The Rise and Fall of Markets

Recently, I read WeWork and Counterfeit Capitalism. An important part of his thesis is that the bad behavior is driven by cheap capital.

So why is capital so cheap?  This is not a lever some corrupt or incompetent official can just pull – it has to be the result of economic forces.  Harry Dent argues that a huge driver of economic waves is population demographics, especially the “baby boom” style of population peaks and valleys.  He predicted in the early 2000s that there would be a housing bust in 2010, based on the underlying population demographics in the U.S. (Reminder, he was right (or lucky?), but off by 2 years.)

Here is how that prediction went: the post WW2 baby boom produced a big (but temporary) surge of births.  Over time we saw new births grow sharply, hit a peak, then decline, forming a big wave in population.  Those children age at a predictable rate, so we can watch that wave flow over time – for example, six years after the beginning of the baby boom, we could anticipate (and later observed) the number of children entering primary school growing, with the peak of new school kids exactly 6 years after the peak of the baby boom.  After that, we can easily predict that there will be a decline in kids entering school.  We can look ahead to any age-related event in life and anticipate a related boom, peak, and bust.  There is also an echo boom behind the baby boom that has the same pattern.  Dent’s prediction about a housing crisis was based on the average age when a person upgrades from their first “starter” house to a larger family house: 44 years old.  So he looked at the echo boom curve for when the peak hit 44 years old, and could predict that leading up to 2010, we would have huge growth in this housing market, and after that, we would have a bust.  He couldn’t predict that we would have all the sub-prime bad behavior, but bad / stupid behavior is probably typical in any high growth market, because the growth hides bad behavior.

So if we had a peak of 44 year-olds in 2010, and that peak of people was large enough and economically strong enough to drive the whole U.S. economy up, then down, when are they planning to retire?  If we assume 65 is the average retirement age (I don’t know for sure, but that has to be close), then 21 years after 2010, or 2031 is when the peak of retirees will hit.  Before that peak, we would expect a steady growth of retirees, just as we saw growth in the housing market.  I don’t know how long that growth precedes the peak, but before we enter that growth phase, we should see all those people saving for retirement, shoving as much cash as they can afford into mutual funds and what-not. And that cash, sitting in stock funds and bond funds, is sitting there waiting for money managers to invest it.  The more cash there is in those funds, the more competition from money managers looking for a high-return place to invest, and therefore two things: low interest rates (no need to pay high rates when there are lots of investors eager for any return at all), and a target-rich environment for scam artists pitching bad ideas about how to generate a big return.

But this won’t last forever. If 2031 is the peak of the retirement age, the beginning of that curve has to be coming. If the beginning of the rise of that retirement bubble starts 10 years before the peak (again, I don’t know), that is just a year away, in 2021.  And this curve will be different from the housing boom and bust.  With housing, we had a boom and bust of new buyers for larger homes. But after the peak, most of those buyers stayed in their homes.  The bust was not because no one needed a home anymore, but only because the number of new buyers declined. With retirement funds, we have two things happening: first, people stop contributing, so there are fewer new investment dollars competing for, and therefore driving down, interest rates. Second, retirees will actually start withdrawing investment money, accelerating the reduction in available capital.

With a huge wave of retirees switching from buying to selling, I think the price of investments will go down (stock market decline?), and interest rates / capital return rates will have to go up. That will solve the “cheap capital leads to bad behavior” problem, but will probably bring troubles (and opportunities) of another kind.

The Ultimate Go Bag

I volunteer with a couple of organizations that respond to emergency situations: Community Emergency Response Team (CERT) and Amateur Radio Emergency Services (ARES). They both want me to have a go-bag ready to support those duties on short notice. Plus FEMA, DHS, State of California, and City of San Diego all recommend that you have some sort of personal emergency preparedness kit.  Thinking about the needs across all three of these go-kits, I realized two things: I definitely didn’t want three separate bags, but one bag that covered all three scenarios would be too big, too heavy, and would be filled with unnecessary items for each mission.

