Being #2 on the Hacker News front page - HackerThings Launch Week | Hacker News Reader

Being #2 on the Hacker News front page - HackerThings Launch Week

hackerthings.com

57 pointscoderdude28 comments

I'm one of the 114 RSS subscribers, and I've been checking the feed every day - congratulations on your rapid success.

I really like the idea of full comment threads attached to each product - at best, other sites have heavily-edited reviews pages.

But please consider using something other than Facebook for commenting. I didn't even see the comment threads when I visited with my main browser, because I have Facebook's domains blocked for non-Facebook sites. Considering your target audience, I doubt I'm alone in this.

Glad to have you as a subscriber!

I understand that some people don't jibe with Facebook and want me to move away from FB comments (I've received several requests). If I do that then I'll have to write an auth system. Not difficult to do but I'd rather focus my time on content right now. Eventually I plan on implementing an auth system and including Twitter and Facebook logins. Once I have that I'll probably move to an in-house commenting system. I can always export the comment data using Facebook's API later to keep the existing comments.

As far as the audience goes, I know you're not alone in this but the people who have boycotted Facebook aren't the majority of users (even HN users). They're just a rather vocal group. ;)

For now I think I've got most users covered for their commenting needs.

Why not use Disqus? It supports Facebook logins, but also Google, Twitter, and generic OpenID.

Disqus is fine I suppose. I've used it in the past but didn't feel the urge to use them again. One thing I really like about the Facebook comment system is when someone posts a comment to a product or in the blog it posts that comment on their wall with a link back to the page. It has helped with generating traffic. Certainly a feature I will try to add for the in-house comments (of course though, that would require extra permissions).

Why don't you simply use Disqus? It should be as easy as FB and much hacker friendlier... I for one can't stand the fact that FB comments don't zoom.

Have you looked at Disqus? Provides the same pre-built commenting system, and you can log in with whatever service you want. http://disqus.com

What do people think of deleting an ignored HN submission and re-submitting in the hopes of triggering lots of voting the second (or third) time around?

I've thought of doing it once or twice but came to the conclusion that it would be kinda unethical. Is it standard practice?

I don't mean to specifically bash coderdude but he mentioned it in his blog post so it made me think to ask.

Bash away. ;)

It certainly shows that the right people on HN at the right time will make or break a submission. It's almost entirely the luck of the draw. With that in mind I didn't feel bad about deleting a failed submission after 30 minutes of no attention and trying again at a later time in the day.

Don't feel bad even if it was unethical. It worked. End of story.

Yeah definitely. It's probably just not as big a deal as it appeared in my mind.

* goes off to resubmit everything...

Great stuff, I think it's a solid niche.

Won't it get tough for you eventually to continually source new products though? I would definitely explore a Pinterest / TheFancy crowdsourced curation model. Even if you only let in an exclusive set of curators...maybe you require they have > 10000 HN points or something...

We've built a bookmarklet for our product and it's great for letting people source content from around the web for you.

Anyways, kudos and thanks for the intro to StumbleUpon Paid Discovery, I plan on checking it out right now.

It may become too much for me but that'll be a problem I hope to have. I'm planning on adding user accounts so user submissions isn't much of a stretch from there. I think keeping up with price changes and availability is the only real growing pain.

Yeah, I'm just saying if you get into the hardcore community, you can probably identify people who would be happy to curate the list for you, even if just to have their names be the chosen few on an about page.

Just build them a bookmarklet tool, or give them admin auth to the creation page. I think the cool thing that you have going for you is the concentration of cool shit you've found. You definitely want to keep products somewhat scarce so that they're still compelling once you've discovered them for users.

There's probably an efficient number here...maybe people only expect 1 new product a day / every 2 days. Your challenge will become hitting enough niche sites to find the coolest stuff, but you dont want too much crap so I think opening it up publicly might be a mistake.

I was more interested in how much money did he make, but he wasn't talking about it after all in the blog.

I haven't made any money from it. Haven't even tried to. The project started as just something fun to work on and grow.

Wow... what a pity though. It would be well earned money. On the flip side, if you actually had to find deals with resellers before launching, it would for sure have taken much much longer.

Release early and often, so they say. There will always be time to make money. If the site gets big then I will have ample opportunity to make some money. If the site were to die out then I wouldn't care if I only made a little money on the side. It wouldn't make much of a difference in the long run. So for now I'm focusing on what will get the site to that next level.

Indeed... good luck!

Thanks!

A question for coderdude: how well did the traffic from StumbleUpon perform? Did it almost entirely bounce straight away, or was there a lot of clicking through to other pages?

Out of 132,000 Stumblers the average pages viewed was 1.42 and the average time on the site was 46 seconds. The bounce rate was 80.93% but it's hard to tell how GA is calculating that since I have so many off-site links. All in all I wouldn't call SU traffic very "sticky" but some of it did convert well (became subscribers or shared the site in other channels).

I didn't know about paid SU... i'm going to give it a try with one of my projects, www.plansharing.com. I've tried Facebook ads and Google AdWords. Neither of which has been very good. Facebook REALY doesn't want to send you off site, and mostly the google ads were run on spam trap websites in my experience.

If SU leads to organic traffic too, that's an interesting alternative.

Congrats man! Glad you persisted with the submissions and made it fly.

Thanks!

What is your stack like? How many servers, what back-end technology, amount of memory, processor, etc?

1 server, shared hosting, Apache + mod_wsgi + web.py, MySQL. 160mb of RAM. Not sure about the other system specs on that host. It held up because I used a CDN for all the assets and because I generated static pages. I did have to monitor it though and increase the RAM at one point.

What CDN did you use? CloudFlare?

Amazon S3 (which ended up costing me more than my hosting, by the way).

Reply on news.ycombinator.com