Hacker News Reader

Comment by bla3 | Hacker News Reader

I think this is both obsolete and subtly wrong. Big corps want products with global impact. When a product is new, it's evaluated for its potential. When it's been around for a while, it's evaluated on how well its growth is matching the estimated impact. Eventually, growth has stopped. If the product lands at an active base of a billion users or so, it's a success and people definitely can make a career out of maintaining and improving it.

And to make working on new stuff a bit less incentivized, promotions to higher levels need to see at least some realized growth and not just getting v0 out these days as far as I know.

View on HN

google will never get a new service to a billion users right now

They have the netflix problem (or does nexflix have the Google problem)

they have killed off soo many projects that no one want to invest their time into new ones unless they reach the multi-year stage, and they can not reach the multi-year stage unless they get millions of users...

It is a catch 22, and neither company seems to want to address or even acknowledge is a problem for them

I don't think Netflix has a particular problem; maintaining series for more than a couple of seasons has always been hard in broadcasting. It's easier to spot now because people have complete visibility: instead of caring about a couple of series in key time slots, they can know care about all series, so it's more noticeable that most series die a sad death. Addressing off-ramps in a nicer way would be an innovation. The problem is that production costs are so high, that keeping a losing franchise running is extremely expensive.

On the other hand, Google plays in a field where maintenance costs are a rounding error. You can basically freeze a product and keep it running forever for pennies, and these products get embedded in our daily lives, which is why killing it looks so bad - I can always pick up a new series with zero effort, whereas migrating off a productive service takes time and effort.

Netflix is different from traditional broadcasting in that they always order an entire season. That makes the cancellations more dramatic because instead of a pilot or few episodes like in the past they instead keep performing a Firefly on their viewers over and over again.

The joke I've heard is "six seasons and a movie" that goes back to broadcast TV series. Something about contracts running six years, so costs balloon after that, and a one-time movie is much easier to schedule a group of actors for. I've long wondered why Netflix doesn't do it more often. I'm sore about the cancellation of "Santa Clarita Diet" and I think it could easily be wrapped up in a movie.

Costs (salaries) can balloon on successful series. It's also the case that, if a series is getting close to the number of episodes that can make for profitable syndication deals, it made sense (in the context of network TV) to stretch things to that point. It's also the case that, as long a series is still doing pretty well, odds are that putting something new in its slot will be a net negative.

Traditional broadcast also just had different incentives than streaming.

A successful sitcom captured a lot of eyeballs that could be translated into advertising dollars. It could also be a magnet for viewers to watch perhaps somewhat less captivating content in adjacent timeslots a la "must see TV Thursdays on NBC."

Netflix, on the other hand, mostly needs to fill its catalog and have enough really must see content so that subscribers keep renewing.

There's overlap between the two models. Both reward series that people really want to watch. But the specifics are subtly different.

This is a joke from Community about open ended TV shows typically having enough narrative steam to get a compelling run of six seasons and a movie at best. Not a financial thing.

Community has crawled through the trenches to get its six seasons and now finally a movie and I wish them the best.

Sometimes shows naturally have a 1 or 2 season arc and sometimes (rarely) they go out strong at 6 or 7. But, yeah, for me sometime around season 5--even if a show has successfully mixed things up a bit--I find tuning in for new episodes more and more optional in most cases.

I wasn't thinking of it as a Community joke, but rather a Dan Harmon quote/meme that he placed in Community (he's nothing if not meta). If my memory is correct (and it does get a bit fuzzy over time), my comment above is paraphrasing him or some other explanation of the saying.

There's no 6-season cutoff. It was just an estimate of a successful show.

The cutoff is 100 episodes, which traditionally is a magic number for US syndication deals. An ongoing series with 100 episodes under its belt will get a good deal for everyone involved.

Most traditional series will hit that number in 5 seasons, but obviously nobody wants to make a deal for a dead series; so you produce season 6 too, to make it palatable. By this point, however, your talent bill has probably grown significantly, so making new seasons is less and less profitable - as well as narratively hard. So you pull the plug and start afresh.

> You can basically freeze a product and keep it running forever for pennies

We’re moving to a security culture where a Cloud product not only has to be constantly rewritten to keep its features (not only upgrade libraries but also swap out the old Apache libs), but logs must be constantly inspected to address security risks, and GDPR issues have to be automatically lodged with CNIL. This “let’s keep this service only” has become as dangerous as leaving a Windows XP computer connected to the Internet.

