Archive | June 2015

No more slack for SpaceX


Millions are still rooting for SpaceX and Elon Musk, even after Sunday’s launch explosion of the Falcon 9.1, and are willing to cut them some slack. But not everyone will think that way. Yesterday, one of my friends took me to task for pointing out on a Facebook post that the SpaceX launcher had fallen from 2nd most reliable launch platform to 7th, with a 93% success rate:

“That he [Elon] got this far in such little time at a fraction of the cost is reason enough to cut SpaceX some slack.”

In the world of social media, or even traditional media coverage, Mr. Musk has slack coming out of his ears. Coverage and conversations online have been wildly positive and supportive. The story of the successful outsider who shakes things up and disrupts a stagnant and bureaucratic industry is a compelling narrative. It isn’t just Iron Man and Tony Stark; this kind of story has had Americans cheering for the ‘underdog’ for a century now.

But for SpaceX, everything changed yesterday. This isn’t like having trouble recovering the first stage of the launcher by landing it tail first on a barge: although SpaceX kept failing to achieve that goal, it didn’t really affect customers. If SpaceX can re-use first stages then their business model looks better, but if they don’t, the folks who pay Musk and company millions of dollars don’t care. The only thing that matters to those customers is getting stuff into orbit.

There may be things about building or launching rockets that a relatively new entrant like SpaceX can do much faster and leaner than the old guard. But launch failure investigations aren’t one of them.

Take a look at the video from yesterday. There’s no fibre optic cable running back terabytes of usable information: the telemetry from a rocket under boost is not much to go on. No black box would survive that explosion. There will be some debris recovered, but after blowing up at 50 km altitude, there may not be much.

It will likely take days to weeks to determine the cause of the explosion. Not because of bureaucracy, but because it is hard! Depending on the results, it will likely take weeks to months for SpaceX to come up with compelling fix. They may even need to have trial launches with no payloads to demonstrate the fix. By the way, that isn’t my skeptical timeline: SpaceX President Gwynne Shotwell said the Falcon booster will be grounded for “a number of months or so” in a press conference on Sunday. Shotwell added that the delay wouldn’t be as long as a year, but there’s no way to actually know that. The Russian Proton M (pictured below, in full explosion mode) rocket had a failure back in 1988…and it took them 27 years before they finally figured out what the problem was!


Next, something that has been a big tailwind for SpaceX is about to become a headwind. For years now they have been winning launch contracts by being cheaper – under $1,000 per pound to orbit. SpaceX has said that they were able to achieve those lower costs by a combination of manufacturing efficiency, lower operational costs, and high efficiency performance in flight. All of which has been accepted and accounts for much of SpaceX’s success. But after a rocket blows up, questions will be raised about whether some of the cost savings has forced compromises around safety or reliability. That may not be the case…but SpaceX now needs to prove that to what will certainly be a more-sceptical customer base.

Just to repeat that point. They don’t just need to ASSERT that safety has not been compromised to save money, they need to PROVE it. Which is much harder and will take more time than many think.

“Hey! That’s not fair! Elon is trying to ‘disrupt’ the space industry. Why can’t we just cut him some slack?”

Because the paying customers say so.  Here’s the money quote:

“I’m not going to stand up and put a billion dollar satellite on top of a rocket I don’t know is going to work.”

Until SpaceX knows what went wrong, has proposed and implemented a fix, and shown to people that the fix works, neither customers nor insurance companies are going to be comfortable entrusting their cargo to the company. The space business is bad that way – a single failure after tens of successful missions is enough to derail a program for months or years. The Challenger Space Shuttle disaster of 1986 occurred after twice as many successes as the Dragon 9.1 had achieved until yesterday.

But I want to close by discussing why I really don’t want to cut SpaceX some slack:

They got lucky.

Yesterday’s loss of cargo was annoying and expensive, but not the end of the world. The International Space Station is down to four months of supplies, instead of the six months they normally maintain. But another rocket is on its way July 3.

What if they were down to less than a month and this had happened? What if something BAD had happened to the ISS on Thursday and the global space industry had worked for 48 hours straight to get a critical component ready for the Sunday launch to save lives? Could the sighs of relief that the Falcon was already scheduled for a timely launch have turned into tragedy and tears following a failure?

