3 ms·
Disclaimer that I’m in the current YC batch with a company trying to solve exactly this (resync.ai has a contact button if you’d like to learn more). I think t
by ethanbond 4y ago
Disclaimer that I’m in the current YC batch with a company trying to solve exactly this (resync.ai has a contact button if you’d like to learn more).
I think this is a real and very substantial problem on a bunch of different levels.
For ICs, they spend a bunch of time finding scattered resources which usually involves pinging each other on Slack “what’s the latest on X” or “where can I find Y.” This incurs continuous cost on the team’s attention.
For PMs, if they’re good they’re in all the different convos and tools and they become the “reducer” for the rapidly changing team state. But this duty takes away from the more strategic/future-looking work that a PM should be doing.
For director++ people, what we’ve learned over the last few months is that they don’t know what their org is working on either. In easy-money times you just keep hiring through the inefficiency but now companies are starting to take account of what their org is actually working on and we’ve heard some real horror stories about these efforts (e.g. discovering you have many more large eng projects ongoing than you have engineers).
The underlying fact that makes a solution difficult for each party is that no one views maintaining this truth as their job, or if they do, they do it in the cheapest way available (see the PM who just sits in every single meeting, watches every Slack thread, every file’s comments) which tends to just move the problem.
The key thing to solve is how do you make curation and updating of “shared truth” and its associated resources a natural byproduct of doing the work you’re already doing. You’ve got to figure out a way to connect the substance layer of the work (various documents, chats, mockups, issues) with the ”meta” layer of the work (indices, decision logs, status reports, etc) with close to zero work from IC level people.
We’ve found this problem emerges around 30 headcount, gets super painful around 100, and continues to just get worse from there.
- manibaur 4y agoThis is interesting. Will definitely reach out. I'm wondering where the AI bit comes into play. Is it in streaming updates across resources without involving grunt work from ICs?
- ethanbond 4y agoHaha, right now: nowhere. The .ai TLD is cheap :) We think there's a lot of interesting stuff we could do if we have this connection between the substance layer of the work and the meta/reporting layer though. Ideas like reconciling decisions made in Slack into a Google Doc, or summarizing weekly updates for people who have to read 20 of them per week. Looking forward to chatting!