Yes and no. Things do need monitoring and support. But hardening is doable, as well. With more and more use solidifying exactly what to allow.

Upgrades and new features often make hardening tougher, due to a combinitorial explosion in exposed surface area.

2 "Seasons" of a netflix show is not even 1 season of broadcast... Most Broadcast shows run 20-24 Episodes per Season. Netflix is 10 at the most, some 6 episodes. per "season"

That is part of their problem, and why the cancellations feel even more abrupt and why people do not want to invest time in them. it is hard to tell a compelling story in 6 episodes of TV. So if a netflix series get 2 seasons, 12 Episode. it is just starting... even though to netflix thinks it should have a captive audience. that would just be the midseason of a Broadcast show

Most of the popular shows on broadcast would never had made it to be popular under the netflix model of 1000 shows @ 6 episodes...

Depends on the country, it is totally ok for some british detective/crime series to have 6 episode seasons, because each episode is an hour or 1,5 long, they are not filled with pointless stuff to make them longer, etc.

I recommend Line of duty.

The short series format works if you write a story that is suited to it. When done right it feels like the middle ground between movies and a full TV series, not as starved for time as a movie and moves much faster than a 12-24 episode TV series. Of course the issue with Netflix is they tend to take what would traditionally be a 24 episode TV series, break it down into small parts, then cancel it at a inconvenient point in the story.

IMO the problem of Netflix is not how the split the shows or their duration, but their quality.

Yeah, I think that's the real issue. People complain about 6 episodes not being enough, but the truth is that a lot of those 6-8 episodes now feels like filler - probably because some constraints (ad cuts on scenes...) have disappeared, so writers feel emboldened. Not everything can be The Wire.

I have a friend who's written Netflix original shows, and I'll say it's the opposite - Netflix has very specific instructions for how scripts must be structured based on viewership data. Netflix knows where people tend to pause, and where viewership tends to fall off, and have formulated guidelines for scripts to counter it. They require scripts to conform to their requirements - add a plot twist in episode 2 at the 60% mark, for example. The primary character should occupy 60% of episode 1, and 40% of episode 2, and so on.

> 2 "Seasons" of a netflix show is not even 1 season of broadcast

I'd say most of their episodes are longer, on average. But that's not entirely relevant anyway, the main cost is in setting up the production as a whole.

> it is hard to tell a compelling story in 6 episodes of TV.

It's even harder to do it in one episode - the pilot.

In the past, a lot of pilots would be produced and never seen by anyone but execs. Then something like 5% of projects would be picked to air, and typically be given two seasons to iron out any problems - because it would look bad on execs who had fought to use a scarce slot for it, so there was a level of continued support for a while. Episodes were very short in length, and being effectively just containers for ads, it was alright to schedule tried-and-tested material, even if performing in a relatively mediocre manner.

Pilots are now less necessary, Netflix will invest on screenplays; but that means that a lot of the selection previously made by execs is now made by viewers. NF execs have less skin in the game, because there isn't a real limit to the amount of picks they can have. Netflix needs subscriber numbers to grow, so it needs big hits more than tv channels did. They have no incentive to continue pushing nonperforming series.

> Most of the popular shows on broadcast would never had made it to be popular under the netflix model

I'd argue that most of the series produced under the NF model would never even have existed on broadcast.

> 2 "Seasons" of a netflix show is not even 1 season of broadcast...

Yes, most streaming-first platforms use shorter seasons than is typical for US broadcast (but that was also often true of US cable, and is also true of, e.g., British broadcast.)

> That is part of their problem, and why the cancellations feel even more abrupt

Mid-season, especially first-season, cancellations on broadcast have always been a thing (as have soft-cancellations by moving—often multiple times in a season—to a less valued timeslot, especially when timeslots mattered more.)

> Mid-season, especially first-season, cancellations on broadcast have always been a thing

A TV or cable network canceling a show is very different from netflix doing it though. When a network cancels a show mid-season fans get pissed but the network removes that show from broadcast and replaces it with shows that do get full seasons and proper endings.

When netflix puts out an unfinished product by starting a show and then not finishing it, the fans get pissed, and it just sits like a giant turd in their library so that month after month and year after year more and more people will start the show, get to where it cuts off and also get angry that netflix wasted their time.

