Patrick McKenzie's 2012 in review

Lots of figures and thoughts.

Patrick has documented the Bingo Card Creator so thoroughly that it ought to be relatively straightforward to describe it in terms of workflow.  Then ... iterate.

Business patterns

A two-part article [one] [two] in Microsoft's Architecture Journal from back in 2004 covering business patterns and how they fit into the design process.  Very valuable.

Interesting service concept

OK, here's a neat little idea: given a site, build a moderately customized quality survey for it and submit it to the Mechanical Turk for a series of cheap evaluations.  This kid built this for his own site, then put the service together as something available for other site owners/builders.

It would be instructive to sketch out the business processes for very simple businesses like this, and provide a kind of "business iteration framework" to cookie-cut not only the site itself but the entire business - automation and all.

Customer/product development

So here's a very interesting article about customer development - outlining an actually rather simple process for finding out the language people use to talk about your service domain, and who those people are.

But let's consider for a moment what that actually means.  When doing customer development, you're kind of understanding a group of people united by a common need - maybe before they even know this.  Your task is to find that need, solve it with something you already have (or, in product development, build something that solves it), and then contact those people in a believable way so that they will also see that their need can be solved.

1. Find the need and the people: customer development
2. Find the solution: product development
3. Communicate the solution: marketing
4. Manage the solution: technical support
5. Manage customers' interactions with the solution: customer support

Each of these areas is kind of a part of any business plan that involves solving needs for money.  Which might be all business, but I'm not prepared to make such a sweeping generalization.

So here's the thing.  How automated can the discovery part of this process be made?  Can I spit out business ideas just by searching yet?

How to start a startup

Some good tips from somebody doing app device for 8 months.

Support tips

Nice article, "How we support hundreds of Web apps".

Startup pitches

A place to start (ha!) with the startup database: TheStartupPitch.com.

Research APIs

FullContact is a personal-information research company that exposes an API.  (e.g. here [HNN]).  This is an excellent basic pattern for a startup.

Blog monetization

Interesting post.  I really need to develop a set of blogging processes that I can drive with minimal effort, based on statistics that are analyzed automatically.

A Google Docs startup idea document

Interesting.

The business case for APIs

Of course, the elements of the business process description are automatically API candidates.  Here are some interesting business-level aspects of APIs.

Cofounders and startup management

Questions to ask potential cofounders. And a neat slideshow on "How to run a 5-whys" - this is apparently a buzzword for a post-mortem.  Which is itself a buzzword, I suppose.

The point is: in a human organization moving at great speed, mistakes simply will be made.  It is essential to understand the mistake and to construct systematic ways of improving - otherwise the mistakes will simply continue to be made.  People are fallible.  "Trying harder" isn't a solution.

The point of the slideshow is that it's imperative to use humor to circumvent people's sense of shame at failing.  Otherwise you won't stop failing.

A practical guide to selling books online

A few pointers to publishing ebooks.

Software maintenance

Actually, "tool maintenance" - you'll be maintaining your processes as well.

A note from one of the authors of BugZilla: suck less over time.  Address issues, know your users' actual pain points, and alleviate them, especially the paper cuts.

Startup registry

I feel like I should start a database.

Blank: open-source entrepreneurship

Steve Blank has a nice list of components for entrepreneurship here.  [hnn]

PG on finding startup ideas

Not that this is new information, but: [here].  And a reaction.

Idea sources

Most of this is kinda useless; some is good.

Sequoia Capital's business plan elements

Another breakdown of business plans, by Sequoia Capital.

How to win as a second mover

More strategy.

Startup Founder's Workbook

Steve Blank has produced a startup founder's manual, with a companion workbook at Zoomstra, here. This workbook consists of a set of checklists with progress markers that would make a dandy first stab at Startup::Declarative.

Zoomstra

Another platform for monetizing content: Zoomstra.

Gabriel Weinberg's flowchart

Startup strategy - codified.

Another dark pattern: Sambreel

Nice strategy: plant browser-resident malware ("Drop Down Deals", etc.) that modifies each page the user sees, then zap the host's ads and show your own. [Atlantic]

CRM for the Web

Open data permits better management of the customer relationship.

Voting record tracker

This isn't so much a startup idea as a "something I could do" idea - given a list of proposed legislation, tag it with the purposes of each bill, then use http://www.govtrack.us/congress/votes to track votes.  Do statistical analysis to cluster Congressmen on like voting (this is probably kinda boring given the Republican lockstep lately).

