I think that this blog post is hiding the fact that a team of engineers had their evening robbed from them, at no fault of their own, for something that was ultimately kind of trivial. There are numerous reasons why it's a bad idea to make engineers burn the midnight oil to rush the finishing of a product -- the first among them being Hofstader's law. If your team actually needed an additional day of work, and you didn't already have your product in the bag, you had no good reason to be courting the press in the first place. Mistakes happen in the news cycle. This is usually why you do an unadvertised soft launch shortly before a press release's publication date. This was a lack of due diligence, and not on the part of the engineering staff.
I think that there's this misconception in the startup world that the shipping of a stable product and its actual introduction in the market are efforts that are in parallel, in real time. In actuality, it makes a lot more sense to implement a lag of roughly one week between finishing a stable product and putting it out to the public. Many shops refer to this as "staging" or "vetting a release candidate". This doesn't seem like something that the folks at Cue considered before diving head-first into a hurried hackathon.
The last thing a responsible organization should do is punish the people responsible for making a stable and useful product by making them rush the last 10% of their efforts. I don't doubt that this kind of hurried time to market will result in another all-nighter down the road. I'll bet you dollars to doughnuts that after the 20th time hearing "Y Sin Embargo", the team was fatigued, annoyed, and ready to take shortcuts. So there, we have technical debt that could have been avoided if everyone just agreed to stick to their guns with the original release date. But most likely, the best solution would have been a soft launch preceding the any publication by at least a couple of days.
Let's be honest: if the immediate traffic from a little pre-arranged press is what makes or breaks your product, you're doing it wrong. As an engineer, if I see you have to put your entire organization into crisis mode over something like this, then I'm going to start taking recruiter calls more seriously.
It's negativity through and through here, and it's impossible to get away from it. Consider the following:
- You have a team that has been working incredibly hard pursuing a dream: that they can build a product, from scratch, that is truly useful to millions and millions of people. This is not an EA requiring engineers to pull insane hours -- this is a labor of love and everyone has considerable equity.
- It's a small team and that means each person has a lot of responsibility. Don't like ever having to pull an all-nighter? Never want to work a minute past 5? That's OK, but it means you have no real responsibility, and you will be compensated accordingly. Others choose more responsibility, and they man up when the shit occasionally hits the fan.
- Have you ever coordinated a massive press push with the top print publications? Then you have no business telling them how it should be done. It's extremely difficult, it requires dates locked down months ahead of time, and with that kind of advanced commitment, planning 100% accurately is beyond difficult.
- After this ordeal, the team posts an honest story from the trenches. Go to any successful startup and you will hear many war stories like this. Nothing is ever perfectly planned, unexpected shit happens. It's frankly awesome that they shared it and gave some insight into what goes on behind the scenes that you don't normally hear about.
- What is your reaction? You shit all over them. Tell them that they're horrible people and that they did it all wrong. Exclaim that you could have done it all right and they have no idea what assholes they are.
I have no idea what drives this extreme negativity on HN, but it's gotten completely out of hand.
One of the intended uses of this site is to exchange best practices and learn from the mistakes of other startups. I think that this blog post falls in the second category. And I don't think that there's anything wrong with me pointing it out.
I'm glad it worked out for the folks at Cue, but I think they could have avoided a lot of headaches with some better planning. I'm not suggesting some radical technique I just made up; I'm pointing out that there exist practices that would have prevented a needless scramble.
The circumstances of this story should not serve as an example to other startups, but rather a cautionary tale of what can happen to a shop without the proper release cycle and adequate project management in place. My tone may have been a little rough, sure. I probably wouldn't have even commented if the post had addressed what went wrong and what they learned from it.
I guess heroic last-minute efforts amount to a more entertaining story than adequate planning, but I think we can all agree on which method is a more reliable way to run a successful software shop.
> It's a small team and that means each person has a lot of responsibility. Don't like ever having to pull an all-nighter? Never want to work a minute past 5? That's OK, but it means you have no real responsibility, and you will be compensated accordingly. Others choose more responsibility, and they man up when the shit occasionally hits the fan.
No, that means you work for mature people who have planning and organizational skills. Having to come in at 3am is not a sign of responsibility, it's a sign of immaturity and failure of management.
Yes, it happens. No, you shouldn't laud it. If they were building bridges, people would be dying - but this isn't bridge building, so it's somehow more okay?
Putting up with that shit is fine when you have founders shares. If you don't, then get ready to clock out and iron the hawaiian shirt. Expecting the engineers to just deal with this is prebuscent.
Furthermore it's worth pointing out when trying to execute a coordinated media push (especially with traditional media), you have no choice but to give a lead time and put an embargo in place. This is standard practice when dealing with publications.
