TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

Read This Before You Build Uber for X

231 点作者 capocannoniere超过 7 年前

31 条评论

mooneater超过 7 年前
I think this misses another critical point.<p>The platform owner only retains control of the transactions, if they keep the buyer and seller from getting to know one another.<p>For Uber, thats easy, because the nature of the service means you want a close car (typically, a different one each time)<p>But for Uber for local babysitting, why would the buyer and seller not just leave the platform permanently after meeting, and transact directly, saving both the fee?<p>&quot;Sharing economy platform&quot; commercial success depends on avoiding lasting local human connection.
评论 #15153822 未加载
评论 #15150736 未加载
评论 #15151688 未加载
评论 #15152973 未加载
评论 #15154750 未加载
评论 #15151482 未加载
评论 #15150835 未加载
评论 #15150877 未加载
评论 #15151334 未加载
评论 #15151558 未加载
评论 #15150925 未加载
评论 #15154728 未加载
kbenson超过 7 年前
Am I the only one that sees &quot;Uber for X&quot; in a somewhat negative light? In my mind Uber does relate to dominating an industry and getting big, but more in a Microsoft of the 90s sort of way, where there a lot of negative baggage associated with it because of what looks to be unrestrained greed. For example, I&#x27;m not sure I would want to advertise my actions as trying to be the next Oracle. That probably only comes across as entirely positive to those that don&#x27;t know much about Oracle or only care that they make an obscene amount of money.<p>Is this just me? Or is it perhaps time to retire this idiom?
评论 #15150506 未加载
评论 #15150450 未加载
评论 #15150431 未加载
评论 #15150480 未加载
评论 #15150374 未加载
the_stc超过 7 年前
Nice timing. We are working on &quot;Uber for Sex&quot; or &quot;Tinder + Buy It Now&quot;. On the lower end of the market, it can be a low skilled job in the sense that a lot of people know how to have sex well enough to satisfy a lot of clients. Certainly going upscale or into niche offerings that doesn&#x27;t hold true. So our system is somewhat split into people primarily booking on cost and schedule, versus people booking on the entire experience. I think we&#x27;ll manage to handle both acceptably enough for all parties involved.<p>On-demand? There are a lot of scheduling difficulties. Getting a sex worker on short notice, even in a popular and mostly-legal city is a nightmare. And it runs into the Airbnb-pre-instant-booking issue: You contact several people, they don&#x27;t all get back right away. Then an hour later, you have several replies, and since you can&#x27;t book them all, some end up annoyed you&#x27;re wasting their time. Not to mention any screening issues or other deal-breakers.<p>And if someone does have some last-minute availability? They can&#x27;t easily fill that slot. It&#x27;s hard to just offer sex work in a few hours here and there - you&#x27;ll spend all your time answering phones.<p>But it isn&#x27;t just the &quot;Uber&quot; aspect that we&#x27;re working. Sex workers lack cohesive unions and shared resources. We&#x27;ll be able to pool for things like security agents. We can handle client verification one time and be done.<p>We&#x27;ll do photography of workers, then advertising and bring clients in. This benefits less entrepreneurial workers that don&#x27;t want to try to manage an online presence, deal with SEO, try to A&#x2F;B test their photos and copy, etc.<p>There are other similarities. Uber had to avoid LE, so do we. Uber has to deal with partial cash payments and reconcile, so do we. Uber offers some insurance, so do we (legal). But we&#x27;ll be far less sexist than Uber, that&#x27;s for sure.
