6 ms·

You can amp up a team by challenging them to give the very best possible user experience, so that your customers loves your product. Exhorting the team to "cru

by tytso 7y ago

You can amp up a team by challenging them to give the very best possible user experience, so that your customers loves your product. Exhorting the team to "crush the competition" or "dominate the market" is not necessarily the best way to amp up a team. In fact, I'd argue it's the worst possible way, since that means that playing dirty patent troll tricks or financing things like the SCO lawsuit is a legitimate way to "win", as opposed to winning by delighting the user.

com2kid 7y ago
> Exhorting the team to "crush the competition" I worked on a consumer product in a highly competitive space. Our competition released products that dropped frames and had laggy UIs. The discussion wasn't "We're going to make a smoother UI so our users have a wonderful and pleasant experience." it was "Hah, how the hell, they have literally 20x the CPU power, a GPU, over 1000x the RAM, and they are dropping frames? Let's show'em how its done!" In certain domains, and when trying to perform at certain levels, you need every single human emotion you can draw on to get enough energy to find the motivation to do amazing work. Is that needed 99% of the time? Nope. But MS sure as hell could have used someone pissed off about their mobile experience being terrible about, oh, say, 3 years earlier than what did happen. Another example. Gmail launched 2004. Imagine if 2002 a senior leader at Microsoft had burst into a meeting of the Hotmail leadership team and started shouting "Why the hell do other email providers even still exist? And why the fuck do I have to still delete emails every couple of days? I want every other free email provider out of business right asap. Lose money on it, I don't care, just make us the best!" But instead, someone at Google got upset about the same problem and made a product that completely dominated the entire consumer email market. If you go into a competitive space (especially consumer!) with the idea of "we'll look at the market research and do what is needed to establish a place for ourselves" then no one cares about you. "We're going to make a product that makes everyone else on the market look like incompetent jerks" is an actual mission statement that can be used to get things done. The competition has a painful first run experience? Yours isn't better, it is perfect. Why? Because screw the competition. Your competitors have fat profit margins? Yours are razor thin, because screw the competition. There product has an industrial design comes in black and ugly grey, you have designer colors, because screw the competition. If you are a multi-billion dollar company, and you have the resources to do everything better than anyone else in the field, then you either perform at that level, or you lose. The iPhone launch was to be a tour de force of this ideology. "They have ugly plastic buttons sticking out everywhere, we have metal and glass. They have a crappy joke of a keyboard, we'll reinvent the entire idea of an on screen keyboard. They have a shitty sales experience in stores? We'll partner with one carrier, and train the living daylights out of their sales people. Their UI is hard to use? We'll launch a world wide ad campaign where we show subtly people how to use the device during the ad." I was at MS at the time. The idea there was "We'll let our OEM partners decide on how best to differentiate their product, they are after all the ones doing the main value ad and we rely on them for manufacturing." Apple said screw that entire model, they just made it themselves, and they made it amazing. There is an incredible energy on teams motivated by the ethos of doing the best and steam rolling the competition through sheer excellence. I'm not advocating for dirty business practices, but holy hell, telling your engineers "do good enough" is going to get you "good enough." When they wake up and go to work and the entire office is saying "we're going to make a product that puts everyone else to shame", you get a great product.
- breck 7y ago
  I agree with this. Swears and cursing are an efficient and motivating way to communicate orders of magnitudes. Curious, what was the attitude like on the early Band team? I absolutely loved it, and was active on the beta, but didn't have a front-row seat to the early development.
  - com2kid 7y ago
    > Curious, what was the attitude like on the early Band team? I wasn't allowed to tell people I was interviewing what the job was for. (Most people figured out that an entire floor of HW and Firmware engineers were working on Hardware.) I had to convince people to join me by telling them about the team and about why they wanted to work there. I said things like "this is a classic Microsoft team, we're going to take something that is elite and expensive and make it affordable for everyone, fuck elitism, Microsoft makes products everyone can use." "We are understaffed and underfunded, this will be the hardest you've ever worked in your life, but we are making something that is fucking amazing and you'll be telling stories about this for years to come." "Want to build an OS? How about an entire graphics system? Say yes and I'll have you at a desk next week doing stuff you never thought you could do." The Band was critically under funded, I was told that it was under 60 million across all the years we were active. Apple Watch famously had a billion. The energy was insane. We were a startup within MS, no one else knew we existed but the C-level leadership and maybe their directs. My team went to Ikea and got our own desks and put them together late one night. We stole (!) a TV from another team someplace in the building, wheeled it into our work area, and watched Twitch Plays Pokemon. The firmware team wrote a runtime from scratch. Then my team took an off the shelf UI toolkit, rewrote almost all the drawing routines, rewrote the font rendering, and rewrote another ~60% or so of the remaining code, and made a butter smooth v-sync'd UI. We had 96 MHz and 256 KB of RAM. 30 FPS locked, the UI could run at well over 60 but we had limited bandwidth to the display controller. One day I read an article about touch latency on various mobile OSs. I said fuck it, we can do better than any of those. We optimized our touch system, max of 2 frame latency, min of 1. An amazing team from MSR came in and wrote us a machine learning gesture recognizer that also did motion estimation so that the screen tracked your finger almost 1 to 1 compensating for errors in the touch sensors. Very understaffed, everyone did many things. Early on a bit of dead weight tried to attach itself to the project, it got cut eventually. Lots of bad decisions. The team spun out of Xbox Accessories. No one there knew software except Firmware, I was hired on because I knew UI, all their other software engineers had no idea. UI was the green light on the Xbox 360 controller. Amazing firmware engineers though. Lots of issues around the cloud and mobile apps, they were started much later than the rest of the product and never had time to go through proper planning, esp. for the Band 1 launch. Our design team was balls to the wall amazing. Our lead designer put so much thought and care into everything that was on the screen, it blew me away. It also taught me how much a real designer cares about their work. If I am lucky I'll get to work with a designer like that again in my life. The battery life would've been a day longer on Band 2 but we had a shitty accelerometer chip that couldn't do step counting on its own (or maybe it could but it required a lot of programming to get it to that stage, I never was made 100% clear on that), so we had to take raw data and process it with the CPU. Horrible waste of power, I always was sad we never were given the time to add that extra 16-20 hours of battery life. Everything we did we wanted to do better. We had very limited hardware, so everything had to be magic in code. The EE team did magic in hardware for sure, the screen portion of Band 2 was less than 2/3rds the thickness of the Apple Watch v1, so they had very little space to work with. Every decision was "how can we do this better than anyone else?"

