Petition To Build Death Star To Get White House Response

Death Star

The internet has succeeded in getting 25,000 signatures in the petition to build a Death Star. By practice the White House must now respond to the petition. I am very much amused.

It’s worth noting some Debbie Downer figured out it would cost $852,000,000,000,000,000 in steel alone to build.

I for one look forward to the White House response. If the president were wise, he’d respond himself with a web video.

Hard Drives Still Expensive After Flooding

From Ars:

The impact of that disaster has passed, and supply levels are back to near where they were before last October’s disaster. But while the flood waters have long since receded, drive prices haven’t fallen nearly as much—as InfoWorld’s Woody Leonhard reports, retail hard drive prices are still about 75 percent higher than they were before the flood and show no signs of coming down. And the manufacturers are posting healthy profits as a result.

Nobody is shocked by this right?

Unfortunately for them, this will be short lived. The major HDD manufacturers don’t really have great penetration in the SSD market, which is growing at an amazing rate due to dropping prices and people wanting faster boot times.

Hiking pricing, unwillingness to adapt product lineup to meet demand. I think we know where this business strategy leads.

On Google’s Made In In The U.S.A. Experiment

Google is reportedly building the Nexus Q in the USA. It’s almost assumed these days that electronics are made in China, with a few notable exceptions in Japan or South Korea. However when you think about it, it makes sense. Consumer technology has come full circle.

Open up some electronics made before 1985. With rare exception, they look almost foreign by todays standards. They are spartan in design. Big circuit boards with a few components, some wires to peripheral lights, motors, whatever the device needs to do it’s thing. It’s almost elementary to figure out how it works. Even fixing it is well within reason.

Now open up a modern device like a modern cell phone. It’s generally a single highly concentrated circuit board with a mess of finely placed parts on and around it. Every year they get smaller and more complicated. Until recently.

A curious thing started to happen. Electronics in an attempt to use less power, become cheaper to manufacture, started simplifying their designs. Kind of. They consolidated many of their parts. For example the complex set of layers in a cell phones screen that separated the backlight from LCD panel from the digitizer were made into one slim component. Multiple chips were combined into a system on chip (SoC). Common things like WiFi a Bluetooth which are almost never exclusive were on one chip. Devices became simpler.

Most of these individual components are manufactured through highly automated means. For example those LCD panels are not made by hand, they are made by machines because a high level of precision is needed. No human etches a CPU or any other chip. Even soldering is increasingly machine driven as most use surface mount techniques like ball grid array which would be nearly impossible to do by hand with anywhere near the accuracy or speed needed.

The result is that these components are made increasingly by machines. These components are assembled increasingly by machines. This changes the equation when it comes to deciding where to manufacturer. The biggest advantage to China was very cheap skilled labor. This is changing. First of all China is starting to become more expensive as affluence builds in China. More notably, the need for labor is able to decrease as designs become more machine centric. While the iPhone is very labor intensive today, but don’t expect it to stay that way. It’s increasingly simplifying it’s design despite pushing the limits. Energy costs and shipping costs are also changing.

Google’s bet is that they’ve simplified the human part of manufacturing to the point where the labor costs are becoming minimal. It’s a reasonable bet. If you look closely at the photos in the NY Times piece, you’ll see lots of humans posing, and a few doing actual tasks related to assembling. Of those assembling, there’s no mention on if they are working on prototypes, or if they are in full scale production (which may involve less humans and more automation).

Factory automation will increasingly bring hardware manufacturing back to the US. But don’t expect to see the jobs of the 1980′s coming back with it. These will be highly automated facilities run by engineers and supervised by increasingly limited staff. They may one day operate like data centers.

Was Instagram Worth $1 Billion Dollars?

I think the question “was Instagram worth $1 billion Dollars?” is incorrect. Any form of that question is. The value of any singular thing is in the eye of the buyer, not the market as a whole. The correct question is “was Instagram worth $1 billion Dollars to Facebook?” To that, I think the answer is yes. To any other company, no.

Value is the relative worth of a good. Water is worth nothing when it’s plentiful. It’s priceless when it’s in short supply. This is supply/demand. Companies are slightly different as they aren’t a commodity, there is only one Instagram. The supply is a constant meaning demand is what dictates value. The demand is based on how much another company (or companies) wants to buy it. This decision is based on how well Instagram’s assets, product line, employees, etc. would benefit them. It’s also based on how much it’s worth keeping away from a competitor, and how much it’s worth avoiding having another competitor. Distractions are costly for businesses. Mitigating them has difficult to quantify but significant value.

Product/Value Curve

A basic company works like this: it creates a product with value that it obtained by customers (freemium products make things slightly more abstract, but this still applies). At some point the product runs out of potential customers either because the entire world who wants it has it (think: Facebook, Google), or customers just grow tired of it (think: Yahoo). The way a company keeps up growth is to introduce new products to add value to the company.

