3 ms·
Maybe they are reading too much Hacker News and the "ship ship ship" we preach over here. I know I have heard some companies thinking like that without realisin
by iaskwhy 15y ago
Maybe they are reading too much Hacker News and the "ship ship ship" we preach over here. I know I have heard some companies thinking like that without realising it's not the best strategy for every company.
- mgkimsal 15y agoIt's a good observation. So many web startups think "ship early!" works for everything, and it doesn't. Certainly not hardware. Yes, for a website, where you can change the system in 5 minutes if you find a small mistake, or you can reindex a db and fix perf issues - yes, get it out there and validate and iterate. Hardware is a totally different beast.
- iaskwhy 15y agoThere are many reasons why you shouldn't just ship and with tablets like the TouchPad there are at least two good reasons: the hardware (for the reasons you said) and the price (which is expensive for something that is not unique or particular innovative). You can keep the features to a low though, just make sure they are tight. Like the Kindle.
- msbarnett 15y agoI don't think Apple really does anything different than what HN preaches. The best way to do "ship ship ship" well is to cut your feature set to the absolute bone, even if that means your initial launch is with just one single feature, as long as that one is polished enough and works well. This is basically Apple's modus operandi; they'll take the hits from the nerds on not having every feature under the sun in return for being able to entice people with a small, well chosen set of features that "just work". The HP's, Acer's et al are stuck in the mindset that they simultaneously need to shoe-horn in every idea they can think of while "ship ship ship"ing by throwing the product out the door as an unfocused, half-baked mess. Aggressive curation of your feature set is a powerful strategy for nearly every company.
- commandar 15y agoSoftware and hardware are totally different ballgames. With software, if you ship and you find that your product needs major changes, you iterate and you're not really out of anything. Ship bad hardware? Well, you end up with things like partners demanding you take back a quarter of a million unsold units.