9 times out of 10 it goes off without a hitch. We were the 10%, so to speak.
hey moe, let me know when you ship a product rather than spending your entire day trolling on hacker news and tearing down the people who are actually doing something with their lives.
It's not that you criticize or critique other people's work, it's that you're consistently a dick about it.
Nearly all your comments are sarcastic and mean-spirited. IMO you're a liability to the Hacker News community.
One of the great things about hacker news is that we can have an open dialogue with the founders of startups. If you're acting like a jerk, they'll stop being so open.
If people like you keep acting like jerks, you're destroying the spirit of HN and the dialogue this community depends upon.
1.) What makes you think this isn't my "real" HN account? Long time lurker, new commentor. Unlike the founders you tear down, you're not using your real name or any identifiable email address or title/company you work for, "Moe." Afraid to put your name on your hateful remarks? Your account is just as throwaway as mine, you can easily create another vitriolic spewing alias safe from any repercussions. How about this - in the interest of open dialogue. Put your money where your big snarky mouth is. I'll say my real name if you say yours.
2.) Ahhhhh where did you offend me? Here for one.
This isn't random. I think it's mean spirited and insensitive that the Greplin/Cue founders and team members are here. Writing under their own names, their classmate from their YC batch is here writing under their own names and after what seems to be a brutal push with no sleep and lots of stress, all you do is swoop in with "bad management" snark and then swoop out.
your nastiness and at best insensitivity and at worst malicious disregard of others is only matched by your arrogance and superiority.
and of course your oh so witty response to Leo Kim of Foursquare, when he rightly called you out on making tests to make mongo fail .... I reached the same conclusion he did.
Troll.
Let's disregard for a minute, your obvious gaping technical knowledge and your sheer lack of Silicon Valley business acumen. Let's say you managed you convince PG that you aren't a prick and somehow you get into YC. In your weekly meet ups would you ever say any of the above comments to peoples' faces?
HN is suppose to an extension of YC, a place where hackers can come together, take about things of interests, their own projects, start ups, etc. Your entire comment history is subversive to that and completely against the guidelines:
"Be civil. Don't say things you wouldn't say in a face to face conversation.
When disagreeing, please reply to the argument instead of calling names. E.g. "That is an idiotic thing to say; 1 + 1 is 2, not 3" can be shortened to "1 + 1 is 2, not 3." "
Hacker news would be better off without you, you and people like you are bringing a cynical, negative, schadenfreude cloud over what is suppose to a culture centered around idealism, optimism and wanting to change the world- not insulting those that do.
What makes you think this isn't my "real" HN account?
The fact that it will be abandoned shortly after you realize the feebleness of your little vendetta. Yes, I can predict the future! ;)
would you ever say any of the above comments to peoples' faces?
Yes, actually I do. As you can see I even respond to throw-away accounts, I'm just straight-edge like that.
Other than that I'm sorry to break your heart yet again, but I'll probably continue to stick to the voting system for feedback rather than anonymous rants.
anyone want a dish with ad hominem and a sprinkling of superiority?
You're right, moe probably hasn't shipped anything. I haven't either.
But that doesn't mean that it's okay for you guys to treat engineers like this. Because one of them might just realize how undervalued and mistreated they are, and think about starting their own comapany...
and when they do they'll realize that building a company is fucking tough and sometimes you might have to ask your team to work long hours. None of the Cue engineers that have commented on here have said that they are undervalued or mistreated. It was a team decision - they want impact, they love their product. They wanted to work.
Frankly I'm disappointed that so many folks on hacker news think working long hours on a product is abusive. Newsflash--journalist work long hours to beat deadlines, musicians spend days in the studio recording their music, politicians and their aides spend long hours on the campaign trail, writers write well into the night. If you want to change the world, you going to have to bust your ass doing it.
And p.s. Moe has continually behaved like an ass on this forum, making snide insulting comments to countless people. He's the last person anyone should be white knighting. He's served an entire buffet of "ad hominem" and superiority. It's time he got called out for his dickish ways.
God forbid people actually work on something because they love it and want impact. In majority of those professions I detailed above, they make much much less than engineers but they still work hard and work long hours. Not everyone is mercenary. Why do you think so many rich people in tech still work? still build? If it's just about shares and profits, then head to finance. though I doubt they'd appreciate the whinging about long hours....
No one should ever be happy about working for free. Being a good team player means sometimes doing it anyways, but the situation that caused it should be rectified first.
How much is your salary really worth if you're working twice as much as someone making the same amount? What's the value of your free time? How frustrating would that be if you knew it could be avoided?
I'm sorry for being a stupid consumer but I still don't understand what cue is doing. I understood Greplin (search everything I have in the cloud), but I don't get what is point of Cue.
Which problem does it solve?
Maybe I'm not such a busy person. Or maybe the problem is that I don't care too much what is happening outside my email account.
