You say that like senior management typically makes decisions based on data and facts, rather than the reality where they just shoot from the hip using their "gut."

By the same logic implied here, we cannot trust senior management to know what they're doing.

PS - Also Google's measures of productivity are terrible, that's why they keep starting/killing new projects ever other week, and killed their golden goose in their so-called "20-percent rule."

I love how engineers pretend that they are the only ones acting rationally. It's as if they are playing chess and can never think there might be a reason behind the other players move, only their plan makes sense.

I've seen an awful lot of either not understanding what a measurement means, or not caring, among management, including out of "top-tier" folks from e.g. management consulting firms you've heard of. Talking really obvious "experimental error" types of things. Taking a measurement as indicating something when there there are obvious confounders which, if you don't look into them, are probably making your number meaningless.

I don't think engineers are the only rational ones, but I think the rational ones in management have noticed that their numbers being meaningful or correct only rarely matters for getting a "yes" or a promotion, so have responded... rationally, since getting useful figures and proving that they're useful is both 1) usually a lot more work, and 2) often not even practically possible.

The other option is that management is lying about what is important to them. I can accept that as well.

It is not at all required for engineers to be acting rationally, let alone for them to be the only ones acting rationally, for workers—of any type—to have a better understanding of what will let them get a good job done for the company than upper management.

This is especially true in large corporations, where upper management is very insulated from the realities of line-level work. It is also especially true in most publicly-traded corporations, where upper management is almost invariably going to care more about the trendline in the company's stock price than in their actual productivity, or even in the well-being of their workers.

If engineers acted rationally, we would not have so many technology fads.

I would argue that is rational for the engineers who become less employable if they aren't doing the latest stuff and creating new libraries.

The interests of engineers are often extremely detached from whether development is cost effective and maintainable by others.

This rational behavior in the face of stupid incentives is the same reason the rational set of management-sorts may not bother to make sure their facts & figures actually demonstrate what they claim they do, even when they know they probably don't. The extra effort wouldn't help them get the answer they're looking for, anyway, and might even harm their chances if they're honest about the level of uncertainty.

The part that's stupid is everyone, even the ones who know better, pretending a bunch of bad statistics and such aren't bad, but if the suits in the room are nodding along you damn well better not point that out. People succeeding in the management track either actually are bad at understanding these things, or (plenty often) do understand but also know to STFU about it.

> and killed their golden goose in their so-called "20-percent rule."

This is factually untrue. The only difference is Google is much larger, making it harder to launch new products.

The reason they start and kill so many projects is precisely because they are measuring top-level metrics and acting on them. A developer is going to look at how much functionality they implement, refactoring they complete or tickets they close and say "Wow, I've been so productive working from home". Management is going to look at growth and revenue, execution against top-level OKRs and the opportunity costs of competing options and maybe reach a different conclusion.

The reality is that neither is probably a WFH issue as a first-order effect.

Yet somehow, these boneheads haven’t driven Google to bankruptcy yet, so I don’t think their decisionmaking authority is in immediate jeopardy.

You don't have to make good decisions when you are riding the Ads Rocket!

There is a great deal of ruin in a FAANG, as Adam Smith would say.

Can't trust the senior managers of one of the largest revenue generating companies on the plant? With probably the highest Revenue per employee ever in the history of modern business (if not highest, up there).

I dunno, seems like they are doing ok at their jobs....

98%+ of Google’s senior managers today weren’t around when the company originally pioneered and grew the business lines that pay those manager’s salaries.

They didn’t build the ship, they’re just passengers.

Hence why most companies fall out of the S&P 500 as quickly as they rise. If you think Google is going to stay on top forever you’re in for a rude awakening.

From what I saw when I was there, a large part of the top brass were there since before IPO...

Statistically speaking, I wouldn't trust the senior managers of any company to genuinely know what measures will improve their business over the long term. Even leaving aside the depressingly large percentage who are fundamentally incompetent at managing (eg, the ones who think "yell louder and louder until people do what you demand" is an effective management style), a huge number of management at all levels are indoctrinated to believe that the next quarter's numbers, and hence the stock price, are the #1 priority at all times.

And even among those who are genuinely trying to see that the business does well for years to come, rather than the next three months and then who cares, even among the ones of them who are genuinely decent at managing a product, far too few understand the positive relationship between how you treat your employees and what you get out of them.

Because we've spent decades in this country being sold the idea that good management means forcing workers to work more hours, under tighter control, because "workers" are stupid, lazy, and/or malicious, and it is "managers" who really know how to do things right, by the very definition of their jobs.

They seem like they are doing a good job because we don’t have anything better to compare to.

America is one of the best countries in the world, therefore its leadership has to be smart and competent.

