Hacker News Reader

Comment by markkat | Hacker News Reader

http://hubski.com We are in the early stages of a social aggregator. It is built off the HN code in Arc, and the style still looks currently grokked. We are looking at innovative ways to move away from this.

Functionally, the site primarily differs in that you follow people and topics, which builds a custom feed. You can also hub (retweet) posts you like, which refers them to your followers. -Not unlike HN meets twitter. Also we are playing up the submission format a bit so that a user's submissions are more like a blog. When activity permits I envision the frontpage to be replaced with a user's feed. A bit more service than destination.

Looking for design feedback that will inform our next UI changes. Thanks! A cool thing to do.

View on HN

{tl;dr} Define your assumptions. Capture data in an effort to verify those assumptions. Modify experiment or assumptions appropriately. Repeat. {/tl;dr}

There's a great answer I recent read on Quora which may help answer how to approach the design of your site. The thing that made the most impact here is being aware of the market your are entering. (http://www.quora.com/Internet-Startups/Should-I-focus-on-get...)

It's easy to argue which stage/market your product falls into, but I think you agree that the experience of your site is going to be the primary differentiator. (Experience Focus Stage)

Aside from the few tips in that thread, the next thing I would focus on is capturing data to verify your assumptions. Your assumption here are that the blog layout and distribution model is better than what's available elsewhere. (Right? If not, what ARE your assumptions/goals?) Custom analytic tools which let you capture custom events are helpful in verifying these things. You might setup a tool like MixPanel to capture key events your users are experiencing.

For example: Maybe you have a specific flow or funnel you're expecting users to take. Are they completing this flow? Where are they stopping? Are you staying on a specific page for very long?

I would also ask people who have never seen your site before to try using it. Watch what they do and ask them to share their thoughts out loud as they browse the site. Do they get the point of the site? Does your vocabulary used on the site confuse users?

Above all else, make sure you're investing your effort in the right things. You might not need to improve your UI. Your UX (user experience) may be hurting you more than the site's looks. The fact that you have a whole FAQ explaining your site's basic terminology is a bit concerning. (I know many of these things didn't make sense until I read your FAQ. How many users do you think will invest that sort of time?)

Follow up questions are welcome. Let me know if I should elaborate further.

Thanks for those very thoughtful suggestions. I am now going to write our assumptions down, and then set out to test them. I also think you are dead-on about the experience being the primary differentiator. The blog aspect is a completely untested assumption.

As you point out, the terminology is an unnecessary hurdle, -especially considering the space we are in. I am going to look at ways to simplify that as well.

This has given me some very good food for thought. I think that I was putting more thought into UI when UX is just as relevant, if not more so. Thanks a lot. It's much appreciated.

p.s. I see you are on Forrst. I might follow up with you there in some time if you don;t mind.

Of course! I don't log in often on there, so only follow up there if it's not too urgent. @mikegreenberg on twitter is the easiest way to reach me.

Glad it helped! ;)

Reply on news.ycombinator.com