Case in point: this last Wednesday, I had a going away dinner for a friend. This was put on my GCal. This came via a Facebook message. Without me doing anything, it put 2+2 together. When I opened Cue that morning, there was the event, with a link to the message.
When I fly in August, it knows enough to put all my travel info (passes, emails, events, etc) together in the same way.
Search is still there, and it still rocks. The added AI on top of that makes this killer - I would have used Greplin to surface that data, but Cue was smart enough to do it for me. It's solving a roughly analogous problem, but doing it in a much smarter, faster way.
We're scraping flight confirmation emails from several airlines, but getting all of the major ones working is challenging. If you've got a flight, it may be with one of the airlines we don't support in production yet. The easiest work-around is to put the trip in Google Calendar; they have a proper API. It sucks that you have to do that much, though. We're working on it. If you have the time, and you're willing to share details over email, we'd love to hear from you at [email protected].
Ultimately, we'd like to make a simple open API for various types of events. How cool would it be if (for example) airline companies attached a machine-readable version of your itinerary to the confirmation emails, the same way they send out HTML and text versions? Then you could integrate that with any tools you like. It would be like the "semantic web", except without all the hand-waving and overly-verbose XML stuff. This would be easy to do -- the budget would be peanuts to them -- but in order to get something like that happening, we need to convince those companies that there's a demand for it, that someone would actually use that machine-readable information if they provide it. Our automated email scraping is an attempt to break the chicken-and-egg dilemma, and hasten the spread of open machine-readable data for things that are currently just auto-generated text, by providing a compelling use for it.
kudos to Daniel and Robby and the Cue team - having just finished our launch, it's always stressful in the hours leading up. Rechecking lists, re-reading docs, running through flows again and again.
The fact you did all that days ahead of schedule is amazing.
Really excited about Cue! Can't wait for the Android version =)
I really liked Greplin (go to https://www.greplin.com/ to understand how it is related to Cue). I vouched for its ability to search twits (and other data source) for such a long time, and now I'll probably have to search for another product, as the team stops focusing on Greplin (and possibly eventually killing it).
Don't mind me though: I'm always happy when a team changes to a product that is more worth their time, even when I liked the former product.
The product that was previously called Greplin is a subset of Cue. The search functionality is not going away. We have no plans to get rid of it. It isn't an edge case, or legacy functionality - it's a core piece of Cue, and it is going to stay that way.
I'm another engineer at Cue, and I'd like to elaborate on this. We need search, and we'll keep on needing it, because our service is not all-knowing. No matter how cleverly we may predict what you're going to want to see next, we can't possibly predict everything; there always needs to be a way for you to search for things. We know this, and that's why, far from deprecating our search functionality, we're always trying to improve it. The goal is to minimize the time it takes to find whatever you're looking for.
This has been one of our most requested features. It is definitely something that will happen, but I can't comment on timeframe.
When the iCal support ships, I expect you can use that to keep us sync'd even before we have native exchange support. An exchange endpoint is how google sync keeps my phone up to date with ical events, iirc.
I'm not. There's lots of us. I'm planning on getting together and having a Mad Max style Thunderdome throw down at some point to get this all worked out.
I'm the Kevin Clark from Powerset (and subsequently Bing).
The cue app has search as a core, though secondary functionality. It doesn't seem to me that they will phase it out, though they may indeed change the way the app expects you to interact with its search interface which perhaps may run counter to your use case.
Cue has as of yet never worked for me. As in, it just says "Loading..." on the Today screen, and never loads anything. Could it be an iOS 6 thing? I'd really like to get it working; it looks so cool!
I think that there's this misconception in the startup world that the shipping of a stable product and its actual introduction in the market are efforts that are in parallel, in real time. In actuality, it makes a lot more sense to implement a lag of roughly one week between finishing a stable product and putting it out to the public. Many shops refer to this as "staging" or "vetting a release candidate". This doesn't seem like something that the folks at Cue considered before diving head-first into a hurried hackathon.
The last thing a responsible organization should do is punish the people responsible for making a stable and useful product by making them rush the last 10% of their efforts. I don't doubt that this kind of hurried time to market will result in another all-nighter down the road. I'll bet you dollars to doughnuts that after the 20th time hearing "Y Sin Embargo", the team was fatigued, annoyed, and ready to take shortcuts. So there, we have technical debt that could have been avoided if everyone just agreed to stick to their guns with the original release date. But most likely, the best solution would have been a soft launch preceding the any publication by at least a couple of days.
Let's be honest: if the immediate traffic from a little pre-arranged press is what makes or breaks your product, you're doing it wrong. As an engineer, if I see you have to put your entire organization into crisis mode over something like this, then I'm going to start taking recruiter calls more seriously.