No, you cannot trust management at google. This site is clear evidence of that:

https://killedbygoogle.com/

I can think of no product at google that is not overshadowed by their policy of non-support. The culture that leads to their innovation also informs trust. If there is any such thing as "too innovative" then google is it.

This is why Azure is doing so much better than google in cloud computing despite the fact that google invented modern cloud computing. People trust Microsoft to support their products.

> “Productivity” is not something a person can measure themselves.

And management can? Please, tell me how. Give me metrics and methods for measurement that can effectively capture productivity. What's the KPI for "strong corporate culture" or "innovation"?

Because if we can't rely on individual experiences, then clearly the only other option is systemic measurements, so please, tell me how!

Not everything can be numerically measured. By what metric is Macbeth better than the Da Vinci Code?

Also, not all measurements should be shared. Let’s say that they have data showing that new hire productivity and skill growth is much lower when senior engineers are WFH. They’re not going to share that data when forcing senior engineers to be in the office, since it could foment resentment of junior staff.

> Not everything can be numerically measured.

So then why shouldn't we trust individual experiences? What's the evidence that those reports are not reliable while management perceptions are if this stuff isn't measurable?

> Let’s say that they have data showing that new hire productivity and skill growth is much lower when senior engineers are WFH. They’re not going to share that data when forcing senior engineers to be in the office, since it could foment resentment of junior staff.

They're already fomenting resentment by making a decision to force people back to work without any evidence to justify it.

That's somehow better?

BTW, I'm not saying they're wrong! They might be completely right. But to disregard individual experience while trusting management is measuring the right things is, frankly, pretty naive. The reality is management is as much about personal beliefs, gut instinct, and good ol' fashion tradition as anything else.

I suspect it's far more likely that management at Google simply doesn't like the idea of fully remote employees, not based on any specific data points, but simply because that's their preference. But assuming that's true, it's better to simply admit that then to make pseudo-scientific arguments to justify those preexisting beliefs.

I agree that Google management made this decision mostly on gut feel. But I think that it aligns with my own gut feel, which is that it’s going to be very hard for a fully remote team to invent the next Waymo car or iPad or whatever, and I also think that it will be very hard for a junior engineer to get the same learning remotely as they do colocated with 20 year tenured engineers. If I was CEO of Google I would have made the same decision.

> I agree that Google management made this decision mostly on gut feel. But I think that it aligns with my own gut feel

There's another term for that: Confirmation bias.

That's fine, but let's not pretend your opinion is any more valid than anyone else's, including the individual employees who feel differently.

The reality is no one, here, is actually basing their opinion on data. Absent that this is basically faith. Which is great 'cuz then we can all just agree to disagree. ;)

The difference is that the individual employee advocating for WFH is doing so because it’s better for them, not because it’s better for Google. The interests of capital and labor are not necessarily aligned here.

Frankly, that makes me even more suspicious of any claims by either side that are not backed by data. Confirmation bias isn't a good basis for credulity.

> By what metric is Macbeth better than the Da Vinci Code?

I like this question. Let’s say that we can create a list of 10 quantitative measures of the story, all binary and all comparative, e.g. which has the best pacing? which makes the best use of supporting characters? Now we can total up the answers and produce a number, e.g. Macbeth is better than the Da Vinci Code in 30% of our measures.

You could repeat this process pair-wise with lots of books and produce a ranking.

Yes but “best pacing” again requires a judgment call, it’s not a numerical metric, and some things don’t add in a logical way (“The Big Lebowski” has almost no plot whatsoever but is an excellent film).

Well, if I was going to pick one organization in my experience that I think is best able to judge developer productivity it would be google. They have a team of organizational analysts who study it and have developed a pretty sophisticated model. Interestingly, in past research they found that remote work is the top predictor of developer productivity.

> Interestingly, in past research they found that remote work is the top predictor of developer productivity.

Positive or negative?

I don't want to misrepresent it, but my understanding is that an expressed preference for working remotely is consistently associated with high self-evaluation of productivity. The abstract says "our results also suggest that software developers’ self-rated productivity is more strongly related to task variety and ability to work remotely."

https://storage.googleapis.com/pub-tools-public-publication-...

any source tho?

Fair point. If the decision was actually about broad productivity, and management was able to make a clear-eyed judgment about that, you'd hope they would reveal the data they used, so people didn't just assume it was a petty control issue. I think a lot of people would assume that's what it was actually about.

Some ammunition for the opposition: this HBR article indicates that companies who managed the transition to remote work the best increased their productivity during the pandemic. So, a question I'd want to ask if my hypothetical boss said "back to the office everybody!" would be why we hadn't been able to adapt as well as the best of our competitors, and why we didn't want to learn from them and get that productivity boost.

https://hbr.org/2020/12/the-pandemic-is-widening-a-corporate...