Organizing Iterate’s internal knowledge

In Iterate we have had multiple initiatives to catalog our knowledge, using systems like Notion or Confluence. At some point someone sets up one of those system, lots of content is created, then the intiative often dies down.

Those systems are also great at categorizing knowledge everyone agrees on, like who to ask when your computer dies or the current state of some internal project. But it’s difficult to keep content people disagrees with up to date. For instance, at our old confluence-page, we have a page for “Core Books Everybody Should Read”. When I read that list now, there are at least one book I don’t recommend anyone reading.

I think the main problem is that even though many will agree on this list, there will be disagreements. Most people (myself included) don’t want to get into a fight over which books to read, therefore the list on confluence basically belongs to the author, and not to the collective.

Another issue with confluence and notion, is that they have a hierarchy. All content lies somewhere in a tree-structure. When the space is new, this makes it easy to find content, its all there. However after a while the hierarcy goes really big, and it’s more and more difficult to find content. Even worse, when you write content it’s difficult to figure out where you should put it. The result is often that it’s not used at all.

Where does people put the content? Often it’s not written. In Iterate, some opinion-peaces is written at peoples own blog. For instance, one of our founders have written many really interesting articles at substack and My co-founder at Unicad have his own blog-ish, and you are looking at my own experiment.

At some companies writing content like this might be frowned upon, it’s distributed and it’s even public! In iterate, when we observe that people does things in a way we don’t want, we don’t try to change their behavior.

Instead we want to embrace the chaos!

In the 1990s, Google changed the Internet. Before google, there was portals where content was categorized. Much like how one might use Confluence or Notion. After Google, there were no portals, instead we searched for what we wanted to read about. (Note: I was 6 years old when google was started, the details might be a bit fuzzy).

We want to do the same with knowledge in Iterate. Instead of having one place where everything is written down, we let all employees select how they want to write opinion-pieces. Then we index it, and makes it all searchable.

As everything we do in Iterate, this is an experiment. Maybe it will work and we will be able to create knowledge from chaos. Maybe it wont and it will stay chaos. Time will show.