评论 #15151874 未加载
评论 #15152201 未加载
评论 #15152129 未加载
评论 #15151808 未加载
评论 #15155001 未加载
评论 #15153491 未加载
评论 #15153402 未加载
评论 #15152315 未加载
评论 #15151807 未加载
评论 #15154069 未加载
slg超过 7 年前
The way I see it, there are three main things that contributed to Uber&#x27;s success:<p>1. As discussed in the article, the service they provide is commoditized making both employee skill and customer to employee relationships irrelevant. If I like the person providing the service, what incentive is there for me to not to continue to work with that person outside of the app?<p>2. They connect people who are looking for a service with an untapped market of unskilled contractors who can provide that service. I would never become a full time taxi driver but I might consider driving for a ride sharing service if I needed extra cash. A layperson like me is never going to work for an Uber for massages.<p>3. There is preexisting regulatory capture in the market that allows the new entrant to provide a much better or cheaper service than the incumbents (although here be legal dragons).<p>Ideas like Uber for massages or Uber for housecleaning don&#x27;t work because they fail all three of those conditions. Companies that meet multiple conditions tend to be more successful. Postmates for example meets the first two conditions while Airbnb meets all three.
评论 #15150801 未加载
s17n超过 7 年前
The article doesn&#x27;t mention providers dealing directly with the customer to avoid paying a percentage to the app. In any service business like housekeeping where you have a long term relationship between the customer and the provider, this is going to be an issue.<p>The article indirectly discusses this by talking about how an app can provide value to pros in specialties with various characteristics, but I think that it should have started with by saying &quot;if you think you can take a 30% cut of every transaction in a service business without actually providing substantial value <i>in every transaction</i>, it&#x27;s not going to happen.&quot;
评论 #15150512 未加载
b1daly超过 7 年前
I&#x27;m amazed that this article and most of the comments implicitly assume that Uber is succeeding, and therefore useful as a an example of a successful platform startup.<p>Uber seems to be failing badly so far, losing increasing amounts of money. The ridiculous corporate shenanigans Uber has been generating seem of a piece with a company that had massive investments, with expectations, and no clear path to profitability.<p>Here&#x27;s a link to a page that has some detailed, and damning, critical analysis of Uber&#x27;s prospects.<p>The gist of his argument is that Uber is gaining market share by substantially selling their product below cost. But they have not found some kind of magic solution to lowering the cost of providing their services by taking advantage of scale or network effects.<p>The author argues that Uber is actually a high cost provider, in comparison to traditional taxi services.<p><a href="http:&#x2F;&#x2F;horanaviation.com&#x2F;Uber.html" rel="nofollow">http:&#x2F;&#x2F;horanaviation.com&#x2F;Uber.html</a>
ahh超过 7 年前
This is pretty interesting and apt. One thing I&#x27;d quibble with is their claim that many professional skills aren&#x27;t fungible; this is true in general but there are many cases where a non-fungible skill is fungible for a major use case.<p>In particular I&#x27;m thinking of doctors. I care a lot about having exactly the right doctor (a good one, a specialist, etc) much of the time, but there are many medical services where I simply need anyone competent to pass their boards with a prescription pad. (I have been taking the same three or four prescriptions since I was six for a few common ailments--I don&#x27;t need someone to tell me special information about albuterol.)
评论 #15151463 未加载
cjcole超过 7 年前
The &quot;skill versus urgency&quot; illustration has me wondering if the author foresaw or even intended the resulting arguments over whether or not, for example, web design requires more skill than dog training. It&#x27;s one way to get people talking about the article, I suppose.
jaypaulynice超过 7 年前
Uber succeeded because they can reliable predict demand for services. They have both historical data (rush hours, regular commute, etc.) and also future data (sporting events, concerts, night life, etc.)<p>I looked at starting an Uber for healthcare when it was getting hot, but I thought again about the problem. It&#x27;s not possible to predict who will be sick when and where and doctors are very expensive. Having doctors go to people&#x27;s homes is a lose lose situation because the time they spend driving is time they could spend helping patients.