First, let’s talk about each mission, and what the requirements might be. None of them are like backpacking in the wilderness, living completely out of my pack. Generally, I will be working a 4 or 8 hour shift, and be fed or given an opportunity to get food.  I may be working outdoors, but will generally have access to shelter. But because these are emergency services, I want to be prepared for meals to be late or incomplete, or shifts to run unexpectedly long.  If I expect to be somewhere overnight, I will bring additional gear. With this bag, I am preparing for a 4-8 hour shift, but won’t be too uncomfortable if that stretches into 24 hours.

CERT: CERT will have two types of operation – Official deployment, and unofficial community response. Official deployments in my city will probably involve back lines administrative support, but might be in a field operating post with temporary facilities. Unofficial deployments most likely involve me walking through my own neighborhood doing search, rescue, and first-aid. In either case I will need some tools and supplies, but only expect to have to support myself for hours at a time, rather than days.

ARES: This is about providing radio communications when other channels are down or overloaded.  Often I will be working out of a hospital or fire-station, so it’s like going to an office building to work. Other times I will go to a temporary field station to support a non-emergency event like a marathon or an air-show.

Personal: This is for when I have to leave my house quickly and unexpectedly due to a problem with my house or neighborhood.  Where I live, that will most likely be wildfire or earthquake. Probably I just put my family in the car and drive out of the danger zone, but evacuations are always filled with unexpected delays and challenges, so some preparation makes sense.

So I made a list of the items I wanted for each mission, and compared them to see where there was overlap and where things were different. I built a basic go-bag that contains all the overlapping stuff, and a set of inserts that contain the mission specific components.

Back of Bag

Right Side with water bottle pouch

Left side

Solar battery on front of bag

Main compartment

I chose the 5.11 Tactical Rush 24 pack. The molle attachment system lets me add extra pouches as necessary, and the internal storage keeps all my gear really tidy. The main compartment has five pouches to organize gear along the front and back walls, and leaves some space for additional content in between. This bag holds all the things that are common across my use-cases.

Rear of main compartment

Top pouch: Outdoor Management

Bottom pouch: Food

The rear of the main compartment (left picture) has a zipper mesh pouch on top, and a draw-string solid pouch on the bottom. Outdoor management (sunscreen, bug repellent, toilet paper & wipes, hand sanitizer, lip balm, space blanket, hand warmers, camp soap, camp towel) is stored in the top compartment (middle picture). The bottom compartment (right picture) contains food (snacks, one meal, utensils).

Front of main compartment (top pouch is at the bottom of the photo)

Water accessories

Personal Protective Equipment

Tools

The front of the main compartment (left picture) has two mesh zipper pouches and a zipper solid pouch on the bottom. I keep chlorine tablets to purify water and hydration supplements in the top pouch (left-middle picture). I don’t expect to need to purify my own water, but the tablets don’t take much space. The middle pouch (right-middle picture) holds my personal protective equipment (gloves, dust mask, ear plugs, safety glasses, and a bandana). Tools (headlamp, knife, duct tape, lighter, paracord, spare batteries, zip ties) go in the bottom pocket (right picture).

Outside top right pocket loaded

Outside top right pocket partially unloaded

Trauma kit

On the front of the bag there are several more pockets – each with their own internal organization. The top right pocket (left picture) holds my trauma kit. In the middle picture, the pocket is partially unloaded. In the right picture you see my trauma kit with supplies for a larger bleeding / shock type injury.

There is an identical / mirror image pocket on the top left of the bag (left picture), which contains my first aid kit and some personal medications (right picture). The first aid kit is for cuts & scrapes, so it has lots of little band-aids and ointments, unlike the trauma kit, which is more about stopping the bleeding from a large wound.

Front lower compartment

Radio reference guides

Misc. front pocket stuff

There is a lower pocket on the front of the bag (left picture) that has several more organizing compartments inside. I made some reference cards for local radio repeaters and procedures (middle picture). The left picture shows my spare radio battery, playing cards, pens and markers, and a glasses-cleaning cloth.

