4 ms·

Show HN Aftermath: Releasing Followords.com this morning

12 points by bcambel 14y ago

This morning I've submitted our project to HN. http://news.ycombinator.com/item?id=4099054 Got pretty helpful feedback. But that wasn't the only thing. 478 people requested invite out of 755. 63% conversion rate. Almost 9K unique visitor visited our website. At some point in the morning I was seeing 400+ active users on the website. Unbelievable. 35 people registered without telling them how to register. ( You cannot register directly, there is only a login form - but they figured out a way :) http://imgur.com/Kbf9J 6K people visited directly from Hacker News, 1K from Twitter ( I guess Popular Hacker News post ) I was afraid that Nginx might not hold up, but that was not the problem. I wasn't ready for it! Stuck on my chair from 08:00 am till 02:00 pm. I am still feeling anxious. People tweeted about us without using our application. https://twitter.com/#!/search/realtime/followords (even in Spanish) Today another lesson learnt: So if you are working on a project. Release it. Ship it. Let people tell you how much your project, website or tagline sucks. Embrace it. Think about it, thank them and move on. I need some sleep, tomorrow is another great day. Thank you all for your contributions.

bmelton 14y ago
I didn't see it this morning, but seeing this now and just signed up for the beta. I'm still not entirely sure what it does, even after reading the linked thread, but I try to do whatever I can to support all the 'Show HN' threads, so even if it sucks, you've got my registration (and I'm not presuming that it will suck.) Nginx is almost certainly not going to be your failing piont, and can certainly scale to beyond 400+ active users, assuming that it's configured correctly and on decent hardware. Your database could crash, or one questionable join could tear the whole thing down, but Nginx is probably the least likely component in your stack to fail. As for the lesson learned, that's spot on, and one that I waffle on every now and again, even though I've done it and know it to be canon. Best of luck. It's a space I've got a lot of experience in, and 'interestingness' vs. 'importance' isn't really gotten right anywhere yet. Hopefully you've nailed it.
- bcambel 14y ago
  Thanks Barry, interesting vs importance highly personal and it's a hard topic.
maxhacker 14y ago
That is an awesome response from a Show HN! Gives me some motivation to finish my project and show it to the world. Also, I just requested an invite to the your service (didn't catch the actual HN post earlier, but I saw this and was intrigued with your project). Thanks for a good post!!
nuhm 14y ago
"478 people requested invite out of 755." Does this mean 755 requested an invite and 478 got one?
- bcambel 14y ago
  755 opened the registration form, 478 completed it.
  - leejw00t354 14y ago
    All you really need is an email. Asking for first name and last name too, even if it's optional, will probably discourage some people when they see the form.
    - bcambel 14y ago
      If they're discouraged because I'm asking their names, than it's fine. There is nothing I can do about that.
      - bmelton 14y ago
        Not to sound contrarian, but the 'something' you can do about that is to not ask for their first and last names. I'm not suggesting that's the right or wrong approach, but you should ask yourself the following questions every time you ask the user to do anything, especially so when asking for their personal information: - Do I really need this? Does the app break if I don't get it? Does the quality of the app diminish? Is there some measurable way in which this impacts me negatively? - Can I rely on this? Can it be verified? If a user enters his name as "Pinocchio Dandelion", can I say that it isn't? At this point, if it IS incorrect, refer back to question #1. By my estimation, generally, if you're asking for names, and it matters, you should probably also be asking for credit card information, or some other way to vet that data. If you aren't expecting to bill them, you probably don't need to ask for it. Of course, you might have something in the app that does need it that we can't see from the outside, but most commonly in these scenarios is that people think they need the information and really don't, which means that you might have captured an even larger percentage of users in the signup funnel. In regards to signups, you really really really want to make it as frictionless as possible. Once you have their email address, you almost certainly have a more 'committed' user than someone who hasn't filled out the form at all. If first and last names are nice for you to have (but not strictly necessary), then you can ask for them later, after you have their email address. Once you have the email address though, that gives you the ability to try and lure them in further. Patio11 (as usual) has more detailed information on this topic, so I'll refer you to his posts as to why having an email address is a really good thing, but suffice it to say, it is.
glimcat 14y ago
"I was afraid that Nginx might not hold up" Seeing articles get nuked by a sudden spike in traffic tends to give one unrealistic assumptions about what servers can handle. Most can take quite a bit if you're serving mostly static content.
mikeyouse 14y ago
Just wanted to let you know, there's a typo on your landing page. "Archieve" should be "Archive". Otherwise, looking good!