ilaksh超过 7 年前
Great article.<p>Another thing to consider before you build Uber for X: there will come a time, perhaps in the next couple of years, when people stop tolerating having monopoly technology companies as intermediaries for everything. Or they will tolerate some business involvement when it really is a deep value add, but there will be much more competition and the types of fees being charged will have to be much more reasonable.<p>The reason these companies become so huge so fast is of course because what they are mainly doing is managing basic logistics and information flow (payment being part of that).<p>The reason we need these companies, or think we need them, is because we do not yet have practical and popular decentralized systems that can replace them. Most people are not aware of the existence of decentralized platforms that could possibly replace the centralized servers and banks.<p>But these decentralized platforms do exist and more innovations happen every day. Popular examples are Bitcoin and Ethereum but that barely scratches the surface of what is actually already a large ecosystem of peer-based systems with many types of capabilities.<p>And I know that it will be really hard to convince most people that these services could exist without centralized servers or payment intermediaries. But maybe look into places like reddit.com&#x2F;r&#x2F;rad_decentralization or r&#x2F;Ethereum to start to get an idea.
评论 #15152813 未加载
评论 #15151623 未加载
clairity超过 7 年前
the article nailed it: for an &quot;uber for x&quot; to work, it requires (1) commodity, low skill service combined with (2) on demand scheduling.<p>one thing it&#x27;s 2x2 matrix implies is that both taxi services and delivery services share a common appropriateness as &quot;good&quot; automatable services. but that&#x27;s not true. they certainly differ on the scheduling metric - delivery is not nearly as on-demand as taxi service.<p>but there is another, more important variable that differentiates these two: lcationality.<p>while a taxi service can start and end anywhere, delivery services are constrained to start at particular locations, which makes logistics and routing the primary challenge to overcome. you have to get vehicles to the pickup point each time, which wastes time and money. delivery companies try to overcome this in a variety of ways:<p>1) trying to deliver &quot;anything&quot; (deliv, postmates) to approximate a more even geographic distribution of pickup points<p>2) subsidizing delivery through membership (amazon)<p>3) making delivery incidental to the core service (instacart is a marketing channel disguised as a delivery service)<p>i&#x27;m sure there are dozens of other ways that companies are trying to overcome this inherent last-mile logistics problem, but it&#x27;s a real differentiator between taxi and delivery.
swsieber超过 7 年前
The solution I think is a little more like what a salon does - a good, shared platform.<p>I think if somebody wrote a Uber for x <i>planning</i> on the 1 on 1 repeat business, they&#x27;d be fine... and how?<p>1) Provide real value for provider - an easy to use billing platform, scouting out of the upfront work and risk, and classification of the customer and (here&#x27;s what I think would be the big one) work schedule management.<p>2)Provide real value for the customer, it be being able to evaluate a provider, get good ratings (and here&#x27;s the kicker) if the preferred provider isn&#x27;t available, they don&#x27;t have to go hunting.<p>3) Let providers recommend each other. Temper that with user ratings and platform specific ratings. Have users report when a job isn&#x27;t done right, or another visit needs to be made, etc. But the provider peer ratings are like &quot;I have too much work to provide x, you should try John, I trust his work.&quot;<p>4) The categorization of work could have meaningful use to the providers - it lets providers specialize further than they normally could otherwise, and provide a platform to fall back on.
rdtsc超过 7 年前
&gt; The reason for this high failure rate is most founders underestimate the intricacies and nuances of the specific service industries they are trying to uberize.<p>(tongue in cheek) So build a platform which allows startups to build Uber for X. The real market is a desire to emulate Uber with lots of VC money pouring in. This is not unlike selling shovels during gold rush.
thwarted超过 7 年前
This is a decent writeup because it defines what &quot;an uber&quot; is and what it means to be a business like that.<p>How many &quot;Uber for X&quot; businesses label themselves that way because of interactions like this one, where customers and investors are so focused on Uber as some kind of canonical startup that there&#x27;s attempts to squeeze unrelated business models into it?<p><a href="https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=syoqjYLDs48#t=19m31" rel="nofollow">https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=syoqjYLDs48#t=19m31</a> (specifically 21m56s and 24m9s)<p>This video is four years old, so it&#x27;s good to know that claims of &quot;Uber of X&quot; are getting long in the tooth. It&#x27;ll help to start agreeing on definitions and get away from the buzzwords.