by tytso 7y ago

com2kid 7y ago

> Exhorting the team to "crush the competition" I worked on a consumer product in a highly competitive space. Our competition released products that dropped frames and had laggy UIs. The discussion wasn't "We're going to make a smoother UI so our users have a wonderful and pleasant experience." it was "Hah, how the hell, they have literally 20x the CPU power, a GPU, over 1000x the RAM, and they are dropping frames? Let's show'em how its done!" In certain domains, and when trying to perform at certain levels, you need every single human emotion you can draw on to get enough energy to find the motivation to do amazing work. Is that needed 99% of the time? Nope. But MS sure as hell could have used someone pissed off about their mobile experience being terrible about, oh, say, 3 years earlier than what did happen. Another example. Gmail launched 2004. Imagine if 2002 a senior leader at Microsoft had burst into a meeting of the Hotmail leadership team and started shouting "Why the hell do other email providers even still exist? And why the fuck do I have to still delete emails every couple of days? I want every other free email provider out of business right asap. Lose money on it, I don't care, just make us the best!" But instead, someone at Google got upset about the same problem and made a product that completely dominated the entire consumer email market. If you go into a competitive space (especially consumer!) with the idea of "we'll look at the market research and do what is needed to establish a place for ourselves" then no one cares about you. "We're going to make a product that makes everyone else on the market look like incompetent jerks" is an actual mission statement that can be used to get things done. The competition has a painful first run experience? Yours isn't better, it is perfect. Why? Because screw the competition. Your competitors have fat profit margins? Yours are razor thin, because screw the competition. There product has an industrial design comes in black and ugly grey, you have designer colors, because screw the competition. If you are a multi-billion dollar company, and you have the resources to do everything better than anyone else in the field, then you either perform at that level, or you lose. The iPhone launch was to be a tour de force of this ideology. "They have ugly plastic buttons sticking out everywhere, we have metal and glass. They have a crappy joke of a keyboard, we'll reinvent the entire idea of an on screen keyboard. They have a shitty sales experience in stores? We'll partner with one carrier, and train the living daylights out of their sales people. Their UI is hard to use? We'll launch a world wide ad campaign where we show subtly people how to use the device during the ad." I was at MS at the time. The idea there was "We'll let our OEM partners decide on how best to differentiate their product, they are after all the ones doing the main value ad and we rely on them for manufacturing." Apple said screw that entire model, they just made it themselves, and they made it amazing. There is an incredible energy on teams motivated by the ethos of doing the best and steam rolling the competition through sheer excellence. I'm not advocating for dirty business practices, but holy hell, telling your engineers "do good enough" is going to get you "good enough." When they wake up and go to work and the entire office is saying "we're going to make a product that puts everyone else to shame", you get a great product.

