Standard economics. As you start penetrating potential audience its going to get saturated after some point until you reach all of the audience. As impressions to get one more click are needed more, google needs to exhaust their inventory more. Apparently CTR is going to degrade, and google will start increasing marginal cost per click. Cost per click is curve is often going to be marginal cost curve, where maximum profitability is at lowest point on the curve. Its not about profitability of google adwords but the profitability of the media you're bidding for. When saturated you can always expand your potential market (audience) by adding more keywords/placements.
I tried logging in -- persona sent me email of confirmation, when clicked it redirected to ebookglue which alerted 'login attempt failed'. you might want to look into that.
I'm sorry about the trouble with Persona -- many people have been reporting issues, and it's unfortunately not been as stable as I would have wanted it to be. I'm looking into whether it's my own implementation (though I followed all the recommended best practices quite closely), or an issue with Persona itself.
Unfortunately, this is a situation where a lot of people on HN probably disable third party cookies (though perhaps a larger percentage of HN readers already have Persona accounts?). Anyway, I'm looking into rolling out my own authentication with email/password soon.
I'd like to try it without logging in at all. The first question everyone will have is if it works on their own site or a site they are familiar with and to see the results. Having to create accounts/login just impedes that process.
I understand -- there's a bit of a problem when trying to prevent abuse, though, and relying on the Persona identity provider makes it easier to prevent abusive behavior upfront, especially since the conversions themselves consume a lot of resources.
I'm still working out the optimal solution, but thanks for the feedback!
In what way does Persona prevent abuse? All you validate is email receipt which can be trivially fudged using mailinator.
It is far better to deal with this sort of thing at the "transaction" level - eg use a captcha or similar for the second and subsequent conversions from a particular IP address.
It's just a rough barrier before I set up something a little better, though I already have logging at the transaction level set up. Even using Mailinator requires a little bit of work, so I figured Persona was a good starting point -- I'll look into adding a "try it out" area on the home page that doesn't require you to sign up though.
yes, there are lots of publications that write quality articles but don't have place to weight them based on community opinions. every article has comments, but those wouldn't be in order of priority of reading. also business news are big enough so that every publication eventually covers it, what remains is the community insights about the news.
that is why I like hn -- because of community opinions than content that is posted.
What else do you call a financial application which hasn't set its cornerstone as data protection.
I saw this app and thought it looked nice, I could see myself using it. I then read a bit more about it in the FAQ and I cannot use it. My financial data may not send the world into an economic downturn but it is something I care about and what to keep safe. If this application cannot do that then I cannot put my financial details into it...
If this was sorted out the app would look very promising. Such an apparent disregard for data protection isn't acceptable for a financial app and in its current state looks like it shouldn't be used.
building app has started to become commodity, most of these are built over the weekend and left after that. What we need to understand is that user acquisition and retention are the most critical factors of the startups in this era. I don't want to go into negative but every hacker knows that we should build something that solves pain for the users and try to monetize that. And every weekend project submitted here seems to be solving some kind of problem at least for that particular niche market. But to grow such apps into business one needs to work on user acquisition, retention, building credibility and customer loyalty which can not be commoditized. These are completely different aspects of business than writing code and should be taken as serious and aggressive as writing code.
I think lean methodology goes best with this. What matters in initial stages is working prototype, validation and then MVP. Most of the initial time is supposed to be dedicated to this than spending it with operations.