tedsanders超过 7 年前
This article is excellent.<p>It takes a common idea (Uber for X), puts it through a simple but appropriate 2x2 lens (skill &amp; scheduling), and produces a valuable insight with broad applicability (most markets are not like taxis).<p>Thanks to Sam Madden for writing it and to you for sharing it.
smacktoward超过 7 年前
The fascinating thing about this writeup is that it&#x27;s even necessary.<p>The analysis it presents is a pretty straightforward one. It&#x27;s the kind of basic contemplation of product-market fit I&#x27;d expect anyone starting a new business to undertake before spending a significant amount of time or money on it. In other words, it&#x27;s a simple sanity check.<p>What makes this one fascinating is that apparently so many people out there have tried to start &quot;Uber-for-X&quot; businesses <i>without even doing a simple sanity check first</i> that it&#x27;s necessary for a third party to do it just to prevent even more lemmings from marching off the cliff.
Grustaf超过 7 年前
&gt;Technology cannot necessarily miraculously make the pro more skilled at their job<p>Well, in the case of Uber that&#x27;s exactly what it did. Uber wouldn&#x27;t be possible at all without consumer GPS routing apps, that&#x27;s one of the reasons (at least ostensibly) for regulating taxis. Cf The Knowledge, London taxi drivers&#x27; training that is so extensive it actually can be detected on a brain scan.<p><a href="https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Taxicabs_of_the_United_Kingdom#The_Knowledge" rel="nofollow">https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Taxicabs_of_the_United_Kingdom...</a>
Triiistan超过 7 年前
&quot;It worked great for local transportation services, so why could it not be applicable across all types of industries?&quot;<p>Dit it really work that great for Uber? Are the drivers doing all right? Did the company make the world progress? Is it even profitable? Are the clients happy about Uber&#x27;s service? These are important questions to consider before trying to replicate their model, and I&#x27;m pretty sure not everybody feel the same way about this success.
jkingsbery超过 7 年前
&quot;It is rare that one plans their day around transportation&quot; - it wasn&#x27;t the main point of the article, but that&#x27;s not my experience. Many of the people I know who, like me, commute into New York from the suburbs plan their day around the train (when is the express train, when is the last train that runs every 30 minutes instead of 60 minutes, and so on).
throwaway100z超过 7 年前
Apologies for an unrelated question but, what is the formal name of this chart indicated in the blog? Scatter plot?<p><a href="https:&#x2F;&#x2F;blog.ycombinator.com&#x2F;wp-content&#x2F;uploads&#x2F;2017&#x2F;08&#x2F;Before-You-Build-Uber-For-X-Chart-1024x731.png" rel="nofollow">https:&#x2F;&#x2F;blog.ycombinator.com&#x2F;wp-content&#x2F;uploads&#x2F;2017&#x2F;08&#x2F;Befo...</a>
评论 #15151122 未加载
capocannoniere超过 7 年前
Submitted this article as I&#x27;m curious about what the HN community has to say. Apologies if this is a re-post
评论 #15150134 未加载
mindhash超过 7 年前
I think urgency should be driven by lead time (when i want it - when i get it).. for e.g. Getting a web design guy may take you upto 3 weeks or more or 45 days if you hire but if you get the same guy in a week or less then the model still works.<p>Comparing Web design urgency with taxi doesnt make sense.
cossatot超过 7 年前
What is an example of an on-demand, app-mediated service that is actually profitable, regardless of whether labor is contract, etc.? Is any one in this space actually making money off customers?
zengid超过 7 年前
Minor side-note, but I think that &#x27;skill vs urgency&#x27; graph underestimates the skills required for delivery service providers. Source: I work for UPS.
评论 #15151146 未加载
jonny_eh超过 7 年前
Web design is less skilled than life coaching?
