About six months ago, I decided to get into electrical contracting. Long
story short, I'm seeing a lot more Teslas on the road these days, but
hardly anyone knows how to do routine electrical work, like bending conduit.
I decided to buy (not build) a firm—day-1 revenue is nice, and lots
of owners are looking to sell as they hit retirement age.
And so, a couple hundred hours in, meeting dozens of owners, poring over tax
returns, and spending countless afternoons on the phone with
banks, business brokers, and regulatory agencies, here's a bit of what I've
figured out.
Markets
Electrical contracting has dozens of submarkets.
Residential/small business "service work": EV chargers, deck lighting
Retail T.I. (tenancy improvement): prepping a "shell" for a Jimmy John's / KFC / Supercuts
New construction: single-family, multifamily (apartments), high-rise
Public/municipal: schools, post office, outdoor city lighting
The customers and sales processes in each submarket are unique; firms typically specialize on just 1-2.
It's the same wire nuts, sold a completely different way. I'm reminded of
software:
There is virtually no difference in the mechanics of work done between $100 an hour, $200
an hour, and $30k a week — all of the leveling up there is in sophistication on who you go after, what
engagements you propose and deliver, and how you package things for clients.
Broadly, markets are "all or nothing" union, or non-union.
Large "prevailing wage" and "project labor agreement" projects tend to be union.
Homeowners, small business owners, and other price-sensitive customers typically don't buy union.
Since markets are all-or-nothing, and shops tend to specialize in 1-2 markets, contractors (at the level of
the LLC/corporation/entity) tend to be all-or-nothing union, too. This gets complicated with companies (e.g.
national contractors) that operate across many union jurisdictions.
Markets also split on:
How much presales work is required to win a job? Detailed proposals with design work and estimates (public
utilities), or just showing up at the door when the call comes?
Do buyers care about licensure? How about insurance/bonding? (Public agencies: definitely, homeowners: maybe not).
Capitalization: Can the firm float $250k in wages before getting paid? How about $500k? Is waiting
90 days to get paid catastrophic?
Unions divide the world into numbered territories called "locals" (sometimes "jurisdictions"); I live in Local 595
IBEW operates an apprentice program. It lasts five years, everyone who wants to be an electrician in an
IBEW shop has to do it, and apprentices earn fast food wages ($18/hr). I can kind of see both sides to this;
it's very long, but the union sees it as an alternative to college where, rather than paying for it, they pay you.
However, the program is over-enrolled, so admissions are competitive and "pre-apprenticeship" programs
are taking hold, extending the overall commitment to timeframes approaching a Ph.D. (6+ years). Reddit discussion
Here in the Bay Area, non-union journeymen cost shops about $80/hr to employ ($45-50/hr W2 plus tax,
benefits, and vacation); IBEW 595 journeyman cost is about $120 ($75/hr plus $45-50 in benefits).
Union benefits are indeed better; the question is whether customers accept the $220/hr bill
rate required to pay them ($250+ for service work).
Unions specify the terms they'll work under in a collective bargaining agreement (CBA—here's IBEW 595's)
Big employers (e.g. airlines, school districts) typically have their own CBAs, negotiated 1:1 with the union.
Most electrical contractors are too small for their own CBA, so rely on trade associations (e.g. NECA) to negotiate on their behalf. What comes out
of this negotiation is a "take-it-or-leave-it" offer for all smaller firms in the Local's jurisdiction. Firms contractually
commit to the CBA using a "Letter of Assent"
Key terms of the CBA:
Which of the firm's employees does it apply to? e.g. an IBEW 595 CBA wouldn't apply to a truck driver, or an electrician in Los Angeles.
Employers must fire any employee who won't join the union after a set number of days (e.g. 8); this is known as a "post-entry closed shop provision" (it applies after a person is hired and starts working), and in right-to-work states, may be partly or entirely illegal (note: this area of the law is quite complex)
Wage and benefit levels (typically given as a minimum)
Frequency of pay: typically weekly
Work rules: overtime, minimum shift lengths, requirement to pay when driving to/from the job, hazard pay for doing dangerous things (e.g. climbing tall ladders)
The union sees itself as:
Ensuring the workforce is property trained, particularly regarding safety
A staffing agency, moving members between firms as projects start and end
A standard provider of a great benefit package (health, pensions) that's portable between contractors
A "cop" that enforces good behavior by contractors (fighting wage theft, unsafe practices)
Licensure
Anyone holding themselves out as an independent contractor, for most types of construction work $500 or more,
must, by law, hold a government-issued contracting license, to operate in the State of California
This does not apply to employees, either of contractors, or firms that would otherwise require regulation (e.g. electrical work at a power plant)
The California Contractor State Licensing Board ("CSLB") administers licensing.