12 points by bcambel 14y ago

bmelton 14y ago

I didn't see it this morning, but seeing this now and just signed up for the beta. I'm still not entirely sure what it does, even after reading the linked thread, but I try to do whatever I can to support all the 'Show HN' threads, so even if it sucks, you've got my registration (and I'm not presuming that it will suck.) Nginx is almost certainly not going to be your failing piont, and can certainly scale to beyond 400+ active users, assuming that it's configured correctly and on decent hardware. Your database could crash, or one questionable join could tear the whole thing down, but Nginx is probably the least likely component in your stack to fail. As for the lesson learned, that's spot on, and one that I waffle on every now and again, even though I've done it and know it to be canon. Best of luck. It's a space I've got a lot of experience in, and 'interestingness' vs. 'importance' isn't really gotten right anywhere yet. Hopefully you've nailed it.

bcambel 14y ago
Thanks Barry, interesting vs importance highly personal and it's a hard topic.

maxhacker 14y ago

That is an awesome response from a Show HN! Gives me some motivation to finish my project and show it to the world. Also, I just requested an invite to the your service (didn't catch the actual HN post earlier, but I saw this and was intrigued with your project). Thanks for a good post!!

nuhm 14y ago

"478 people requested invite out of 755." Does this mean 755 requested an invite and 478 got one?

bcambel 14y ago
755 opened the registration form, 478 completed it.
- leejw00t354 14y ago
  All you really need is an email. Asking for first name and last name too, even if it's optional, will probably discourage some people when they see the form.
  - bcambel 14y ago
    If they're discouraged because I'm asking their names, than it's fine. There is nothing I can do about that.
    - bmelton 14y ago
      Not to sound contrarian, but the 'something' you can do about that is to not ask for their first and last names. I'm not suggesting that's the right or wrong approach, but you should ask yourself the following questions every time you ask the user to do anything, especially so when asking for their personal information: - Do I really need this? Does the app break if I don't get it? Does the quality of the app diminish? Is there some measurable way in which this impacts me negatively? - Can I rely on this? Can it be verified? If a user enters his name as "Pinocchio Dandelion", can I say that it isn't? At this point, if it IS incorrect, refer back to question #1. By my estimation, generally, if you're asking for names, and it matters, you should probably also be asking for credit card information, or some other way to vet that data. If you aren't expecting to bill them, you probably don't need to ask for it. Of course, you might have something in the app that does need it that we can't see from the outside, but most commonly in these scenarios is that people think they need the information and really don't, which means that you might have captured an even larger percentage of users in the signup funnel. In regards to signups, you really really really want to make it as frictionless as possible. Once you have their email address, you almost certainly have a more 'committed' user than someone who hasn't filled out the form at all. If first and last names are nice for you to have (but not strictly necessary), then you can ask for them later, after you have their email address. Once you have the email address though, that gives you the ability to try and lure them in further. Patio11 (as usual) has more detailed information on this topic, so I'll refer you to his posts as to why having an email address is a really good thing, but suffice it to say, it is.

glimcat 14y ago

"I was afraid that Nginx might not hold up" Seeing articles get nuked by a sudden spike in traffic tends to give one unrealistic assumptions about what servers can handle. Most can take quite a bit if you're serving mostly static content.

mikeyouse 14y ago

Just wanted to let you know, there's a typo on your landing page. "Archieve" should be "Archive". Otherwise, looking good!