评论 #15151075 未加载
michaelmior超过 7 年前
&gt; It is rare that one plans their day around transportation<p>I do this pretty regularly as do many people who rely on public transit.
odammit超过 7 年前
I only build the Netflix of Ubers.<p>I seriously hate these phrases. It&#x27;s OK to use adjectives, nouns, verbs and to say the word &quot;app&quot;.<p>I was talking to a guy at SBYP once that said he was building &quot;the Netflix of snacks&quot; and I asked the guy how he planned to digitize said snacks and he looked at me like I was the asshole.<p>I get it but use some words to describe your thing, man.
ap46超过 7 年前
Way to demotivate other startups!
评论 #15152293 未加载
评论 #15150587 未加载
wellboy超过 7 年前
Insightful article, but a bit hard to distill the distinct points. From my own experience at an Uber for X, there is one overarching theme that dictates the whole dynamic that I think isn&#x27;t addressed. The article makes the following points.<p>1. &quot;The more skilled the service, the harder is to gain traction, because quality in service isn&#x27;t reliable.&quot; This is true to some extent, but there is a bigger underlying dynamic, namely that the fewer people can offer that service, the harder it is to gain traction, because it takes longer to get that service on demand (e.g. 20 mins instead of 1 min), which decreases the user experience.<p>2. &quot;There are services that require more time for for scheduling and some require less.&quot; The article doesn&#x27;t really say which one is better, but the latter is better, because it makes it more spontaneous, people can make the decision to demand the service easier and do it more often.<p>3. &quot;Maximizing conversion through services where no sales pitch is required, because quality is almost always the same (taxi).&quot; This goes back into point 1, where low skill all have the same quality. It adds the factor that when quality is the same, people are less picky and make more purchases.<p>4.&quot; Maximizing profits through retention to get repeat customers.&quot; I didn&#x27;t understand this point, since critique for Uber for X is that repeat customers would bypass Uber for X, in order to avoid fees, such as it happened with Homejoy, Plumbers etc. This can be prevented through making ratings valuable, i.e. giving pros incentives to not bypass the platform.<p>The overarching themse is critical mass. All of the mentioned points explain what factors prevent critical mass form being accomplished. Summing up these are<p>1. The more skill you need, the less service providers you can get per square mile, the longer it takes to fulfill a service, the harder this makes it to get traction.<p>2. The shorter it takes to schedule the service, the more people will make the decision and book the service.<p>3. The less the quality is a factor, the less people will worry about making the purchase.<p>4. Repeat customers is where the money is. Make it easy for the consumer and pro to maintain these relationships, BUT, keep them on the platform.<p>5. This point I add myself, it is locality and it is the most important point. It means that you have to figure out a way to get traction in a new city in a scalable way, because you can get traction in SF with masseurs and clients, but you need to do it all over again in Fresno, Seattle, London, etc. There is some traction that comes from other cities, but it is little. For some services, traction in a handful city is enough and you cna be a $100m company, but for most, you need to be able to get traction in several dozens of cities quickly.
grabcocque超过 7 年前
Uber&#x27;s main success was to take something that already existed (taxi services) but then found a way to work around local labour laws on a technicality.<p>Because Uber drivers are all self-employed, they&#x27;re not unionised or protected in any way. Thus, Uber pays its drivers a pittance, and is able to undercut and undermine existing taxi services who are forced by unions and law to pay drivers a living wage.<p>That&#x27;s the key to being Uber for X. Ask: can I use the internet and shady business practices to exploit workers to undercut a business that previously workers were protected in?<p>If the answer is Yes, then you can be Uber for X.
评论 #15150741 未加载
评论 #15150633 未加载
评论 #15150713 未加载
评论 #15150842 未加载
评论 #15150754 未加载
评论 #15151038 未加载
评论 #15150721 未加载
评论 #15150594 未加载
评论 #15151498 未加载
评论 #15151134 未加载
评论 #15150608 未加载
评论 #15150949 未加载
评论 #15151552 未加载
undoware超过 7 年前
X11 is getting a rideshare service?<p><i>Sigh</i> well I guess it has modules for everything else
评论 #15150336 未加载