There are dozens of license types, falling into three general categories
General Contractor, Class A (so-called "General A"), the "public works" category for pavement, dams, and other "civil engineering"-type work
General Contractor, Class B ("General B"), for buildings
Class C "specialty contractors": electrical, pool, HVAC, plumbing, 20+ others
"C-10" is the main license type for electrical contractors. This is a large, general, comprehensive license that allows solar installation, indoor and outdoor high-voltage (90 volts or higher) power wiring, and low-voltage signaling/telecom lines (Ethernet, CCTV).
Some of these are separately licensed, e.g. C-7 low voltage systems contractor—the C-10 lets you do it all.
CSLB licensing requirements: (California): (a) 18 years old, (b) pass one or more exams, and (c) document 8,000 hours of journeyman-level work experience.
Per the CSLB (phone call), work is certified using an affadavit-like certification of work from an: (a) employer, (b) journeyman (unlicensed),
(c) licensed contractor (General-B/General-A/C-xx), (d) fellow employee, (e) union representative (foreman),
(f) business associate, or (g) client (certification of work experience).
CSLB will waive (a) two years (4,000 hours) for completion of a two-year college degree, (b) three years (6,000 hours) for completion of certain degrees, or a union apprenticeship program.
California additionally requires anyone working for a C-10 contractor, to be further certified by the Department of
Industrial Relations, as a certified party or, barring that, a "trainee" which
requires enrolling at an educational institution, in addition to work experience.
People of the same trade seldom meet together, even for merriment and diversion, but the conversation ends in
a conspiracy against the public, or in some contrivance to raise prices.
It's remarkable that Adam Smith was writing in the 1770s, some 250 years ago; the quote
above is from An Inquiry into the Nature and Causes of the Wealth of Nations. And it's as
true now as it was then.
If you want to understand why healthcare is so expensive, why Illinois is teetering on the
edge of fiscal ruin, or why housing is so unaffordable in the Bay Area, that quote will get
you pretty far. The issue is competition, and regulatory capture.
Right-thinking individuals never want to hurt others; they just want what's best for
themselves. And even their clients.
Medicine has a long history of quackery—miracle cures,
bloodletting, all kinds of crazy
stuff. Perhaps it made sense at the time. But as science advanced, it became clear that some
things worked, and others didn't. And so today, every facet of healthcare is regulated—who
can practice, what they can charge, whether drugs can be sold, and to whom, even whether hospitals
can be built, and what equipment they can contain.
As this thicket of regulation developed, by politicians with input from the doctors, drug
companies, and insurers, they made sure to get a seat at the table. Until what emerged,
year by year, compromise after compromise, was something everyone could live with; the
output of a political process that let politicians "get something done", ostensibly for
the benefit of their constituents, that the regulated parties could live with.
That the regulated parties could live with—that's the rub. Competition—the bane
of the successful. When you've made it, sales are up and to the right, and politicians are
calling on you, asking your input on major legislation, there's only one direction you can
go: down. Who wants that?
So the predictable result is that competition suffers. Not because anyone wanted to kill it,
it just wasn't a priority—we decided we wanted other things more.
Nobody says "I want to be fat", they'd just rather eat dessert every night, than carefully watch
their weight.
Nobody says "I don't want competition", they just enact rules like:
State-by-state occupational licensing. Why hire a $350/hr lawyer in California when you can
get one who knows the law just as well for half the cost in Nevada, or even Mexico or India?
Does orthopedics, dentistry, massage therapy, or psychiatry work differently one state over?
But: who's going to make it "their thing" to make it possible?
A drug safety process that takes a decade, costs billions of dollars, and ensures the
only route to market runs through gatekeepers such as Pfizer, Merck, or Genentech, who get
a cut of the action
Banking regulation that have led to—count them—five new banks in the United States
from 2013 to 2017.
Five!
Want to open a new hospital? Many states have lengthy review processes—sometimes they even tell you
you can't close.