Tool pouch, lozenges, notepads

Maps, compass, poncho, garbage bags, nitrile gloves

Contents of the tool pouch

In the internal slip pocket of this front pouch (left picture), I keep a belt tool-pouch, throat lozenges, and notepads. In the zipper pocket behind the slip pocket are maps, compass, rain poncho, garbage bags, and a small pouch with some nitrile gloves (middle picture). The right picture shows the contents of the belt tool-pouch: flashlight, EMT shears, multi-tool, pen and marker.

Solar battery on front of bag

Solar battery flipped up

Lights on

I also wanted to cary an external battery to recharge my personal electronics (iPhone, iPad, etc.), and being able to charge it from a solar panel seemed like a nice touch. In the left picture you can see it strapped to the front of my bag. The middle picture shows it flipped over (it can be strapped into this position) to show the built-in lights. If I need to light up my work area, or be sure people can see me when I am walking (maybe during search & rescue, or fog or night), I can turn on the lights (left picture).

Water bottle pouch

Holds 1L of water

Built-in 2L hydration

Of course one of the most important things to cary is water. The bag has room for a 2L hydration bladder (right picture), and I added a water bottle pouch to the side (left & middle pictures). The pouch will hold a 1L disposable or a reusable bottle.

Anyone 878 radio

Radio pouch on shoulder strap

Padded top pocket

Radio ear piece

I am a ham radio operator, so I would bring a hand-held radio (Anyone 878 in the left picture) for all of these scenarios. I added a radio pouch (also from 5.11) to the left shoulder strap, and ran an external speaker / mic over to the right shoulder strap (left middle picture). In addition I have an earpiece that lives in the top padded pocket (right two pictures).

This covers all the basic items I want with me for any of the three go-bag scenarios. When I deploy to ARES or CERT events, I will want to add some mission-specific stuff. I didn’t want to have to pack a list of things, because this defeats the purpose of a ready-to-go bag, and my experience is that that I will miss something in the packing process. So instead, I created a couple of inserts that cary the mission-specific items for each type of event.

For some ARES deployments, I bring a lot of radio, power, antenna, and other gear. But for some activities, I might only need my hand-held radio and a few more accessories.

Molle plate as base of insert

Insert with pouches

Top pouch open

Powerpole accessories

J-pole antenna & accessories

The base of my mission-specific inserts are the 5.11 Tactical COVRT molle inserts (top left picture), to which I attach various pouches. I got these also from 5.11 Tactical, but the beauty of molle is that you can find compatible pieces everywhere. In the top middle picture, you can see two pouches and a 25′ coax cable attached with velcro straps. You can also see some luggage tags I made to attach to EMCOMS power and antenna lines, and (not visible) one that includes my FCC license. The top right picture shows the smaller pouch opened to show two zippered compartments. One side holds several Anderson PowerPole adapters and cables (lower left picture). The lower right picture shows the contents of the other side of the pouch, which is my J-pole antenna, 6′ coax extension, 50′ of cord to hang it, and a bunch of RF adapters for connecting to the radio and other antennas or coax.

Screwdrivers

Bottom pouch

Contents of pouch

I have a set of screwdrivers with interchangeable bits tucked under the molle straps of the lower pouch. One of the screwdrivers has 4 sizes each of flat and Phillips tips; the other one has 8 sizes of star tips (top left picture). The top right picture shows the bottom pouch, which contains more radio batteries, a charging cradle & AC adapter, a 12v adapter, and my ARES field guide (bottom picture). If you noticed in the previous pictures the cable with PowerPoles on one end and a barrel connector on the other – that works with the charging cradle, so I can charge my radio batteries in a car, or from any 12v source.

Insert fits in main compartment

Perfect fit

Bag get a little thicker

The insert has a handle on top for easy handling, and is a perfect fit in the main compartment. The bag gets a little bit thicker, but works great.

CERT insert works like the ARES one

In the main compartment

Helmet on a lanyard

CERT insert

I use the same approach for my CERT gear, but it has the special case of the hard-hat. That fits better clipped to the outside of the bag, but I wanted all my CERT gear to be a single unit that gets added to the main bag.  I solved this by putting a lanyard on the carrier plate with a carabiner to clip to the hard-hat (right picture). When the insert goes into the bag, the lanyard snakes out the partially zipped bag and clips to another carabiner on the side of the bag (right middle picture). The only time I separate the hard-hat from the CERT gear is when I put it on.