In a perfect world every product would achieve maximum potential and a company’s growth would be at worst a diagonal line, consistent and upward moving. or a curve moving upward at an accelerating rate. We know that’s not realistic.

The reality is most companies have lots of overhead when diversifying the product line. Not every product is a winner. The most notable example of this is Google. They have many products, to the point where I bet few if any Googlers know them all. Most however aren’t generating substantial revenue. Sergey Brin has been putting some effort into reducing Google’s product complexities because of this. Google. Apple famously reduced it’s product line when Steve Jobs took back the company and is highly profitable. Facebook has scaled reasonably well as it’s product can essentially be broken down to: communication (direct messaging, broadcasting), photo sharing, contact organizer. This on a graph looks something like this:

Value Curve

Avoid Distractions

Getting back to the topic on hand, what is Instagram worth to Facebook? Facebook needs to keep on or above the line. That means they need to keep rapidly developing and improving products There’s a reason Zuckerberg put an emphasis on “Stay Focused & Keep Shipping“. Instagram was a distraction to Google. It was the only viable competitor to Facebook’s massive photo sharing business.

Flickr stalled long ago, Google can’t figure out Google+ or Picasa. SmugMug et. al. are to small to matter to Facebook. Instagram was the only photo sharing business with growth that could concern Facebook. Avoiding the distraction that Instagram would be is worth a substantial amount to Facebook.

Instagram being bought by someone else could also be a distraction. Keeping it away has a similar value too.

The Future

Facebook has pretty much hit market saturation with its current audience. Facebook is having some trouble in some parts of the world where other social networks have taken hold. To keep growing that means they need to keep growing users. Obviously getting kids to sign up (once they turn 13 thanks to COPPA) is what Facebook needs to keep itself going. Instagram has managed to gain serious traction in the pre-teen market. These are kids who grew up with text messaging and photo sharing. Two of Facebook’s major product offerings. This is a major threat to their future. Why join Facebook and be observed by “friend” parents when you can use your phone (or iPod touch) and remain completely below the radar? Many of them have been doing this for a while now. All they need is WiFi (lots of free SMS apps in App Stores). The incentives to change workflow to Facebook have dropped. Switching over is hard anyway. Just because Child A turns 13 today doesn’t mean all their friends do. Being under 13 on Facebook means fearing being suspended constantly. Facebook purged 20,000 accounts daily in 2011. I’d bet that number is higher now.

These kids have a system of communication that works for them. It’s a very workable system. Facebook isn’t part of that system. That’s a big problem. Buying Instagram means they have a vehicle to convert these new users.

Engineering

Instagram scaled quite well with a dozen employees and managed to build an app that makes photo sharing extremely simple and fun. The app’s workflow for photo sharing makes Facebook look terrible. Buying this up and integrating this is worth a lot to Facebook. Just like FriendFeed gave Facebook some serious engineering chops, this will as well. the Instagram team’s vision and expertise shouldn’t be discounted. Buying a company for its engineers isn’t uncommon these days.

I wouldn’t be surprised if Facebook launches (or rebadges Instagram into) a Messenger like app that focuses on just one of Facebook’s tenets. Instagram’s Facebook integration just needs a little more tweaking and it would work very well. Keeping Instagram communicating with Twitter and Flickr will prevent alienating users.

HP Produces More Of A Discontinued Product

John Gruber questions the point of HP’s decision to do a final run of TouchPad manufacturing. I’ll propose a likely theory:

iSuppli says the Bill of Materials for an iPad 2 (32 GB GSM) is 336.60 when you add in manufacturing. That same iPad retails for $729.00. This is common sense. There’s R&D, marketing, shipping, and of course profit. Keep this in mind. The retail price is not the break even. It’s a profit. HP is selling their Touchpad’s at $99 and $149 I believe, for 16 GB and 32 GB respectively. A loss, but not quite as substantial as comparing to retail pricing would lead you to believe.

Secondly, it’s important to keep in mind that costs aren’t incurred as products are produced. Supply chains often require commitments. HP likely spent considerable funds securing the parts for the Touchpad. They also spent money tooling the factory. This money is already spent. Contracts were signed (they might be able to get out if they pay a penalty + accept some bad will with vendors they may need in the future). Costs that exist regardless of their decision. This is like selling tickets to a sports event you can’t attend at a loss, because it’s better than being stuck with tickets you can’t use and being out 100% of the cost.

I suspect the primary purposes of this last production batch are as follows:

  • HP already incurred the majority of the cost with R&D, parts, etc. Using up the inventory they have is a way to recoup some of these funds, vs. selling back to the vendors or finding other interested parties. Given it’s a mobile device, parts may even have been custom fabricated to meet the specs and confined space.
  • HP wants to preserve it’s relationship with it’s supply chain.
  • HP isn’t giving up on tablets, they are giving up on WebOS tablets. Might as well get some tablets out there and find out how the hardware does in the wild so building v2 with new software can learn from v1. Again, most of the costs were already incurred.
  • HP isn’t (officially) giving up on WebOS, they are just giving up on WebOS tablets. Until they figure out what to do with it, either license to someone, use on other products, spin it off might as well keep the ecosystem alive so it retains some value. HP invested a lot of money in it. HP has almost 600 employees on it. Loosing a little more cash on hardware to keep demand in the ecosystem up for a few months may not be a bad investment.