Charter school bans, ensuring everyone, not just the poor, can experience terrible, one-size-fits-all
solutions that would make Soviet bureaucrats happy
I don't know how to ensure regulatory processes maintain, or increase, competition. But I do know that,
if you give business a seat at the table, they're going to do what's best for them—not necessarily
consumers, and definitely not competitors, especially the most deadly and innovative ones out there—the
ones that don't yet exist.
I think this is why, in my heart, I still lean toward small-government libertarianism. It's not perfect, but
the surface for regulatory capture seems smaller. A compromise might be making every rule expire in some fixed
period of time—if it's still relevant, we can renew it when it comes up.
Tepper and Hearn
wrote a book on this topic. Early signs are that the Biden Administration is going to be tougher on antitrust; that's
at least a sign they acknowledge that concentration is a problem, which is promising.
Two other things I've realized thinking about competition and regulation:
I used to be suspicious of the conceot of "systemic racism". As I thought about it, I realized competition is
basically its economic equivalent: something possible not because anyone's evil, they just want things other than
"promoting competition" or "racial equity" more. Also, if you're on top, you definitely aren't taking steps to lose that
position.
There's no substitute for effective lobbying—maybe the activists got that right. It doesn't matter whether
you're the National Education Association (teacher's union), AFL/CIO, NRA, or American Medical Association. People
who engage in the process get what they want—period. Maybe the federal government should have a smaller role
in peoples' lives, but that's not the world we live in, today, in 2021. If you want a voice, show up.
I wrote before how one consequence of getting old is being less surprised. It's
comforting to think I'm getting better at seeing the patterns, and I think there's
some truth to it.
But I also think seeing the patterns—predicting where things are headed—is
getting more difficult. Perhaps, as Ray Dalio says, I'm just an ant, too zoomed
in to see the larger trends. Perhaps with a different lens—more distance, greater
abstraction—things would be more predictable.
I'm still not convinced that's true, though. Sitting here in the last month of the decade,
there have been a lot of surprises:
(1) Facebook purchased Instagram in 2012 for $1 billion. I'd just arrived in Silicon
Valley, and it seemed unthinkable to plop down $1 billion for a social networking site
with no tangible assets, only a handful of employees (JD above says it was 13), 30 million
users and no revenue. People called it "the Youtube deal of the 2010s", recalling Google's
$1.65 billion purchase of Youtube in 2006. Just two years later, in 2014, Instagram looked
impossibly cheap when Facebook upped the ante to $19 billion, later increased to $22 billion, for
nascent social networking app WhatsApp. For comparative purposes, that price was the
entire market value of Kellogg's—Raisin Bran and Corn Flakes—for no tangible
assets, 70 employees, and $20 million in 2013 revenue.
(2) At around the same time (2012), I remember wondering whether Facebook itself, the big
growth story, would come back from $20 after falling almost 50% from their IPO price around
$38. Eight years later, with stock testing $300/share, I think we can consider that one
answered—$1 billion in sales by 2012, today, $70 billion in sales after 15 years in
business. Nothing, ever, in the history of capitalism, has grown this fast.
(3) Also around 2012/2013, I remember getting recruited by some upstart ride-sharing
company, with sloppy engineering practices and a legally questionable business model.
That company was Uber. I sure didn't think they could overcome the
taxi raj.
(4) Intel, the preeminent semiconductor company and pillar of Silicon Valley, was Apple's
first port of call to manufacture the iPhone's chipset. Failing to see how the engineering
investment could work—$20 CPUs inside $200 iPhones, when you're used to making $150 CPUs
in $1500 PCs, in a "niche market" (high-end phones) no less—Intel said no.
Apple shipped 218 million iPhones in 2018.
(5) Along the way, Apple became the first trillion-dollar company in 2018. But
why stop there? In mid-2020, they became the first $2 trillion market cap company.
(6) "The young don't vote". Until Barack Obama showed up in 2008 with "Hope and
change"—and a sophisticated understanding of how the Internet was reshaping
campaigning, and fundraising. Then Donald Trump picked up the torch, sailing to victory
on "Make America Great Again"—and being a rare member of the over-70 set who's great
at Twitter. Alexandria Ocasio-Cortez wasn't "supposed" to beat Joe Crawley any more than
Donald Trump was "supposed" to beat Jeb Bush or Marco Rubio, but like him, she's very
good at using modern media tools, particularly Instagram and Twitter.
(7) Throughout the 2000s, the fed funds target rate hovered around 4-5%. Today,
sovereign rates are negative all over Europe, the 10-year treasury is yielding 0.89%,
and asset prices are in the stratosphere across the whole economy.