CERT insert

Tools

Vest & CERT guide

Duct tape and builders crayons

The CERT gear includes chem lights, markers, my ID card, and three pouches worth of stuff (left picture). The lower right pouch is a roll-up drop pouch, again from 5.11. It contains my water & gas shut-off tool, a small pry-bar, and a 10M climbing rope (left middle picture). The tall pocket holds my CERT vest, a CERT Field Guide, and 2 dust masks (right middle picture). Included, but not shown here, are several door stoppers to keep doors open behind me during a building search. The smaller top pouch holds more duct tape (each of these little flat packs is 5yds, so altogether I have 20yds), plus some construction crayons for marking search and rescue symbols on buildings (right picture). Still to be added to this pouch are rolls of plastic ribbon for marking victims for triage.

This arrangement of a main bag plus mission-specific inserts means that I have a go-bag ready for several different scenarios. I just put the needed carrier into the bag, zip-up, and head out the door.

Product Manager Portfolios

I recently saw a question looking for templates for a Product Manager portfolio. I have been a product manager for about 10 years and don’t really have one of these, so here is how I interpreted the question: This is a newish product manager getting her first or next job as a PM, and was asked for a portfolio or somehow feels like she needs one to showcase her skills.

I think the first step is to identify what type of product manager you are. Some focus more on market research and product specification, some focus on creating marketing content, and some focus on growth hacking & promotion. Of course, some have end-to-end skills that include all of these. (Note this is not an exhaustive list, just some loose categories.)

When I hire product managers, I think of them as having three primary qualification areas. These are:

1. Expertise in our product – this is especially important if you are working on a technology product.

2. Customer domain expertise – understanding the life of the customer, and what problems she is trying to solve.

3. The skills, practices, and processes of Product Management

I usually expect that I can only get two out of the three areas from someone that I hire, and have to train for the third. For example (my industry is software), I might hire someone from an internal tech support or field engineering position. That person probably knows a lot about our product and customer needs, but not much about being a PM. Or, I might hire an experienced PM from another company, so they have PM skills and probably some experience with my customers, but don’t know my product.

Each of these qualification areas has some basic skills that I do expect to find from everyone:

1. Product skills vary by industry, but even if someone doesn’t know my product, I hope they have the fundamental technical skills for my industry, and can learn the important details of my product quickly, and connect it to other concepts in the industry.

2. Customer expertise also varies by industry, but I expect someone I hire to have some experience with my customer. For example in the makeup industry, personal use consumers have different needs & daily process from TV & film makeup artists, but hopefully my PM understands the fundamentals of makeup, or film & TV, or at least wears makeup. (I don’t know anything about this industry, I was just trying for an example that isn’t tech.)

3. PM skills & practices can (like anything else) get very advanced, but there are some basics I hope even someone new can bring. Examples are writing, analytical sills, math, basic business, people skills, visual communication skills, etc.

To build a portfolio, think about where you have skills in each area, and find or create some artifact that helps demonstrate one ore more skills. A blog can show writing, communication, analytical, and logical reasoning skills. You might also be able to show some industry knowledge, depending on your subject area. If you can point to a webpage you helped design, or write, or promote, or whatever, that is good. If there are products in the market that you worked on, point to any of the marketing material (although be clear about your contribution, so people don’t assume you created the ad if you didn’t).

In my last job interview, I was asked for a writing sample, and I pointed to my blog (not frequently updated, but there are some good examples of my writing & thinking), the website for my current product (I didn’t build the web page, but developed the message & the product), and a couple of other websites that were hobby projects that I thought showed some important skills.

Social Jobs to be Done

I am a huge fan of Jobs to be Done theory (JTBD). It has really revolutionized my thinking about why customers buy a product, and what makes a product great. The founding principle of the theory is that customers don’t buy your product because they want your product, but because they have a job to do, and they think your product will help them. In thinking about the job that the customer is trying to do, the theory breaks them down into groups: Functional Jobs, Social Jobs, and Emotional Jobs.

