connections, Degreed, Notion - tools for the enterprise are the same as the ones we use in our personal lives or very similar to them. 

Characteristics of modern enterprise systems

Having said this, I’d like to go back to some characteristics of modern enterprise systems that McAfee outlined back in the day.

1. Freeform. The system doesn’t impose an organisational structure or a structure for the content one contributes to the collective. “People come together as equals within the environment created by technology, and do pretty much whatever they want.”, said McAfee.

2. Frictionless. Participation on such a platform is easy and the barriers to entry are as close to zero as possible. 

3. Emergent. “Patterns and structure appear” in the system in time. Most of this structure comes from the users through the use of bottom up mechanisms such as tags, reviews, stars, likes and lightweight interaction mechanisms. The cream rises to the top, with little intervention. This last, but probably the defining parameter of these systems stuck out to me as the most powerful as well.

Point to point solutions create walled gardens

So did McAfee’s vision come true? It must have, right? Aren’t all those solutions I named previously, exactly what he was advocating for? Well yes, and no. Several tools I mentioned make collaboration free-form, frictionless and probably even emergent, but they do so “point to point”. Let’s take the example of Workspace by Google - Google’s collaboration platform. It’s all about team collaboration - send someone a message; chat with the team; start a video call; share a document with your team; create a site to catalogue your work; store all your team documents in a shared folder. It’s beautiful, but it focuses on making you and your team more productive. It does little to generate signals for other people who may be peripherally interested in the work you do, so they can learn from your experience.

You’d argue that you could share things you want to share in a large, community mailing list. Sure, but that would only get your message to the people who already know they want that message. What about those who need to discover it without being part of a mailing list? What about those who are interested in your work, but not that of your team or your community? Should they have to eavesdrop on all the Google groups you’re part of, in the secret hope that someday you’ll share that invaluable piece of knowledge? And drink from the firehose of emails from everyone else, in the meantime?

By the way, this isn't about Workspace by Google’s limitations. To be fair, many platforms behave exactly the same way. They focus on making certain groups of people more effective when working with each other. They don’t really help all these groups come together as one organisation that builds off each other’s tacit knowledge.

So what capabilities do you need?

We need all the collaborative features that the popular suites offer us. To farm tacit knowledge, though, these features need to align with few key capabilities. The more capabilities your toolset provides, the more robust your knowledge strategy will be.

5 key capabilities your knowledge ecosystem needs

1. Serendipitous discovery. Outside work, how do you discover useful information? You don’t sign up to lots of mailing lists, do you? Between Siri, Google assistant, your favourite news apps, and your social networks, you just stumble on the information, don’t you? Often you find actionable advice from sources who aren’t your direct contacts. How can we make the enterprise mimic this behaviour?

2. AI assistance. Speaking of Google assistant, Facebook or any other platform, the system seems to know you and your preferences. When you say that you need to see certain information, it remembers. It recognises that my partner loves Bollywood and that I geek out on photography, wildlife and politics. It’s creepy and, much as I hate to admit it, it’s also effective. We need to discover ethical, privacy respecting ways to achieve a similar outcome in the company ecosystem. How can your knowledge systems understand your people’s interests and help them discover the content they care for? 

3. Reputation patterns: If you have an interest or hobby, the internet makes it painless to know about the people who have a reputation in that space. Political activists, photographers, travel writers, home cooks - you can find influencers in just about any space. Yes, some of this is media manipulation, but a lot of it is based on the strength of these people’s contributions to the collective that is the internet - a.k.a their digital exhaust? How can our enterprise platforms also surface the experts on the topics that we’re interested in? At a more basic level, how do we preserve people’s digital exhaust?

4. Means for “expression”: The good and bad of the internet is that it provides us a soap box to express ourselves. You can post photos on Instagram, videos on YouTube or have your own website and do whatever you want there. Now sure, you invite yourself to commentary - but outside of hate speech, that’s mostly a good feedback loop. People who “follow you”, have a way to know what you’re thinking or see the content you’re producing. You can “post in peace”, knowing well that you’re not creating spam for anyone. People can choose to see your content. Or not. Their choice. How do we create these friction free mechanisms for people to air their ideas without spamming many people?

5. Structure, after the fact: The beauty of the internet is that structure continues to emerge after the fact. A bunch of good samaritans keep Wikipedia so current that it’s endangering the entire encyclopaedia business. People maintain their own structures too, through tools like Pinterest, Diigo, Evernote, etc. How does the enterprise provide “sense making”? Once you have a large volume of free-form content that people have produced, how can everyone navigate this knowledge ecosystem easily?

---

Those five provocations bring me to the end of part 1 of this series. In part 2, I’ll build on these ideas to share an alternative to the “stocks” based approach to knowledge management. Until then, I’d love to hear if there are other capabilities of the consumer internet you think we should mimic in the enterprise.