Even if a netflix show seems to under-perform, it's in netflix's best interest to make sure every story has some kind of conclusion because some percentage of subscribers will still be able to find value in it. A finished product is a win for their library.

Otherwise the unfinished show will just sit on their servers being forever unwatched by anyone aware that Netflix screwed the show's fans, or it becomes a trap that will only make anyone who does start it angry at netflix.

Netflix should either just commit to concluding any show they start, or remove any shows that get left without a conclusion from their library entirely (throwing away all of the money they invested)

> So if a netflix series get 2 seasons, 12 Episode. it is just starting

True Detective season one: 8 episodes and done => one of the most acclaimed TV shows in many years.

Mare of Easttown: 7 episodes and done => one of the most acclaimed TV shows in many years

If you make actual stories instead of trying to create ongoing scenarios, you can hit it out of the park in less than 10 episodes.

Those are 1 hour episodes with no commercial breaks, so the run time is pretty comparable to a season of network TV.

But that's tangential to the point: once you've watched the 10 episodes, what is there to keep you subscribing?

That wasn't the point at all. The nominal point was a hand-wavy claim that you can barely get started with a small number of episodes, which is clearly not true.

A friend who's written Netflix original shows says that Netflix has a lot of data regarding user watch behavior, and provide specific directions for how shows should be structured to maximize viewer retention. As a theoretical example, they know people tend to lose interest around 60% mark in episode 2, so that's where you have to insert a plot twist.

The result is shows that keep your interest just barely enough for you to keep watching …until you grow even tired of that.

I think the "hard to tell a compelling story in 6 episodes" thing just depends on the type of story. Previous replies mention several excellent shorter-run series.

Often, if I find out a show is a miniseries (or I guess they call them "limited series" nowadays) I am much more likely to watch it. Too many shows go the route of "how many seasons can we get this to run" instead of "how long will it take to tell this story with good pacing to conclusion?"

I get tired of watching shows that seem designed to run indefinitely. They all have the same sort of pacing that stretches things out unnecessarily just to pad seasons. Shows like that feel like basic cable to me when I watch them. I'll take a good 6 or 8 episode arc over indefinite seasons any day.

Amen. Knowing that the show runners had the ending in mind when they started making the show gives it a much better chance of being a cohesive package than a vehicle for somebody's desire to build a franchise.

> Google plays in a field where maintenance costs are a rounding error. You can basically freeze a product and keep it running forever for pennies, and these products get embedded in our daily lives, which is why killing it looks so bad

That would be nice if it was true, but that is not how it plays out at most tech companies.

I have specifically targeted this in my career, thinking about engineering productivity and even spent a number of years working at a company that provided software for manufacturing and explicitly learning about lessons learned in that field.

What usually happens is maintenance costs are not taken into account when building a product and a team is not disbanded after building something. They continue to run the larger teams/orgs until one day some VP looks at the expenses and decides to simply kill the entire product.

So in a world of having hiring freezes, having someone come in to help reduce total costs, from AWS to headcount is a way to free up people and get "new hires" without additional cost.

Some companies and individuals seem to get this instinctively, but many are happy to not make waves and try to improve the system. For some in management, adding value without headcount doesn't mean anything for them individually so they wont.

Increasing the total value over total costs of teams and orgs is my jam. I can take over products that have high maintenance burden, but are valuable, dramatically reduce the maintenance burden, then turn around with my free headcount and go and do it again and again growing the total value produced. This is the 10X-100X game.

when I was doing research on ratings, one thing I noticed: no matter how popular or unpopular a show, its an almost certainty that the number of viewers viewing something goes down over time.

The first season almost always has more viewers than season 2, season 2 almost always has more viewers than season 3. The only real difference between the extremely popular shows is that the decrease from season to season is not as pronounced, but its still there.

I was like wow, I'm surprised they even bother with season 2s.

I think google needs to stop launching new products under its own name.

If they dont explode as desireable but are sustainable, spin them out.

Then again, maybe we are all over estimating how sustainable these products really are.

It’s worth noting the competitive space for Google Glass Enterprise is still busy - products like Vuzix and Realware are succeeding at the same use cases.

So if you were in charge of procuring such a solution three years ago and opted not to go for Glass because Google has a history of killing projects, you saved your firm a lot of wasted time and money.