The Functional Jobs are relatively easy – e.g. I need a cheaper way to get to work, or I would like to be able to check my email without carrying / starting up my computer. For functional jobs, we pretty quickly get to the kinds of performance metrics we are used to thinking about with products – the cheaper way to get to work might be satisfied by a car with better gas mileage, and checking my email might be satisfied by a smart phone or tablet. But what about the other two, Emotional and Social Jobs?

Christensen makes a point in his book to say that if the JTBD only has Functional components, then a person (or a computer) can quickly weigh the relative benefits, and there will be one obvious winner. All successful products will be forced to optimize to those dimensions, and you end up with a commodity. He wasn’t suggesting that there is no room for choice or variation in your product, instead he was highlighting the importance of thinking about the other two components present in most jobs. I think a lot of us, especially if we come from an engineering background, tend to ignore and misunderstand these fuzzier-seeming components of a customer JTBD. This is especially difficult when we try to understand the difference between an emotional job and a social job.

As I try to apply JTBD theory in my work, and teach it to the people on my team, this lack of clarity shows up pretty fast. My first advice to the team was “be aware that these factors are important, look for evidence of them in our customers, and capture them in our descriptions whenever we can”. In other words, take a swing at it, but we really don’t understand how to examine, much less compete in these areas. We could possibly be blind to two-thirds of the problem our customers are trying to solve!

So now I am collecting examples that I think help highlight social and emotional jobs, and the difference between them. Today, I was looking at Apple iMacs, and something along this line caught my eye. Look at this photo comparing a regular 27″ iMac with the iMac Pro.

There is a huge difference in price between the two computers, and the iMac Pro offers much a much more powerful range of CPUs, bigger storage, and (not shown in the picture) more memory and better GPUs. From a functional perspective, we can see the difference between the two computers. One is much more expensive, and much more powerful. I can decide if I need the extra capability, and whether it is worth the extra price.

Now notice the colors of the two computers. The iMac Pro is Space Gray, and comes with a matching Space Gray keyboard, mouse or trackpad. Why? It doesn’t have anything to do with the performance of the computer so it doesn’t solve a functional job for me (like enabling me to do advanced AI modeling). What about an emotional job? Maybe I just like the way it looks, so when I sit at my desk, I enjoy the aesthetic environment more? That makes sense, and we see that the laptops are offered in multiple colors to satisfy that emotional job. But why isn’t the regular iMac offered in multiple colors, so I can satisfy that job with either version?

I think Apple has limited the Space Gray to the iMac Pro to satisfy a social JTBD. When you spend that much more for a computer, you are probably doing it for a reason – you work on more advanced problems than most other people, or you can afford to buy the top of the line no matter the cost, or you are smart enough to know how to take advantage of the extra performance, or maybe the boss values you more than your colleagues. I work from my home office, so no one will know if my computer is Silver or Space Gray and there is not much social benefit to me of having that distinctive color. But imagine that I am a programmer or researcher or designer or video editor, and across all the adjacent cubicles and offices, there is a sea of 27″ silver computers. If mine is one of only a few Space Gray ones, it stands out as something different, something special, and by association, separates me as different and special.

A college professor once told me about an insurance company where all the employees had identical office furniture, right down to round trash baskets. Managers got square trash baskets. In this ocean of uniformity, the small differences were glaringly obvious. The professor said that if someone suddenly ended up with a square trash basket, other employees noticed and asked about when they had been promoted, and if they hadn’t been, accused them of some sort of social fakery for having a status symbol to which they were not entitled.

I don’t think Apple expects many people to buy the much more expensive Pro computer just to show off (although it seems inevitable that a few will be sold for this reason). I do think they made the Pro stand out visually as an extra benefit for these high-end customers. They are different / special / more valuable in what they do, and therefor what computer they choose (and how much they spend), and Apple wants to help them signal this to their social group.

When we are designing our own products, we should notice if the job our customer is doing when using our products is something that would elevate our customer’s social status, and look for ways to help them signal that to their social group.

