The Forum Benchmark - Part 4
December 7, 2020
In part 4 of our IT Industry Forum Benchmark blog series, Vipul follows up on his earlier post on how we classify expert forum contributors and reveals the Best in Class forums in our analysis.
Three of the “best-in-class” forum statistics in our benchmark are open source forums (see below). In those cases, Stack Overflow is either the primary forum or has a significant volume of posts that rivals those on the branded community. For Kubernetes and all its variants, discussions happen in many StackOverflow tags, Github, in addition to each distribution/vendor community forum making finding it challenging for developers to find the best answer or expert.
In our publicly available benchmark here, we publish the number of community posts for a vendor or open source technology on Stack Overflow in addition to the ratings vs Best In Class. We update our benchmark continuously, so these comparisons are also updated regularly.
We were surprised that MVP’s still carry most of the weight in generating accepted answers even for most open source forums. To recap our prior post, MVPs (Most Valuable Player) are a subset of Community Experts who have posted higher than mid-point number of replies, and also have higher than mid-point number of resolved posts.
Here again is our diagram of the three zones of classification for the Community Experts.
Looking at three highly active open source community forums:
- Postgres: 8 MVPs generate 37% of accepted answers
- Kafka: 5 MVPs generate 36% of accepted answers
- Magento: 5 MVPs generate 39% of accepted answers.
Why is this important? These super-active and highly reputed MVPs are stretched thin, limiting the potential of the community forums to resolve posts in <24 hours. A clear opportunity is to empower Rising Stars and Active Experts to grow their contribution to MVP-level stats.
High Impact Contributors
We know that experts have a significant impact on the effectiveness of a forum. How may we quantify this impact? We started with a simple question: What percentage of top contributors together provide 80% of all resolutions in a forum? We call this: High Impact Contributors. This measures the fact that some contributors resolve a large percentage of posts (hence have high impact).
This is an effective measure of "expertise spread" across the community contributors. Forums should have a larger High Impact Contributors value, to avoid depending on the top select few (including MVPs) for the majority of resolutions.
For example, we have observed that proprietary vendor forums have a high concentration of expertise in a few individuals. These forums have a low value for high impact contributors. For example, about 13% of contributors resolve 80% of posts in Cisco forums. On the other hand, open-source product forums such as Kubernetes,Nginx, Postgres tend to have a higher value for High Impact Contributors. For instance, in the case of Nginx, 72% of contributors resolve 80% of posts.
The higher the percentage of High Impact Contributors, the less likely that a forum's effectiveness depends on a few key individuals. Nurturing more individuals to become experts increases High Impact Contributors, thereby avoiding dependence on the select few.