Here's what I take from all this:
(1) Communication is an essential human activity and whoever mediates/controls it is
enormously powerful. Looking back, it's not hard to see how Facebook, then WhatsApp and
Instagram—would be so powerful, as communication platforms. I think the effect size
took people by surprise, though.
(2) I wanted to think there were some folks at Facebook in BD or Corp Dev who saw the
potential of Instagram. Like, maybe there were some insiders who had this all figured out?
Go where the best are?
I guess not:
Claims were that he literally did it *without the board*. CEO-to-CEO
When building a new business, focusing on strategy will guide you toward building a business that's
valuable, and defensible.
Markets, competition, network effects. There's a big-picture, first-principles quality to strategy
that reminds me of math, or physics. dy/dx. F=ma. I think technical people gravitate toward strategy
because it feels like math; it's intellectually comfortable, something we know how to reason about,
and manipulate.
But what strategy misses is customers—the human element—with all our quirks, our short attention
spans, credit card balances, and proclivity to eat things we shouldn't.
Go ahead, build that perfect work of strategy—infinitely defensible, hugely valuable—in
theory. Too bad nobody actually wants it.
If you want to keep customers front-and-center, follow Y Combinator's advice: make something people want.
Get that right, and worry about the rest later.
(Sidenote: I think this is related to #14 and #15 from Slava's
57 startup lessons: Most
investor advice [Ed note: strategy] is very good for optimizing and scaling a working business. Listen to
it. Most investor advice isn’t very good for building a magical product. Nobody can help you build a magical
product — that’s your job.)
Economic geography fascinates me. It's something I've written about a lot
over the years—industry clusters, what's made where—and since
COVID, it's top-of-mind for an even greater number of people.
Probably because of how unequal the pandemic's effects have been. Stadiums,
hotels, restaurants—all closed. Whereas those working in software,
media, and a lot of manufacturing might be excused for forgetting there's
even a pandemic.
The industries getting ahead are largely those engaged in producing
intangibles—thinking wealth into existence—realized
as code, books, movies, electronic schematics, whatever.
The double-whammy: the folks working in restaurants, casinos, and malls
were never paid much to begin with, and now they're out of work, even as their
higher-paid peers are still at it, many from home.
But more than anything, I've been thinking about how few people in the economy
actually do this kind of work (produce intangibles). You can drive hundreds of miles
down I-80 and probably not see a single software company—they're clustered,
packed into office buildings, in maybe 20-30 US cities.
The US has Disney, Google, and Boeing. The sixth-biggest company in Italy? Poste
italiane, the post office.
Richard Florida was
right all along: 21st-century wealth has a lot more to do with battery-control software
on EVs, and video games, than stadiums or casinos.
The concept of "paper of record"—a single place to get the news—is relatively new.
Speaking for the United States, many publishers got started with the explicit aim of promoting
a particular political view; historians call it the
party press era,
from about 1780-1830, where editors openly endorsed a political platform.
I'm not sure whether that tradition came from the UK, but something like it lives on. In the UK,
The Guardian was always understood as the paper of labor/working class, The Financial Times for
businesspeople (similar to the WSJ), and the Daily Mail a sort of regional semi-tabloid, a
bit like Chicago's Sun Times.
The difference is that, whereas in the US, you expected to get "fair and balanced" from a professional
paper, the UK never had that expectation. News was well-reported, but written from a particular viewpoint.
I'm hopeful the future of news in the US can become more like the UK. Maybe you have your paper and I have
mine, but we both agree that one source is never "the whole story".
Before you call me a conspiracy theorist, yes, I realize not all papers are created equal. I might
prefer the editorial view of The Economist over the New York Times, but I'd never suggest the Times
isn't a great paper, employing many responsible journalists doing their best to report the news. They
just have a point of view, like any human would.
I had this thought myself but saw another guy with the same idea yesterday on Twitter:
Wondering if the New York Times might be a happier ship if the US had the UK model of multiple newspapers that are all perfectly open and comfortable having a political position on the masthead. The guardian is a left-wing paper. That’s not a problem - there are others.
Most people aren't early adopters. Super-important lesson for early-stage product management.
Even if you deliver something really good, late adopters will do their usual thing: wait.
Wait for their friends, their neighbors, kids, even people at their church. They just don't want
to be first.
The bad news: it can be very discouraging pitching a new idea to late adopters. They just won't get it.