Any launch can fail, and I am not suggesting that SpaceX should be held to some higher standard or perfection. But I am comfortable in saying that they need to be held to the same standard as everyone else who tries to put a critical payload on top of thousands of kilos of fireworks, and lights a match.

No slack in space.


Nielsen media chart tells the wrong story

Nielsen Original

The Nielsen Total Audience Report is essential reading for those trying to predict the future of the TV and media industries. So I was shocked when I ran the numbers from the Q1 2015 report, released on Tuesday, and saw that one of their charts tells a profoundly incorrect story.

The media coverage has focused on the graph above, which at first glance seems simple. Americans 18-34 watch less TV than other age groups, and across all five major platforms (TV, radio, PC, smartphone and tablet) they also spend less time consuming media each week. True enough. But the chart also seems to show that 35-49 year olds spend less device time than average, and the column for those over 50 is about the same height as for all adults.

As I was crunching the numbers, things started to look weird. You have to convert the hours and minutes watched to hours in decimal form, and add each of the five categories together, but the results do NOT match the picture that the Nielsen chart portrays. The youngest age group does have the lowest total usage at 51.7 hours per week. The average adult is 65.1 total hours, while 35-49 year olds are 67.4 weekly hours. I have redone the chart TO SCALE, and added gridlines, below. I tried to keep the same colours and so on as Nielsen:

Revised Nielsen

Although the chart shows Gen X as noticeably lower than all adults, they are in fact higher. Next, the population 50+ isn’t about the same level as all adults, they are using these devices 73.7 hours per week.

You might think I am making a mountain out of a molehill, but that mis-depiction distorts some important facts. The first is that although 18-34 year olds spend about 13 hours less on the five devices than average, those 50+ are about 9 hours more. This a critical truth about media measurement: young people spend more time studying, socialising, and being outdoors while older people are more sedentary, and spend more time on various media devices. That’s not a new trend – it has been true for decades.

Now that we’ve established that, take a look at the chart below. It is a stacked chart, where each category on the X axis will add up to 100%. It hides the fact that 18-34 year olds spend less time on devices, but it does correctly show how they allocate that time. Of the total time spent on the five devices, younger Americans spend about 42% of their “device time” watching TV. That is lower than the average for all adults, but it’s not terribly different than the 49% for 35-49 year olds. The original Nielsen chart doesn’t show that relationship properly.

Nielsen percentages

Don’t get me wrong: younger Americans watch less TV than older Americans, and less than they did three years ago (about 25% less, in fact.) And they will likely watch less in the future. But for now, the “generation gap” between millennials and Generation X is much narrower than you might think.

[Edited. I had a wonderful phone chat with the folks from Nielsen. Other people noticed the error too, and they have already redone the chart. If you download the report, you now see the version below. Kudos to Nielsen for the quick response, and obviously caring enough to do it EXACTLY right.]

Nielsen Chart Revised

Does “Generation Rent” mean the end of banks?


The TechCrunch article was headlined “Millennials Are Destroying Banks, And It’s The Banks’ Fault” and the first sentence was “Millennials are rejecting home ownership across the land.” The article went on to say how millennials want to see banks “completely destroyed.”

Based on some research I’ve done, this is not an accurate statement of generational attitudes: most millennials are dealing with banks now, and expect to deal with them in the future. They are indeed sometimes different in what they want from their financial service institutions, but those differences are more complicated than you might think based on the media stories around the demographic trends.

Let’s take a look at the home ownership part of the argument. Mortgages are a big business in the USA: about $1.2 trillion per year of origination annually as of 2014. If millennials always rent and never buy, that would be a profound change for banks and other lenders.

Have 18-34 year olds really stopped buying homes? No – my oldest daughter’s boyfriend just bought a condo, and he’s not even 25. Our neighbour’s daughter got married last May, and she and Dan (32) just bought a place.

Are Josh and Dan the only ones buying? No – according to the data, nearly four out of ten (38%) 25-34 year olds in the US owned their own homes as of 2012.

OK, but that’s lower than in the past, right? Yes – in 1980, just over half (52%) of 25-34 year olds were homeowners. That demographic has therefore  seen ownership decline by over a quarter in 30 years.