The reason: Paul Broun (Arkansas-R) says science is Satan's work ([e.g.]) - so how do we find his voting record on science-relevant legislation?

Update: Oh.  We look at votesmart.org, of course.

Research sites / searchable databases

Really, any area of interest can turn into a searchable database.  I should really be able to get this kind of stuff done in my sleep by now.  Specifically, I was thinking of DIY hardware - sourcing things like sensors and so on.

The basic process would be:

1. Get a basic semantic picture of the domain (i.e. what items are of interest? What attributes distinguish them? Who is active in the area? Who blogs about it?)
2. Crawl the Web by means of link following, search term identification, and so on.
3. Find the valuable data sources and determine the database schema that best encodes them.
4. Scrape data on a periodic basis into your database, and make sure you have keyword-rich databases available.
5. Provide some kind of commenting/forum functionality.

There are, of course, lots of this kind of thing out there - they all suck.  A small community of devoted fans can make or break this; you can become the definitive guide to a very small domain.  Monetize with ads and with sales.

Getting Real

I'm reading Getting Real (37signals's ebook) and - while it's written in a very minimal style - it's got some great suggestions.  I'd really like to get started writing some candidate workflows.

Another landing page selector

Here's a code-free landing page builder that gives you the ability to let customers select a pricing model preference during their interest registration.

Dark, dark

SOPA as a ransomware virus.

I should note - again - that I have no intention of writing a ransomware virus or instituting any scam to make money.  It's just that a pattern language for business structures should be able to describe these as edge cases - not only in terms of how they work (both technically and from a business process perspective) but also how they use peoples' expectations to propagate.

Speaking of dark patterns

Here's a great one: Roulette Bot Plus.  It's simple: in return for Paypal donations from your winnings, these guys will give you a free roulette-playing bot for online casinos.  It allegedly takes advantage of the fact that their random number generators aren't actually random, which is pretty darned plausible!

But it turns out it's actually a scam by some of the online casinos.  It'll look like it works for a while, then when you put your actual money into play, it'll go all-in and unfortunately lose.  Ah, well - better luck next time!

On the one hand, I'd love to play with bots of this nature now that I don't live in the States (of course, my bank is still in the States...). But at the same time, this scam fascinates me.  I know there are lists of scam patterns (like dark patterns) but it remains a goal of mine to encode them in a machine-readable and -executable manner. (Yes, I'm working towards humanity's demise.  Ha.)

Leanpub self-publishing

Since I'm kind of embroiled in writing book-like things lately, here's a way to monetize them on a really small-scale basis.  Worth a shot.

Business plans

This was the post to SP that made me think I was getting outside the box over there: an outline of a standard business plan.

Repurposing, a little

It's been bothering me for a while now that my Semantic Programming blog has some startup-type information in it.  So I'm moving all that over here - back dated to its original date of publication.

The idea of most of that is Startup::Declarative, a semantic domain for my Decl language that describes the structures of a startup business, its business processes, actions to be taken, strategies, and so on.  A full declarative specification for a business is, in fact, its business plan - in machine-readable form.

Since this blog is about startup ideas, I've decided it should be about startups in general. I may end up specifying some of the ideas formally or something, if it should ever get that far.

How to Build a Web Startup

Steve Blank has another distillation of the startup process.  I love this stuff!  It's like popcorn.

Business model development

The NY Times just published a fascinating article on a guy who sold reviews for ebooks on Amazon, B&N, and other venues.  It didn't last long (goes against Amazon's ToS) but at its peak he was drawing $28,000 a month.

I love reading about business hacks like that.

Here's an exercise for the reader (and for me, should I ever get the time) - design the business processes for that business, including any interface code for the various roles (reviewer, author, etc.) and the interface code to Amazon for submission and tracking of reviews.  (That latter should be done with Javascript in the reviewer's own browser.)

It should be brain dead easy to design and iterate a business of this nature.  Less than a day, at most, between idea and execution.  That's my goal.

Update: just realized I saw the same story in an NLP context last week: textual analysis to determine whether a review is false or not, with the same usual counterintuitive linguistic markers.

Adapting price points for SaaS

Patrick Mackenzie again, this time with an article about adapting price points in software as a service.  He's all about the autostartup.