Exactly.

Indeed. I got so disappointed with their cancellation of The OA that at this stage I'm pretty unwilling to start watching a new niche show unless it's already completed its full arc (or was defined from the start as a "limited series").

Exactly, ironically I was finally just starting to get comfortable enough with Stadia being around long enough to buy a few games on it, literally weeks before the shutdown was announced.

Specifically I re-bought the Jackbox games which were perfect for it since I could now simply cast them when visiting friends/family. I also recently purchased a budget Chromebook for travelling and probably would have bought more games for that.

I would bet that the majority of non-techies couldn't name a single discontinued Google product.

Outside of the tech world Google still has a stellar reputation. Most people don't seem to care/know/notice any of the privacy issues, support issues or product continuation issues.

So I think your statement has some merit for certain types of products - ones that are targeted at the tech crowd - but overall I don't think it's generally true.

I would bet that the majority of non-techies couldn't name a single discontinued Google product.

There are 283 products listed on Killed By Google. If you're right then Google must have a massive problem communicating with non-techies. What's the point in launching anything new if they're so bad at getting users?

Sure, that list has a bunch of items. Have you looked at what they are? Here's a random sample of five items from it:

1. YouTube for PS Vita - a YT app for hardware that's been discontinued. Nobody would know of this because the PS Vita was a tiny platform that Sony quickly abandoned; normal people did not own Vitas.

2. Postini - email security software that Google bought, stopped selling standalone, and integrated into GMail. Nobody would have heard of this because it was enterprise software, not consumer.

3. Gesture Search - a toy Android app for searching for contacts / apps with handwriting recognition rather than a virtual keyboard. Nobody would have heard of this because it's a tech demo solving a tiny problem that people just didn't have. But if people had wanted this kind of search, it would have been integrated to the OS.

4. Building Maker - software for making building models on Google Earth. Nobody would have heard of this since it was incredibly niche. How many people want to model buildings into Google Earth? 10k?

5. Google Schemer - some kind of location-based activity sharing. This does appear to be a legit consumer application, though not one I'd heard of.

We can analyze another course of action from companies such as Microsoft. Not saying that is better or worse but showing an alternative in big corporations.

Microsoft launched similar initiatives every time for decades even failing at prior tries. In the UI front they "play" with Silverlight, WFP, Windows Forms, MFC and a lot more. Try and retry.

I also think that these issues are worse when you are dealing with hardware: do they open the codebase to continue playing with the devices? I am on the European side of not only repairing youself but opening "dead" code or somethink like that. It is just an idea.

Microsoft is suffering form the reality that Azure is the OS that matters, WinDev seems to be getting new devs without much background on how Windows used to be, and wiht constrained resources versus anything Azure.

Also note that DevTools is no longer .NET only, rather any language that has tier 1 support on Azure.

Problem with promoting on potential is that it’s rife for nepotism and poor judgement.

People want to get promoted, managers want their people promoted. Inflating impact can quickly become the name of the game.

> I think this is both obsolete and subtly wrong

Maybe you are right, not sure. But this is my impression from working at Big tech. New thing is easy to measure, compare these: "launched X in region Y" vs "prevented possible incident X from happening" (your manager would ask, how do you know if that incident will happen? maybe you spent too much time for over optimizing the solution?)

Focusing on individual product growth is just one strategy an organization may follow. And it's a strategy that if followed blindly will eventually make said organization be seen as unreliable to customers.

I wouldn't pay Google for any services right now. I prefer products and services that last more than mere single digit years (looking at you Stadia).

This sounds like a B-school idea, not one based on observed reality.

> promotions to higher levels need to see at least some realized growth and not just getting v0 out these days as far as I know

I don't think that's very far.

Yeah, when your cycles are 6 months long that’s about as long as you need to maintain anything, if that. Even 1 quarter might be enough. The effort to promote “Landings not launches” was not particularly successful :/

all the incentives for promotion or getting a high rating is on creating a new product that has cross team impact. Even if your new product improves your team / current product, No - that doesnt have cross team impact, no promotion / high rating for you.

There are exceptions, but those are outliers, not the norm. so as an individual, you have to do what's best for you - that means don't take your chances at improving an existing product, make something new and force other teams to use it.

Reply on news.ycombinator.com