I’ve been thinking a lot about corporate knowledge management systems recently.
If we think about a company as an organism, then a knowledge management system is essentially the (collective) brain that keeps that organism alive and running. A corporate knowledge management system should contain every single bit of codifiable information within the company resulting in a library of all projects, processes and procedures.
In an ideal state, it is the single source of truth that helps to inform every individual in the firm about what everyone else is up to. Information should be easy to add (input) as well as easy to search and find (output) resulting in quick knowledge transfer between different employees.
In reality, however, this hardly ever is the case. As anyone who has ever worked at a larger company can attest to, company knowledge bases always end up being a huge mess.
What starts with a neatly organized Confluence wiki, over time morphs into a multi-headed monster consisting of millions of notes and documents that live across Google Docs, Dropbox Paper, Asana and half a dozen different wiki tools. Most docs will be outdated, some will contradict others and the one you are really looking for only shows up on page 14 of your search results.
It seems like things usually start to fall apart once a company surpasses the Dunbar number of 150 employees. This is probably when people start to realize that all the different documents of explicit knowledge they were amassing over the years have been held together with implicit knowledge.
It’s easy to find – and understand – the right documents when you know every other person in the company, but once you’re past that point, you need a system to organize all the data so that people can make sense of it.
The idea behind tools like Notion is to solve this problem by using just one tool for all your different knowledge documents. Instead of Google Docs AND Asana AND Trello AND Airtable, you just do everything in Notion. This reduces complexity because you don’t have to switch and search across different apps. At the same time, Notion forces you to think about a system that makes information easy to find with its folder-like structure and links between different databases.
I’ve never used Notion with more than half a dozen people myself, but from what I’ve heard from people at larger companies, Notion knowledge bases also don’t scale very well beyond a certain number of users. Once too many people start contributing to it, things become bloated and unnavigable.
A friend at Stripe recently suggested – half-jokingly – that we should hire a librarian to organize all our internal data and documentation. The more I think about it, the more I like the idea. Perhaps every company should hire a Chief Notion Officer once it hits 100 employees?!
An alternative approach to Notion is a knowledge management system that can live across different tools and without active manual curation because it’s based on really powerful search. The folder structure of your Google Drive, for example, doesn’t really matter because looking up documents via search is faster and more convenient. Meta search tools like FYI are supposed to offer the same but across different productivity tools.
But again I’m skeptical that this really works beyond a certain amount of users (and thus documents). I remember even Google’s internal search engine doing only a mediocre job of surfacing the most relevant documents (and even if it did you weren’t sure if there wasn’t a better or more up-to-date version of it).
I’m sure we’ll get there eventually, but until then we probably need a mix of automated search and manual human help – which is where Slack comes in. I’ve always thought Slack plus Notion plus Spoke would make a really powerful product (and I’m surprised Slack hasn’t made any major acquisitions in this space).
If you think about it, Slack is basically a search engine powered by humans: Most Slack messages are just questions. It’s 911 for when everything else fails. So if Slack had access to your entire knowledge base, it could answer at least the most commonly asked questions automatically. The rest would still get answered manually by the channel participants. Or your Chief Notion Officer.