Aha! So this is a terrible thing for banks? Probably not. Home ownership rates vary over time, and for a whole bunch of reasons. Interest rates, affordability, youth employment and wage rates, government policies, consumer or student loan debt levels, and overall economic conditions all play a role. Take a look at the chart below: the home ownership rate for Americans under 35 (the bottom line in purple) has always been low. It was over 40% in 1982, fell to the mid-30s level by 1994, rose to the mid-40s before peaking in 2004, then was already trending downward even before the 2008/2009 mortgage crisis and recession.

Ownership by age

Still, that decline since 2004 could reflect new anti-ownership attitudes from millennials? Maybe. One of the big problems with making broad statements about different generations is there are some effects that are persistent over time (called cohort effects), and others that are transitory (age effects and period effects.) For example, today’s millennials are unlikely to start subscribing to print newspapers or buying CDs when they turn 35. It is unclear if the current relatively lower level of home ownership is the beginning of a trend, or just a reflection of some large and disruptive event that has delayed ownership, but not permanently.

Ummm…like the worst recession in 70 years, which saw millions lose their life savings due to home ownership? That’s a reasonable assumption! 🙂 You can see from the chart below that home ownership in the US (this is for ALL age categories, not just 25-34 year olds) rose at the beginning of the 20th century to 1930, then dropped following the Great Depression. That was a real drop, and significantly affected the US banking system. But by 1950 home ownership was rising faster than ever before, and was well above the previous peak. It is possible that the current decline in home ownership is the start of a long term trend, but it is also possible that once the memories of 2008/2009 fade, ownership and mortgages will rise again. The only way to tell the difference between cohort effects and period effects is to wait a decade or two.

Long term ownership

But we still know that millennials hate banks? I don’t think so. If you look at the image below, you can see that the reasons that millennials haven’t bought yet varies by region. They may not have the down payment, they have too much student loan or credit card debt, or they don’t know where to start/worried about their credit score. None of those seem to indicate that the banks themselves are the problem: in fact, the survey suggested that more than half of millennials are planning on buying in the next two years.

Millenial home reasons

“Generation rent” can more accurately be called “generation rent…for now.”

In my next column, I want to take a look at how millennials are likely to do business with banks? Are they going to be digital only, fully self-serve, and do everything over the internet? Can banks need to get rid of human mortgage lenders?

Duncan’s Reading List & Book Reviews: May 2015

2015-06-05 08.07.21

Nearly 3,000 pages read this month, thanks to some long flights. I have been told those little screens in front of your seat can display movies as well as maps. I’m not sure how that could be more fun than what is on my printed page…

The Count of Monte Cristo – Alexandre Dumas, 1844-45 (1,243 pages)

One of those literary ‘classics’ that almost no one has ever read: we usually only know it through bowdlerized children’s versions, comics, movies, or popular culture references. But unlike some other classics which can be inaccessible to modern readers, I adored nearly every page of the book! It was written in serialized form, with chapters being published weekly, and you can feel the momentum and the cliff-hangers working. There are many characters, but you keep them straight after a while and the payoff from the depth of plot all comes together in the final 300 pages. This is the easiest thousand page book you’ll ever read, but it is not for kids: if the tortures, murders, and other deaths weren’t enough reason for censors to create an expurgated version, the poisonings, extra-marital liaisons, attempted baby killings, hashish sessions and lesbian affairs put it over the top. I have only read this translation by Robin Buss, but I found it very good. Now I need to go buy the Three Musketeers! 10/10.

Caliban’s War – James S.A. Correy, 2012 (595 pages)

Abaddon’s Gate – James S.A. Correy, 2013 (539 pages)

Cibola Burn – James S.A. Correy, 2014 (581 pages)

Books 2-4 of The Expanse series keeps us largely with James Holden and his spaceship crew, but also tells the story with POV slices from several new characters. Roberta (‘call me Bobbi’) Draper the Martian marine, and her new best friend and senior UN politician Chrisjen Avasarala (not just female characters, but non-white female characters!) were my favourites. Spoiler alert – Detective Miller is back. Kind of. Don’t get me wrong: this is not Dumas level literature. But it is solid science fiction space opera. Fun, provoking, and it doesn’t insult your intelligence. All three books were equally enjoyable, and I will buy Book Five which is being published in June 2015. 8/10