Tesla Disruption

Many casual observers view Tesla as a disruption to the traditional auto industry, because they have delivered an electric car with range and high-end appeal that has been so far unanswered. The assumption is that the existing major automakers are incapable of understanding an electric powertrain. Professional observers, like Horace Diedieu, are more skeptical of the disruption claim. Horace is not a skeptic of disruption in general – he trained with Clayton Christensen himself, and speeks eloquently and often about disruption theory. But after investigating the history of automobile production and Tesla’s methods, his conclusion is that Tesla is not disruptive. I intend to argue that Tesla does follow Christensen’s pattern of disruption. My argument will be, perhaps, less rigourous that Mr. Diedieu, but I hope to be as persuasive. Note: Much of the analysis of Tesla versus the rest of the industry has moved forward to speculation over how autonomous vehicles might remake the auto market, transportation, cityscapes, and eventually society. I am starting with the assumption that wide-scale autonomous vehicles are far enough in the future that there is worth-while discussion to be had on how a company like Tesla will impact the traditional automakers in the meantime.

Disruption theory has several principals. The one that Mr. Diedieu focuses on suggests that disruption occurs only when the business model changes. If a new technology / feature / capability comes along and allows a company to adopt it without changing their business model, then it is called a sustaining innovation, and incumbent providers are generally expected to be able to defeat new market entrants. Mr. Diedieu argues that the most important component of the automobile business model is the manufacturing process. His research suggests that every significant change of leadership in the industry has been accompanied by an important change in the manufacturing process. Because Tesla is using traditional methods for most of their manufacturing, the change from an internal combustion power train to an electric one is almost certainly a sustaining innovation for the existing auto industry.

Another principle of disruption theory is that the basis of competition changes. Existing providers of a product focus on satisfying the ever-increasing demands of their best customers for improvements in a few critical performance metrics. Meanwhile, a new entrant provides a lower-level of these performance metrics to less demanding customers, and focuses on other metrics that are more interesting to that audience. One of the classical examples of this is the computer hard disk manufacturers working to deliver better cost per gigabyte, storage density, and total capacity to their large, demanding mainframe and mini-computer customers. PC manufacturers did not require the highest-end capacity or storage density in their applicaiton, and so could be satisfied by relatively low performance in these metrics. And because they were buying much lower capacities, they could tollerate somewhat higher cost per gigabyte. Because PCs would sit on or under a user’s desk, however, there were significant concerns about overall size and noise levels. The performance measurements for high-end applications could be ignored, and competition (in this market) shifted to new metrics.

If we think simply about exchanging an internal combustion power train (engine, transmission, fuel system, engine management system, exhaust system, polution control systems, etc.) for an electric power train (batteries, electric motor(s), motor controller, charge controller), we can easily believe that electric power trains are a sustaining innovation. Existing auto manufacturers should be able to make this change with as little difficulty as a model-year design change. Any benefits that Tesla gains from using an electric power train also accrue to the traditional manufacturers, and they continue to have the advantage of large-scale production. Instead, consider the basis of competition for each type of car manufacturer.

The competitive dimensions for a car can be grouped into a few buckets: power train (which directly or indirectly affects fuel economy, performance, ride smoothness, noise, etc.), styling (including exterior and interior appointments), and the increasingly important automation features (adaptive cruise control, intelligent braking systems, traction control systems, obstacle detection and avoidance systems, parking assistance, lane-departure, navigation, and eventually, self-driving capabilities). Note: safety is also a competitive dimention, but once you put the intelligent features like ABS, traction control, obstacle detection, etc. into automation, all that is left is structural features and restraints like airbags and seatbelts, most of which are mandated across all manufacturers by law). Auto manufacturers purchase many components from suppliers, especially bolt-on safety features like air-bags and braking systems, and most especially the infotainment systems. However, they invest lots of time, money, and internal expertise on their power trains. The power train defines one of the most important characteristics to distinguish one type of car from another. When a customer focuses on features like fuel economy, acceleration, corning performance, towing capability, freeway passing, reliability, maintenance costs, purchase costs, and to some extent, vehicle size / capacity, they are indirectly choosing one power train design over another. Many of these metrics are incompatible with each other – better fuel economy and lower maintenance costs are the antithesis of acceleration and towing capability, and so they cannot be maximized at the same time.

