France Telecom CEO Leaks iPhone 5 Ship Date: October 15

“If we believe what we have been told, the iPhone 5 will be released on 15 October,” Mr. Richard told a group of reporters, accord­ing to TechCrunch. He didn’t say if that date was for the iPhone’s world-wide launch, or if it could roll out in the United States ear­lier in the month.

Mr. Richard is the same CEO that let slip Apple was plan­ning on using micro SIM cards in the iPhone 4, which does add a lit­tle more cred­i­bil­ity to his iPhone 5 launch date.

Why does any­one tell this guy anything?

Chinese Inmates Forced to Farm Gold Online

Liu says he was one of scores of pris­on­ers forced to play online games to build up cred­its that prison guards would then trade for real money. The 54-year-old, a for­mer prison guard who was jailed for three years in 2004 for “ille­gally peti­tion­ing” the cen­tral gov­ern­ment about cor­rup­tion in his home­town, reck­ons the oper­a­tion was even more lucra­tive than the phys­i­cal labour that pris­on­ers were also forced to do.

“Prison bosses made more money forc­ing inmates to play games than they do forc­ing peo­ple to do man­ual labour,” Liu told the Guardian. “There were 300 pris­on­ers forced to play games. We worked 12-hour shifts in the camp. I heard them say they could earn 5,000–6,000rmb [£470–570] a day. We didn’t see any of the money. The com­put­ers were never turned off.”

Will this spawn a fair trade vir­tual goods economy?

 

Why the DMV Website Sucks

There is a lot of right­ful com­plain­ing about inef­fi­ciency of gov­ern­ment con­tract­ing. There exists no short­age of rea­sons why they’ll over­pay for things: it’s not their money, crony­ism, nepo­tism, lob­by­ing bribes. But why are the soft­ware prod­ucts so ter­ri­ble? Why don’t the $550 mil­lion dol­lar data­base ini­tia­tives yield soft­ware that can do sim­ple things like, uh, search the database?

Why is the DMV web­site so bad that it makes me want to go to the DMV? (Nothing else in the world can do that.)

It’s not because they’re under­pay­ing. When it comes to gov­ern­ment work, the deck is stacked against good soft­ware prod­ucts because the bid­ding process is bro­ken. And when it comes to find­ing and hir­ing good devel­op­ers it breaks in a par­tic­u­larly insid­i­ous way.

Please Submit in Septuplicate

One of my old jobs was as a Marketing Assistant for a large gen­eral con­trac­tor. A big part of my job was putting together pro­pos­als in response to Request For Proposals (RFP’s) by com­pa­nies who wanted to build a sta­dium, hotel, casino, etc…

We also did gov­ern­ment jobs. I knew right when they came in because the envelopes were three or four times as thick as is typ­i­cal. I dreaded them.

A large pro­posal for some­thing like a 14-floor casino/hotel, would be assem­bled into three four-inch ring binders for pre­sen­ta­tion. They would then be sent along with an elec­tronic copy on disc to the entity who issued the RFP. It was a lot of work, but when your ask­ing some­one to pay you 20–150 mil­lion dol­lars that’s not unexpected.

The com­pleted gov­ern­ment pro­pos­als, on the other hand, could  con­sist of between 15 and 30 (not a typo) ring binders. It would take me 15 min­utes just to load it into FedEx via hand truck from my car. And this would be for some­thing like a mil­i­tary bar­racks, I’m not talk­ing about sky­scrap­ers or sta­di­ums. Not to men­tion that it was much more com­mon with gov­ern­ment jobs to have sev­eral rounds of RFP’s.

Firstly, it was com­mon for gov­ern­ment RFP’s to request five to nine hard copies of a response, where two or three was typ­i­cal of a pri­vate RFP. On top of the bulk of the mate­r­ial we had to pro­duce, they also had to be fas­tid­i­ously put together in weird phys­i­cal for­mats with cer­tain labels at cer­tain exact coor­di­nates at the begin­ning and end of each sec­tion among oth­ers. It was also typ­i­cal for elec­tronic copies to be requested in soft­ware for­mats that were 10 years old.

