[website] Open Tech lead - Core role

[mnajdova]

• I have followed (at least) the PR section of the contributing guide.

Preview: https://deploy-preview-42189--material-ui.netlify.app/careers/react-tech-lead-core/

[mui-bot]

Netlify deploy preview

https://deploy-preview-42189--material-ui.netlify.app/

Bundle size report

No bundle size changes (Toolpad)
No bundle size changes

Generated by 🚫 dangerJS against 341bd2c

[mnajdova]

We target role IC6 or above, should it maybe be IC5 and above?

[Janpot]

Not sure whether it's better, but we could also write this in the present tense

You oversee the development of ...

[Janpot]

A few suggestions, but at your own discretion.

[oliviertassinari]

@mnajdova I think we need to start thinking about how we will reopen this role. Code-infra and Core Tech Lead looks to me like two full-time roles. Today, Jan covers both responsabilites, but I think this is overwhelming (impossible to cover by a single person), so there is a need to pick a side.

---

As a side note on this, as it's related to staffing. I think we can define "correctly staffed in the core team" with:

1. 2020 level of all the metrics from the KPI board. https://www.notion.so/mui-org/KPIs-1ce9658b85ce4628a2a2ed2ae74ff69c#c02a5c878ee846e1b23b2918a5bdd4aa.

For example, (for the ratio and PR volume): https://tools-public.mui.com/prod/pages/communityPerMonth

(issue starting around 2021).

2. The sum of the contributions stats of the maintainers match the sum of what we could sustain in the past (during the high growth phase in % market share in React, for example, this was sustained from 2018 to 2021:

2a. https://ossinsight.io/analyze/eps1lon

2b. https://ossinsight.io/analyze/oliviertassinari

(interactions happening on Twitter or Stack Overflow counts very much as well, but it's not visible in ossinsight, internal org Slack or internal meetings should likely be significantly discounted, it's not useful by the community, especially when discussing controversial topics, it's precisely the value of having this open).

The sum average was above 1,000 issue comments and 1,000 PR reviews per month (retaining high-quality contributions, no making tradeoffs on this). So this level would be outstanding IMHO.

I think it was amazing we could do this with only two people back in the days, the fewer people you need to do this, the fewer the information is spread between minds, the fewer connections are lost, the better the depth of the execution. So overall, it's not just about target, but also getting there with the smallest possible team.

Now, I don't think we can realistically reproduce this, not with the current org setup, it was an anomaly. It's really really hard to reproduce: I assumed that both Sebastian and I were working 80+ hrs/week, we were not really paid for it during those times (what mattered was a sense of ownership and having a lot of fun), and we were both Vercel hiring targets. I guess we were both junior as well (in the great scheme of things)

It's also worth mentioning that other projects have found success with 50% of those metrics. Instead, they chose to spend the extra time on Twitter or with the community on Reddit, conferences, etc. So those contributions metrics are definitely not the only way.

From a team organization perspective, decisions were very centralized, I would green-light almost everything, this allowed us to move fast and with decisiveness, I would make mistakes sure, but correct them quickly. Our timeout to merge PRs without reviews was 3 days, this was going faster as well than today.

---

I don't know exactly where we are with this 2. metrics today. I think we should measure this. This should be possible with a custom OssInsight SQL query / API request. Since Material UI focus got split into Base UI + Pigment CSS + MUI System + Joy UI + Material UI, all of "Core" counts.

As a target, I think we could start to set a target at 50% of this so 500/month.

• If we are above, great, them to me it's only a matter of keep up with the other KPIs, e.g. point 1. We have a lot of open issues, we don't have that many issue open er month compared to the past, and yet, it starts to go over 2k total open issues.
• If we are below, we work on filling the gap, and then see how this feels. See if we need to go further or it's enough.

Task created: https://www.notion.so/mui-org/eng-Measure-core-team-activity-stats-128cbfe7b66080e6bc28ed0640bd5661

A disclaimer: those are not metrics that we can directly use as leading indicators (as soon as they are gamed, we lose), oss insight didn't exist for the time frame I shared, it wasn't possible for us to look at them back in the day and try to boost them up, we didn't care, it's only in retrospect that they are interesting. What we care about back then was: notification down to zero, zero open PRs, zero open issues (in this order).