An electric power train can be designed to outperform an internal combusion power train for all of these metrics simultaneously, for a sub-set of vehicle types. Tesla competes in the mid-size luxury sedan market, and other than price and range between refueling, can easily beat every related power train metric for all other vehicles anywhere near its class, at the same time. Further, a Tesla can best the fuel economy metric for the most fuel-efficient car while simultaneously besting the acceleration of the highest-accellerating car. Some of this is due to design choices by Tesla, but most of these benefits come from the generic characteristics of an electric power train.

Traditional auto manufacturers absolutely can exchange their internal combustion power trains for electric ones, but this will eliminiate much of the traditional differentiation between brands and models of cars. Every car will be efficient, quiet, reliable, inexpensive to maintain, and high-performance at the same time. There will be little room for Porche or BMW to brag that their cars are better performance cars than a mid-grade Toyota or Honda. Likewise, there will be little room for Honda and Toyota to claim that their cars are more efficient, reliable, or inexpensive to maintain than Porche or BMW. If we assume that passive safety systems are mandated by law and therefore not a strong differentator, all that remains for competition are styling, automation, and in the case of a new entrant like Tesla, large scale operations.

To date, Tesla is meeting the minimum requirements for styling, but is not able to win on these dimensions alone. They are currently far behind on scale operations. The electric power train gives them a temporary advantage, which traditional auto manufacturers may be hesitant to adopt, but which they will as soon at the tide of consumer demand shifts far enough. This leaves automation.

Traditional auto manufacturers are far behind in this space. Although they have basic automation features in many cars, they are almost always purchased as bolt-on components from a supplier, are rarely integrated with other automated systems, and are generally designed to be static. Tesla, on the other hand, is developing their car as a automation platform, with a range of sensors and control points to deliver these active safety and automation features, most of which can be improved via an over-the-air software upgrade. This means that their cars can improve in this important competitive dimension at the speed of software.

This difference cannot be overstated. A new car shipping today might have a navigation system that was designed several years ago. The infotainment vendor designs a system, the auto manufacturer designs the new system into a new model of car (which only changes every few years), begins a production run, and starts selling cars with the new system. An infotainment system might allow minor updates, but this process is also very slow and sporadically applied. Imagine if your laptop computer could only have the software available when it was originally made, with few or no updates. This is normal in the auto industry. Tesla is delivering a more integrated, software-controlled system. It is designed to receive updates via wireless signals throughout its life. Although Tesla hardware doesn’t get updated any faster than other cars, the software that defines how it works does. A Tesla on the road today can receive a performance or convenience or even safety improvement from a wireless update.

This is a huge difference in how features are made (from software, not hardware), and how fast they can be designed and deployed. And developing integrated computer / software-based systems requires a specific style of management that the traditional auto makers have so far failed to demonstrate, and which cannot be purchased as a bolt-on component.

So the disruption argument can be summarized as: the electric power train neutralizes one of the most important performance metrics between different styles and makers of cars, and the automation platform of a car becomes the new area of innovation and competition. Tesla has built their company based on being first to the electric power train, but is making most of their investments in growing the automation platform. Traditional auto makers can adopt the electric power train, but this just brings them neutral with other makers, and destroys some of their differentiation. The automation platform becomes a new requirement where they have little or no historical expertise, and so far, are making few investments.

Tesla still needs to learn how to build cars at scale (GM is currently building the electric Chevy Bolt’s at higher volumes than all Teslas combined), but if they can learn to deliver continuously improved styling (which expertise can arguably be purchased), and maintain their lead in the automation platform, they can certain grow as fast as they like, and have the potential to capture a significant share of the automobile market.

So the final question is: can Tesla learn to build cars at high volume faster than traditional manufacturers can learn to design and operate as a computer hardware / software company? There is no clear theoretical reason why one should be more difficult than the other, but we do have one clear historical guide. Apple learned to scale-up production of iPhones faster than Nokia learned to be a software company.