Software and web devel­op­ment isn’t con­struc­tion, but I’m will­ing to bet the process of bid­ding is sim­i­lar. There is a lot of unnec­es­sary time and effort that goes into bid­ding the gov­ern­ment. (‘Helps if you know a guy, too, natch.) So when New York pays a con­sul­tant fee of $722 mil­lion dol­lars for time clock soft­ware that’s seven years behind sched­ule and doesn’t work… I think it’s a gross waste of tax­payer assets, but it doesn’t sur­prise me.

Filtering Out Greatness

A focus on obnox­ious minu­tiae and sep­tu­plet hard copy requests in and of itself can­not inflate project costs seven to 10 times. But bad programers do.

This whole process is dou­bly sin­is­ter for soft­ware in that it’s going to be par­tic­u­larly dis­taste­ful to just the sort of per­son that would make a great developer. Good hack­ers see prob­lems and are dri­ven to solve them. They are mad­dened by sys­temic inef­fi­ciency and seek to cor­rect it. A bid­ding process like this is so point­lessly waste­ful and time con­sum­ing that it weeds out the exact sort of per­son you want work­ing on your tech­ni­cal prob­lems. Which is how you end up with ter­ri­ble city gov­ern­ment sites and on a larger scale CIA, FBI and DHSA data­bases that are unsearch­able, and won’t talk to each other.

Of course sales and mark­t­ing isn’t doing the hack­ing on projects of this size. Inside any com­pany that can suc­cess­fully engage in this byzan­tine slog that is Government con­tract bidding, there must exist a cul­ture of quite accep­tance of delib­er­ate medi­oc­rity. No mat­ter what depart­ment it lives in, cul­ture is noth­ing if not a dis­ease. It’s infec­tious, and this sort of cul­ture is the antithe­sis of what makes great devel­op­ers great.

There’s been a lot writ­ten on exactly what makes devel­op­ers great, so I’ll quote Paul Graham’s essay Great Hackers:

It’s pretty easy to say what kinds of prob­lems are not inter­est­ing: those where instead of solv­ing a few big, clear, prob­lems, you have to solve a lot of nasty lit­tle ones. One of the worst kinds of projects is writ­ing an inter­face to a piece of soft­ware that’s full of bugs. Another is when you have to cus­tomize some­thing for an indi­vid­ual client’s com­plex and ill-defined needs. To hack­ers these kinds of projects are the death of a thou­sand cuts.

The dis­tin­guish­ing fea­ture of nasty lit­tle prob­lems is that you don’t learn any­thing from them. Writing a com­piler is inter­est­ing because it teaches you what a com­piler is. But writ­ing an inter­face to a buggy piece of soft­ware doesn’t teach you any­thing, because the bugs are ran­dom. [3] So it’s not just fas­tid­i­ous­ness that makes good hack­ers avoid nasty lit­tle prob­lems. It’s more a ques­tion of self-preservation. Working on nasty lit­tle prob­lems makes you stu­pid. Good hack­ers avoid it for the same rea­son mod­els avoid cheeseburgers.

What about the gov­ern­ment pro­posal process, rid­dled with nasty lit­tle bor­ing prob­lems, is going to attract great tal­ent to work on our country’s infra­struc­ture? Nothing.

The $400,000 dol­lars per year all those New York city time clock con­sul­tants were get­ting paid isn’t just for prod­uct devel­op­ment. $300,000 of it is to com­pen­sate for nav­i­gat­ing the night­mar­ish gov­ern­ment bid­ding process laced with pro­ce­dural road blocks to new­com­ers, crony­ism and incom­pe­tence. But throw­ing money at the prob­lem doesn’t work.

People who can write func­tional soft­ware would find all this dis­taste­ful, much less the good ones. They aren’t going to be spit­ting out ele­gant code and cre­at­ing great user expe­ri­ence. If they work at a com­pany with even a pocket of cul­ture some­where in the build­ing that can not only put up with this bull­shit but is good enough at it to win the con­tracts, it’s a poi­soned well.

