4 ms·
Once again, companies blaming strategic problems on ICs rather than real culprit: leadership. Or, rather, the lack thereof. Having worked at both Google and Fa
by cletus 4y ago
Once again, companies blaming strategic problems on ICs rather than real culprit: leadership. Or, rather, the lack thereof.
Having worked at both Google and Facebook I can tell you it's contradictory because in some cases you have an embarrassment of riches, hundreds or even thousands of heads, virtually unlimited resources (CPU, storage, networking), etc. Some make sense like Google+. I mean it was a failure and probably came way too late to succeed no matter what Google did but I understand trying. Maps, Docs, Youtube, Photos, Drive, Chrome, Android... all of these make sense.
I also understand you can't necessarily predict "winners" so to a certain extent you have to try things and expect failures.
Interestingly though every project I listed there (apart from Drive and Photos) was an acquisition.
On the other hand, you have projects desperate for people that turn into abandonware because they don't get sufficiently funded, even when they have PMF.
There are a ton of middle managers at big tech companies who exist only to get promoted and to empire build. You could, in my opinion, take everyone from L7 (M3 at Google, M2 at Facebook) to VP and fire 75% of them and be perfectly fine.
Both of these companies are now in what I call permanent reorg churn. Every few months you'll get an email saying your mananger's manager's manager's manager now reports to a new manager as part of a broad reorg. You've never met any of these people. This is a meme internally.
But what you have to understand is that reorgs are a way of avoiding the appearance of failure while appearing to be doing something. Don't get me wrong. Bad organizational structure can set you up for failure and a good org structure can help you succeed but reorg churn is none of this.
Reorg churn is simply changing the structure every 6 months. Nothing is ever in place long enough to determine if it succeeded or failed. People responsible for those decisions have probably moved on.
Additionally, at Google in particular, the amount of process required to do anything is insane. But don't worry. Bureaucracy busters has another 3 surveys for you to fill out to improve things. I once spent a quarter just babysitting a launch calendar entry.
The checklist to launch anything is insanely long. Even getting a small amount of resources requires Machiavellian machinations.
But sure, there are too many employees. Got it.
- Willish42 4y ago> There are a ton of middle managers at big tech companies who exist only to get promoted and to empire build. You could, in my opinion, take everyone from L7 (M3 at Google, M2 at Facebook) to VP and fire 75% of them and be perfectly fine. > Both of these companies are now in what I call permanent reorg churn. Every few months you'll get an email saying your mananger's manager's manager's manager now reports to a new manager as part of a broad reorg. You've never met any of these people. This is a meme internally. > But what you have to understand is that reorgs are a way of avoiding the appearance of failure while appearing to be doing something. Don't get me wrong. Bad organizational structure can set you up for failure and a good org structure can help you succeed but reorg churn is none of this. So many nails being hit on heads. Bravo. I see a lot of discussion in these threads around how hard it is to measure IC productivity, but nearly nothing about how to measure middle manager productivity (spoiler: you can't because their credit is based on work done by the people below them). In the middle of this hiring freeze stuff I got yet another reorg email from my company about my great-great-grand-boss, who I've never met, switching around to add a new layer of middle management new hires. Each of these is worth at least 5 IC headcount, probably more. I don't see a lot of criticism aimed at how _that_ band of the headcount doesn't match productivity...
- origin_path 4y agoChrome wasn't an acquisition, unless you count the company that only made sandboxes. But that hardly counts. Photos wasn't really an acquisition. Yes they acquired Picasa but Google Photos has no real connection to Picasa.
- pradn 4y agoThe bureaucracy to launch a project can be high. I think some of it is essential. A company really does need to ensure every project launch adheres to regulatory/branding/legal guidelines. We do have to adhere to a high a11y/i18n standard. A smaller company can just choose to ignore a bunch of these things. The problem is when these requirements mean you have to hunt down a lawyer who can flip the bit and they're busy with other things for the next few weeks. Same goes for security reviewers. Another thing is the ambiguity for what exactly is required for a11y/i18n. These things can be improved. I work in Google Cloud. After the main technical work is done, a feature launch requires probers, integration tests, metrics, alerts, dashboards, updating the gcloud CLI tool, updating client libraries in several libraries, writing internal support playbooks, writing external docs, writing an external business-side blog post. This can all be construed as bureaucracy since it's not the main feature itself. But we do need all of these things to provide a good customer experience. I've written the main code for a feature that took about 2-3 weeks and then spend 2-3 months doing the rest of these things.
- strikelaserclaw 4y agoto be honest, you can never be an extremely efficient org with a huge head count like google and to be honest i don't even see what their problem is, record profits year over year, how long exactly is that trend supposed to continue? in perpetuity?