Email

Email is definitely a startup component. Here's an interesting new development in email, though: bad @gmail.com subscribers can downgrade your email reputation - and now that downgrades your Google search results as well! Yikes! This is actually good - it's all part of Google's antispam effort - but it's counterintuitive for people who are following what they thought were the rules for mailings.

Another component list

Here. I oughta have a Wiki - oh yeah! I do! That came to nothing.

Digital currencies

I don't even know if this a component or what, but here's a link to a list of resources for digital currencies, financial crypto, and so on.

Fake it before you make it

An interesting post from Andrew Thomas: the concept of building wireframes as a marketing tool to get out into your target before (or while) you work on the application itself, and generalizations of that concept (e.g. put links for possible extensions into your existing application, then work with the people who click them, etc.).

Upshot: a cookie-cutter wireframe + business process service would rock.

Business patterns

What I really want to end up doing with startups is treat them like programming. That means libraries, components, a language for expressing them, a substrate to run them on, and design patterns. Swombat talks about dark patterns. Think about it. Think harder about it.

Data reconcentration

Here's an interesting post for a small-scale business strategy. Republication of data, living off traffic and adverts. I think this makes a lot of sense, actually - and come up with a cookie-cutter solution and do it over and over, and voila! Income!

And in general, Big Data is turning into a Big Issue. Viz a data man being hired by Cheezburger. It's definitely a place to be going.

Marketing for Hackers

A post I need to read.

Gold farming

By bot. Seriously. The news item about Chinese prisoners gold farming led Charlie Stross to post on it (natch) and one of his commenters dropped a link to a pretty fascinating how-to article (NSFW illustrations).

Point being: one node can earn (right now) something like $2/hour. I don't know how much of that requires any human input at all; responding to chat requests seems to be a bot test. Let's assume that with good programming you could get that to essentially no human input. The article asks, what if you play 12 hours a day every day: $700/month. How about ten nodes? Now you've just built a money machine that you can live off of.

The point being, it would be a good programming project, and I have a hard time thinking that the skills and code you develop for it couldn't be repurposed - and you'd be earning money for free. I feel the same way about pokerbots. Worth thinking about, anyway.

Startup components

Again: not an idea, just a component. I should categorize these at some point.

Recurring billing service provider: PintPay.com.