Rules

Every rule, process, and procedure in your organization was created based on some combination of legal, technical, practical, and arbitrary limits.

Does your company ship everything in just a few standard-sized boxes?  Maybe it is because they fit nicely into a UPS truck, so you get a discount.  Or maybe your first box vendor recommended them, and you have just gotten into the habit of using them.

Are you forced to use 16 digit SKUs for all of your products?  Maybe it is because, years ago, someone arbitrarily set this limit in a database.  Or maybe it is because all of your distribution partners have inventory systems that use 16 digit SKUs, and won’t be able to manage your products if you use longer ones.

Some rules are made because if we do it any other way, it impacts our accounting all the way up to the annual shareholder report.  Some rules are made because someone had to pick a number, so a couple of people got together and guessed.

We get advised all the time to break the rules.  Facebook has (or had) the famous motto of “move fast and break things”.  This approach has the advantage of helping to clear out the rules that were based on assumptions or obsolete technical constraints.  It also tends to create a lot of havoc when the rules encoded important legal or safety constraints. (A supplier of medical devices probably wants to be a little more conservative when changing processes.)

If the consequences of breaking a rule are small, and the opportunity for breaking out of legacy assumptions is high (like for a new social-media company), “move fast and break things” is probably a good rule.  You probably need to supplement that with a team or process that follows-up and fixes any critical issues that result.

If the consequences of breaking a rule are large (people’s health or safety is a risk), you need a more formal change review process.

Most organizations are in the middle of these two extremes, so you need an approach somewhere in the middle.  The key is to unpack the rule and look for legal, technical, and practical influences, and confirm whether they are still relevant.  The rest is probably based on arbitrary decisions that can and should be challenged.

Naming the parts is not enough

Some people can name all the parts of a car, and have strong opinions about the various parts.  That doesn’t mean that they know how to build a car that will work, or one that other people will want to drive.  And it definitely doesn’t say anything about their ability to build cars at large scale with a price and features that will attract a large customer base and generate a profit for the company.

The same is true for building software.  We know that many successful software companies today are made from web pages, databases, and maybe some machine-learning.  Knowing this this doesn’t help us know how to build a product people will like, or how to build a successful software company.

First we fear it, then we are mad when it isn’t perfect

Self-driving cars seem like they are just around the corner, and one of the benefits they hope to offer is more safety. Depending on who you ask, this might be a huge boon to the world, or an impossible dream, or the beginning of the AI apocalypse. There is almost always fear about new technology, but there is also a predictable path for successful innovations: First we fear it, then we overestimate the benefits, then rich people get it, then everyone gets it, and finally we take it for granted and start complaining that it isn’t better.

Today, autonomous vehicles are in the “fear and overestimate” stage. Another life-saving automobile technology, airbags, were a new thing when I was young, but are now in the “take it for granted” stage. An article in the WSJ (2015-09-02, print edition: “Fewer Air-Bag Replacements Needed”, section B2) had a picture with a caption that caught my eye. “Faulty Takata air-bag inflaters have been linked to eight deaths world-wide.” The article is not completely clear, but it seems like those deaths are from the air-bag failing to inflate when required.

We have some work to do to get there, but soon enough, today’s fears about allowing AI into our cars will be replaced with outrage that some manufacturer’s AI didn’t avoid enough accidents.

Our Competitive Advantage is Ease-of-Use

When developing a product, often a desired feature is “easy”. Perhaps this is even the planned differentiation: being easier than the competition. But “easy” doesn’t actually exist. It is like “cold” – physics has no concept of cold, only its opposite, heat. To make something cold, you must find a way to remove the heat. Likewise, if we want to make something easy, we have to find its opposite: hard.

We will find it nearly impossible to specify (or later market) “easy” in our product. Instead, we can work to understand what is hard about existing tasks, and remove those hard things. Now we have something concrete to specify in our product development, and something specific to market to our customers.

Follow the Money

GigaOm Pro Blog provides a list of likely trends, mostly in the infrastructure that powers cloud computing.  They don’t have a lot to say about software, which I believe will also be a critical turning point over the next few years for both public and private clouds.