Overall, it seems surprisingly logical to produce another batch. It costs HP a lot of money to cancel the product so quickly. They are taking the loss regardless. Might as well try and reap some rewards and recoup some cash from it.

Hot Dog Bun Math

It’s the 4th of July weekend here in the US. Today’s BBQ1 got me thinking about that conspiracy theory from Father of the Bride regarding the mismatch of hot dogs quantities and hot dog bun quantities. Steve Martin’s character goes nuts over the discrepancy.

This is really an exercise in Least Common Multiples that no teacher seems to exploit (at least none that I ever had).

Research tells me that the reality of this joke is a little more complicated than 8 hot dogs and 12 buns. It may even vary based on location. From what I can tell the most common hot dog packages are 8 and 10, while the most common bun packages are 10 and 12. That means a 10/10 purchase is a win in terms of efficiency. I suspect there are more combinations, but 8/10 and 10/12 seem to be the most common. Here’s a table of the possibilities:

Hot Dog Qty. Hot Dog Bun Qty. Least Common Multiple Least Hot Dog Packages Least Hot Dog Bun Packages
8 10 40 5 4
8 12 24 4 2
10 10 10 1 1
10 12 60 6 5

This leaves me to question: who profits more from this? To figure this out, we’d need to know buying habits of people and costs involved in producing, packaging, shipping these goods. I don’t have that on hand, but I can draw a pretty graph of how many packages of each you’d need to not waste food:

So it looks like we’ll be eating hot dogs in sandwich bread and making tiny sandwiches out of left over hot dog buns for years to come.

1. Technically you grill hot dogs (hot and fast), not BBQ (low and slow) but American etymology is funny.

More On The Housing Market

Gary Shilling of A. Gary Shilling & Co believes that the housing market has another 20% to go before it will bottom out. BusinessInsider has a pretty extensive slide show explaining how he came to that conclusion.

I noted a few weeks ago that I don’t think the housing market has bottomed out. Despite what various talking heads may suggest, the numbers and historical trends suggest it still has more to drop. It still seems to expensive for there to be enough of a market to warrant the price. This is a cyclical problem. Unless housing becomes more affordable, or potential buyers become more affluent (unlikely) housing prices can’t be stable much less climb in any meaningful way. Eventually prices must drop.

I’m glad to see validation that my head scratching isn’t totally unwarranted. This has bugged me for several months. People claim it’s bottomed out, but there doesn’t seem to be any rational reason to think that other than wishful thinking, which I’d argue is optimistic, rather than rational.

Median Home Prices vs. Median Income

The national median existing-home price in July 2010 was $182,600 [source]. The median household income in 2008 was $52,029 [source]. To account for some rising income and play it conservative we’ll call it $55,000 for 2010. I haven’t found any official number later than 2008. We should also note that this estimation ignores the current recession and its impacts like reduced hours, layoffs, and pay cuts.

The old rule of thumb is the maximum house you can afford is 3X your gross income minus obligations (debts). More than that and you’re odds of running into trouble making payments and affording upkeep/taxes becomes too high. That means the median affordable house is $165,000 if the household had no obligations. We of course live in a world of car loans/leases, $100+ cell phone bills, expensive broadband internet, and rising food prices. According to MSN Money the average household owes $8,000 in credit card debt alone [source].

For a more specific example, New Jersey’s median household income in 2008 was $70,347. Per Capita New Jersey is one of the most affluent states in the country. The median value of homes in 2000 was $170,800 [source]. The median price in the Northeast in July 2010 was $263,800. Assuming the median income rose to $75,000 the maximum affordable house is $226,041, again assuming no other obligations.

I should also note this rule of thumb was before the bubble and before banks became borderline paranoid about lending money.

The home ownership rate in the US is 67.8% [source (xls)] which means it’s a pretty representative part of the US population, hardly a niche group which we could suggest lives outside these numbers.

Using this back of the envelope calculation I propose that the housing bubble for existing homes doesn’t end until either the median existing-home price falls to meet the maximum affordable house index, or the incomes rise and debts fall to meet the existing-home price index. Only then will the supply and demand curve finally meet and the market volume will increase. Personal income and wealth are too tight for either the supply or demand side to cave right now. I further propose that housing prices will fall faster than incomes will rise.

Translation: the housing market hasn’t finished correcting.

Christmas Economics

Jeff Walden linked to a great paper:The Deadweight Loss of Christmas by Joel Waldfogel. Given the economy this several year old paper is extremely relevant and interesting. Something to keep in mind when merchants report their holiday sales statistics.

I remember a discussion on this in college (likely an economics class). It’s a slight misnomer that Christmas is an economic stimulus in the way people think it is.

Continue reading