breck 7y ago
I agree with this. Swears and cursing are an efficient and motivating way to communicate orders of magnitudes. Curious, what was the attitude like on the early Band team? I absolutely loved it, and was active on the beta, but didn't have a front-row seat to the early development.
- com2kid 7y ago
  > Curious, what was the attitude like on the early Band team? I wasn't allowed to tell people I was interviewing what the job was for. (Most people figured out that an entire floor of HW and Firmware engineers were working on Hardware.) I had to convince people to join me by telling them about the team and about why they wanted to work there. I said things like "this is a classic Microsoft team, we're going to take something that is elite and expensive and make it affordable for everyone, fuck elitism, Microsoft makes products everyone can use." "We are understaffed and underfunded, this will be the hardest you've ever worked in your life, but we are making something that is fucking amazing and you'll be telling stories about this for years to come." "Want to build an OS? How about an entire graphics system? Say yes and I'll have you at a desk next week doing stuff you never thought you could do." The Band was critically under funded, I was told that it was under 60 million across all the years we were active. Apple Watch famously had a billion. The energy was insane. We were a startup within MS, no one else knew we existed but the C-level leadership and maybe their directs. My team went to Ikea and got our own desks and put them together late one night. We stole (!) a TV from another team someplace in the building, wheeled it into our work area, and watched Twitch Plays Pokemon. The firmware team wrote a runtime from scratch. Then my team took an off the shelf UI toolkit, rewrote almost all the drawing routines, rewrote the font rendering, and rewrote another ~60% or so of the remaining code, and made a butter smooth v-sync'd UI. We had 96 MHz and 256 KB of RAM. 30 FPS locked, the UI could run at well over 60 but we had limited bandwidth to the display controller. One day I read an article about touch latency on various mobile OSs. I said fuck it, we can do better than any of those. We optimized our touch system, max of 2 frame latency, min of 1. An amazing team from MSR came in and wrote us a machine learning gesture recognizer that also did motion estimation so that the screen tracked your finger almost 1 to 1 compensating for errors in the touch sensors. Very understaffed, everyone did many things. Early on a bit of dead weight tried to attach itself to the project, it got cut eventually. Lots of bad decisions. The team spun out of Xbox Accessories. No one there knew software except Firmware, I was hired on because I knew UI, all their other software engineers had no idea. UI was the green light on the Xbox 360 controller. Amazing firmware engineers though. Lots of issues around the cloud and mobile apps, they were started much later than the rest of the product and never had time to go through proper planning, esp. for the Band 1 launch. Our design team was balls to the wall amazing. Our lead designer put so much thought and care into everything that was on the screen, it blew me away. It also taught me how much a real designer cares about their work. If I am lucky I'll get to work with a designer like that again in my life. The battery life would've been a day longer on Band 2 but we had a shitty accelerometer chip that couldn't do step counting on its own (or maybe it could but it required a lot of programming to get it to that stage, I never was made 100% clear on that), so we had to take raw data and process it with the CPU. Horrible waste of power, I always was sad we never were given the time to add that extra 16-20 hours of battery life. Everything we did we wanted to do better. We had very limited hardware, so everything had to be magic in code. The EE team did magic in hardware for sure, the screen portion of Band 2 was less than 2/3rds the thickness of the Apple Watch v1, so they had very little space to work with. Every decision was "how can we do this better than anyone else?"