I’m not hold­ing my breath, but the best way to fix this would be to scrap the clunky bid­ding process and replace it with one that’s an inter­est­ing prob­lem to solve.

Then again, they’ll never be able to hire the per­son who could do it.

 

Noah Kagan's Original Mint.com Marketing Plan

Aaron didn’t want to hire Noah at first, and told him no, so he put this together and got him to yes. I’m very happy he did.

I’ve always been fas­ci­nated with Mint​.com. They are a com­pany that truly lives or dies by the qual­ity of their UI. Mint​.com itself is just an inter­face for Yodlee. By adding a good UI to the exist­ing Yodlee prod­uct, Mint gen­er­ated enough value in a two year period for Intuit to buy them for $170 million.

Obviously a good UX that doesn’t get used is worth­less, so mar­ket­ing is impor­tant. This is how they got started.

Wufoo Buys SurveyMonkey for $35 MIllion

While the terms of the trans­ac­tion for the Tampa, Fla.-based Infinity Box–makers of Wufoo–were not dis­closed, sources said the price was $35 mil­lion in cash and stock.

Maybe they can put their heads together and come up with a name that the for­tune 500 will take seri­ously. (I kid, I kid — con­grats to the WuFoo team, they’ve been putting out a killer prod­uct for years.)

Social Media Ad Revenue Will Never Match Search

Everyone is wait­ing for Facebook or MySpace to start turn­ing out ad rev­enue like Google. It is not going happen.

In 2007, Microsoft (in)famously val­ued Facebook at $15 Billion, or $323 per user. This was at a time when their annual rev­enue was $0.73 cents per user, plac­ing Facebook’s pre­sumed reten­tion rate at 100% and their aver­age user life span right around 400 years. (Oops)

Expectations have cooled a bit since then, but not by much. And that’s bad news for social media hope­fuls. Search-engine-like ad rev­enues are not on the hori­zon for any of the social net­works for one rea­son: Search engine mar­ket­ing ROI can­not be beat by a social network.

Yes, these web 2.0 giants have had expo­nen­tial growth. Yes they have mil­lions of eye­balls and lots of mind­share online. But you can’t assess the value of mind­share with­out think­ing about what state all those minds are actu­ally in.

People vis­it­ing Facebook or MySpace are there to con­nect with other peo­ple. A social net­work­ing web­site is itself an end. It’s not a means. People on these sites have reached their des­ti­na­tion. They’ve no momen­tum going that will push them to leave by click­ing on an ad. This leaves the momen­tum prob­lem up to the adver­tiser to solve. No mat­ter how you slice it, gen­er­at­ing momen­tum is HARD. (It’s fun­da­men­tal physics, and in this case the metaphor keeps on delivering.)

That’s what makes search engine mar­ket­ing so pow­er­ful. People vis­it­ing a search engine have come there specif­i­cally to leave and find some­thing else. They have momen­tum. They just need a shove and they’re off. Add in the fact that their search terms or key­words pro­vide mar­keters a con­text for exactly what kind of shove is needed, and you end up with a mar­ket­ing envi­ron­ment that may be impos­si­ble to beat when it comes to value for adver­tis­ers. The money will go to search engine mar­ket­ing, because in the long run it’s a mat­ter of ROI for advertisers.

UPDATE: This was orig­i­nally writ­ten two years ago. Right now, Facebook is val­u­ated at around $41 bil­lion. Google’s mar­ket cap is hover around $190 bil­lion. Will they catch up? They might, but it won’t be because of adver­tis­ing rev­enue. It will be because Facebook is sit­ting on a data mine like no one else… except maybe Google.

Small stuff = work that has zero chance of being mentioned in your obituary.

What’s “small stuff?” Roughly, work that has zero chance of being men­tioned in your obituary.

Fantastic quote from Paul Graham’s December 2005 essay, Good and Bad Procrastination. It may be the best one-liner I’ve seen in his entire col­lec­tion of essays. (and that’s say­ing a lot)