Evolving Product Team Typology at Doctrine
In early 2020, when I joined Doctrine, our platform was already providing various legal content types. That content, from court decisions to legislation or lawyer directories, was leveraged to bring legal insights to our customers. Despite the various content, the main focus was, however, still mostly on making court decisions accessible and intelligible and we did not have a well-established practice of scaling legal content types acquisition and distribution. In that context, we had 2 product teams* centered around what we will call “functional” scopes:
- One centered around acquiring, enriching, and making data available on the platform. That was the “Insights” team.
- One focused on making the content discoverable, mostly focusing on the search & recommendations features. That was the “Discover” team.
Not to mention our fantastic Foundation team which encompassed various non-directly product-related aspects of the platform from the Kubernetes infrastructure up to some common pieces of Software architecture like authentication.
This team typology was working well at the time, as again, the two product teams were primarily working on decisions content and were experts in their technical domain adding new functionalities to existing content. However, in the last couple of years, our strategy evolved and we progressively started to add a lot of new content including, for example, legislative assemblies documents or collective agreements. Regularly adding new content types was not scaling with that team typology. Indeed, each time new content had to be added, the two teams had to get up to speed on the content and synchronize with each other.
For that reason, to support our scale, we decided, as the first step, in 2021, to evolve to a content-centric typology. In early 2022, we created content-centric teams that were in charge from acquisition to availability in the search and recommendation engines of a subset of Doctrine’s content. That way when new content had to be added or enhanced there was no need for two teams to intervene and ramp up on the subject. Each team is now fully responsible for its content and for adding new content to its scope. As an example, you don’t anymore have to understand all the subtleties of how legislative assemblies work when you don’t belong to the team in charge of their content. Considering how complex those matters are, this is a big time saver.
Note that as we did not want to lose the functional scope knowledge we also added to the content scope of each team some “historical” functional scope.
We ended up with two content-centric teams:
- Maât is in charge of lawyers and enterprise directories, legal commentaries, and collective agreements from a content perspective. Additionally, the content discoverability functional scope from the historical Discover team is also in this team's scope.
- Thémis is in charge of legislative assemblies, legislation, and court decisions from a content perspective. Additionally, the acquisition and enrichment functional scope from the historical Insights team is also in this team's scope.
If this new organization is solving the issue of scaling the available content types as we have seen in practice by adding yet another content scope team in 2022 to address a new market (the “Distribute” team), it did also come with some shortcomings. In particular, even if we made sure to assign the historical functional scope to some content-centric teams, the deep functional expertise around, for example, the search engine was at risk to be lost or not considered strategic enough by the content-centric teams lacking investment in the long run. We knew about that from the very start, however, we did not want to rush and we wanted to make sure we build new teams at the rhythm we can recruit and correctly onboard new engineers and products into the team.
For that reason, we decided to progressively (re-)introduce functional-centric teams whose objectives would be to be the owner of the non-content-specific functional part of Doctrine. We started spawning those teams from the content teams when they grew enough to be split in two. We actually just spawned our first functional-centric team in October 2022 and are looking forward to the second one somewhere in 2023. In the process, we also made sure to rationalize the scope of the content teams by having three clearer focus points around primary legal sources, secondary legal sources, and enterprise & paralegal information. The schema below shows you where we should be at some point in 2023:
All of this was possible because we were able to grow the team staffing. When it comes to the results we still closely monitor our KPIs following those changes to see how they impact our efficiency. However, this is still a work in progress as we still need to complete the staffing of the first of the funtional team and prepare for spawning a second one next year. For that reason, we are heavily recruiting both in product and engineering. If you want to be part of the journey of building the first legal intelligence platform in one of those squads just get there and apply!
*there was a short period where the structure was slightly different but I’ll skip it for the sake of simplicity
** in yellow content squads, in orange functional squads, in green content squads with functional responsibilities
