huwshimi

User signup/acquisition analytics for SaaS startups. Always open to entertaining offers from someone excited about the user lifecycle space.

How to make that work for non SaaS startups, I'm not yet sure.

This actually inspired my current project, which is all about measuring the results of your marketing (http://flockmetrics.com).

I think several things are helpful here. Understanding the key benefits of your product and figuring out who will most benefit from those features.

Once you know that you need to figure out how to explain your product in a way that those people can understand how it helps them and then you need to find those people.

I personally have found once I've been through that process and can see that I have a product that will make peoples lives easier it motives me to want to get the product to those people.

I contacted people I knew personally that were involved in startups.

I asked a friend who was involved in a startup incubator to email the incubator's founders mailing list asking for beta testers.

I posted on the @flockmetrics Twitter feed and my own personal one.

I posted the site on http://betalist.com.

I cold contacted some people I thought would especially find it useful.

In the end I found it much easier to get feedback from those I had up front personal contact with.

Edit: he also just suggested https://wit.ai/ which is now owned by Facebook.

I know of sourced.tech as they contacted me about a position recently, but the job they contacted me about wasn't relevant enough to my skills and experience.

Your competitive advantage will probably be in the quality of the target lists you create. If you can nail that on a broad scale you'll have something with a lot of potential (this is probably also the hardest problem).

I'd consider starting by running some experiments for how to create relevant target lists.

Feedback welcome!

They're making a portable high res multi-monitor setup that looks like it might suit you.

I think the new homepage has lost a bit of the personality that Twitter once had and I think they'll find that people have trouble understanding what exactly the bar of user avatars is all about.

HN is self moderating (or rather community moderated). If people don't like certain posts they won't vote for them. If HN starts banning websites because some people don't like them then it takes the power away from the community and then you are on a slippery slope of moderated content and a world of approval conflicts and power struggles.

You or I might not agree with every post on HN but we have the power to not vote, or if you've contributed enough to downvote.

Of course there are always exceptions. Spam and illegal content are examples of this. But then we have the ability to flag those posts.

A moderated HN is one I would not use.

It has probably the best set of node editing tools of any vector editor I've used. I now use it exclusively for illustration, web design etc.

One thing you didn't mention with the two companies is possibility for promotion. If you don't know I'd recommend talking to each company about room for growth and opportunities to take on different roles.

The blog post that thread links to is worth a read.

I personally do some Django development on a netbook and have found it excellent, especially while commuting. I have not had a problem with being under powered (I don't know how Android development would compare) and the only downside has been the lack of screen realestate. Thankfully I bought a netbook with a reasonable keyboard size (Asus Eee PC S101).

I mean that the developer needs to set up rules for what gets displayed for the A/B test and then track the results. I guess you could create some kind of API they could hook into for this.

I may not notice replies here, so feel free to email me about this (email in profile).

There really is no such thing as the right language, but you can find a language (and framework) that suits you well.

For me the language and framework that I work the best with is Python and Django (with jQuery for frontend).

I would recommend picking a couple of languages and frameworks and try building a small project (or even a prototype of your app) in each one. That way you'll probably figure out what works for you and if the framework is going to meet your needs.

You seem hesitant to just pick something and go for it, so this way at least you can try a few thing out before you launch into it and find out halfway through that you don't like the way things work. It might seem like it will take a bit longer this way, but buy jumping in with one language and framework it'll only be quicker if you fluke it and happen to pick something you really like.

You should always have absolute minimum number of fields required to start using the service.

Don't have any optional fields (there are plenty of better places to ask for that information).

If there is information required to use certain parts of the service then ask for it when they want to use that part.

http://tumblr.com and http://posterous.com examples of this. I'm actually a fan of the single password input too... I implemented a similar form recently: http://hellohype.com/accounts/register, only with a username field instead of a url field.

I'm not a big fan of going to the extreme of Instapaper though. They don't require a password at all, so a malicious user can access and modify another user's account.

It might be worth trying to get in touch with someone like Patrick McKenzie. His username here is patio11 and his blog is http://www.kalzumeus.com. He's done a lot of writing about A/B testing and has even released a testing framework: http://www.bingocardcreator.com/abingo. I'm sure he has a lot of thoughts that would be very beneficial (I'm sure there are other people around here that could impart their knowledge too).

The nature of A/B testing often requires things to be pretty tightly coupled with the app itself. If you're providing a service for this it'd be interesting to see how you overcome that.