This is a pet topic of mine to think about - so my (probably over-long) two cents below... For context, I worked at Google for ~4 years, and have also been at

This is a pet topic of mine to think about - so my (probably over-long) two cents below... For context, I worked at Google for ~4 years, and have also been at two other FAANGs, as well as done time at VC-funded startups. I'd like to think I have a good gamut of experiences to compare against both in the realm of smaller companies and massive megacorps. Google's problems are (from most severe IMO to least severe): - Poor senior management. Other posts touch on the middle-manager-heavy ranks of Google, but IMO that misses the forest for the trees. A bloated middle management layer is but a symptom of a sclerotic upper management. To add some color to this: - Google has a fundamental ethos that I haven't seen described in other writings and lacks a good term for it. In my head I call it the hypothesis of the free-range engineer. The foundational belief is that upper management's job is to get out of the way of the top talent they've hired. Product ideas should come from the bottom of the company - and even more importantly, they should succeed through intra-company competition. This is probably the singular most impactful organizational cultural belief at Google, and it informs how the entire company is structured. My position: the entire hypothesis is crap, and is a huge part of what holds the company back. - The net result of Google's culture is that upper management sets very little product direction, in favor of letting individual teams set their own priorities and ship independently. This means a lot of duplicative products and a lot of products that don't integrate with each other (see: Workspace account sign-in and... literally everything). Upper management has a light touch and is extremely loathe to mandate standards, resulting in a really mind-blowing level of product fragmentation. Much of this seems to be a deathly allergy that the company developed after the total and utter failure of Google+, where upper management did exert direct influence over product. The lesson they took from G+'s failure wasn't "have a better top-down product strategy next time" and was instead "top-down product strategy bad". - Upper management turnover is excessive. All big companies suffer from re-orgs, but Google practically makes re-orgs a sport. In addition to upper management being extremely loathe to give product direction, upper management (by this I mean Director-and-up) rotates alarmingly often, and each time it's accompanied with a re-org where projects are killed or deprioritized. I don't think in my entire time at Google I ever saw a multi-year roadmap actually executed to completion - each and every time some senior leader is replaced and the entire roadmap and strategy rewritten. - Google has poor product strategy overall. The company tends to swing for a lot of home-runs but structures these endeavors poorly. There are a lot of ambitious PMs advocating for huge multi-year megaprojects that end up being total failures. Upper management is credulous and doesn't demand proof of viability prior to sinking a ton of resources over multiple years into a thing (this feeds the turnover problem above - where inevitably when a multi-year thing flops heads roll, when the real problem is that upper management did not exert scrutiny over the project sooner). Big, ambitious goals are good - but viability needs to be validated quickly and continuously throughout the development process so that requirements can change (and yes, so that the project can be killed expediently if the fundamentals aren't there). Again, a failure of Google's upper management. - Continuing on that theme, Google's upper management is naive and credulous when it comes to product forecasts. PMs seeking to have their projects approved (especially big megaprojects) will make up absolutely bonkers projections for fundamental success metrics, and are often not seriously challenged on them. Senior leadership is ready to believe numbers that are facially absurd, and then are surprised when the product fails to live up to the unrealistic expectations. This phenomenon also drives the turnover/re-org problem above - since when the product whiffs the metrics by a mile somebody has to get the boot - but the real problem was believing the bad projections to begin with! - Promotion-driven development is a problem. But unlike most prognostications about Google I don't think it's their main problem. Leadership is at the core of Google's problems and promo-driven development is just an exacerbating factor IMO. But there is a serious problem with how Google structures incentives for ICs and middle management. The biggest concrete effect of this is that after launching, engineering teams empty out rapidly as people sense that no promos are forthcoming for working on the product, and look for greener pastures. This exacerbates the effects from above, where -- through an array of leadership failures -- the product was already on thin ice. The collapse of the engineering teams on it further buries the product and sends it on a death spiral. Anyway, this is already running way too long. At this point I think there needs to be a fundamental change to how Google pursues product if they hope to turn this boat around.