The good news: after the early adopters have moved on, the late adopters will still be around. They'll
let you grow, and grow. For decades.
There are still people, today in 2020, who primarily use desktop software to do accounting—Sage 50 or
Peachtree. They aren't moving to Xero or Quickbooks Online—they don't see the point. What they have
works well enough.
My friend just sold his cloud devops consulting company for $10-20 million. They're helping companies—software
companies—move their application delivery to public clouds.
There are still millions of computers running Windows XP. It will probably take another 10-20 years for the majority
of cars sold to be electric.
The challenge, when doing customer development, is to figure out whether someone is an early adopter in
this type of product, and keep that in mind while speaking to them. My dad, for instance, would happily buy a new
hand tool from a hardware store, but won't buy an electric car anytime before 2025. Maybe 2030.
I think it's best to bias a little toward early adopters when doing customer development. After all,
you want to be where things are going, not just where they are today.
It is so tempting, when someone curses you out, to yell right back. To scream, get in their face, to react.
Don't.
The moral high ground is too important.
When you disapprove something someone does, the basic choice is like-for-like—to fight fire with fire—or
to draw a line, and say, that behavior might be good enough for them, but it's not good enough for me.
You want to be better—to control the terms of the debate. Not let the other person drag you down, to have
to fight them the way they want to be fought. Also, there's always someone who can yell louder.
Dianne Feinstein doesn't yell. Kamala Harris, California's other senator, loves making scenes, getting in peoples'
faces, and playing the "tough prosecutor" role. Feinstein knows that with only 100 members of the chamber and
a six-year term, today's opponent is tomorrow's ally. If you want to last 28 years as a Senator—the longest-serving
woman in the chamber—that's the only way to play it.
What I take from Feinstein, Gandhi, and others, is that toughness and results don't have to come at the cost of
bad behavior. Not everyone has to like you, but you don't score points being disagreeable, bad-tempered, yelling,
or using foul language.
Another group who understands the power of good behavior is Black Americans of the Civil Rights Era. Rosa Parks didn't
curse anyone out—she just sat there, knowing that said enough.
The standard you walk past is the standard you accept. Pick a small set of non-negotiable rules that matter to you most and enforce them ruthlessly. — Slava Akhmechet
People being people, any group of hundreds will have a few bad apples. Beyond that, people aren't angels—we get
mad, and hungry, and say things we don't mean, and come to regret.
That much is a given. So the question I can't help but ask, when a police officer steps on someone's neck to the
point of strangulation, is what the other officers on the scene did? What went through the other officers'
heads, standing there while their coworker strangled a fellow human? Was that OK? Or was there some reason—rank,
personal loyalty, fear of ostracism—that the other officer couldn't say, "You might want to let him breathe;
you're killing him".
Firsthand experience has taught me that accountability and a sense of mission—key building blocks of organizational
culture—don't happen by themselves. They are intentionally-created things, shaped every day by strong leaders
who reinforce good behavior, and speak up when people do things they shouldn't. Over years, sometimes decades.
Without oversight, organizations turn inward. In business, that means forgetting your customers, letting expenses go, and
putting your own career/agenda first. In my condo building, that meant a manager who ignored board directives and played
favorites, until he sent us a "resignation letter" hoping we'd beg him back—we showed him the door. At the DMV,
organizational rot is what causes five-minute conversations about your kids with coworkers, when the line is out the door.
My own views on police accountability have changed. It might be because I'm living in deep blue America, or maybe I've just
seen enough organizational dysfunction to understand how things go wrong. But I see now how "serve and protect" might
not match on-the-ground reality. "Serve and protecet" certainly wasn't the mantra in my 800-person condo building, why
would it be for the police? Doubly so given a historically strong union, muscular political organization, and norms that
include going after "cop killers" much harder than others, and letting friends and family walk free from speeding tickets?
I wonder what it will take for these norms to change. The wanton destruction of life and property we've seen this week
are a terrible price to pay. (I know the protesters might not be doing the looting, a complex issue in itself.) The
price we'll pay for this is hard any time, doubly so during a combination economic meltdown/public health crisis. What
scares me is that it might be working. If I was a cop, I'd definitely be thinking twice about how I act after this week.
"Avoid others' germs. Like getting sick? I don't." Kind of funny, looking back. I wrote
The Future is Distributed 18 months
ago, long before anyone was talking about COVID or pandemics. These days, the grinding monotony
of shelter-in-place has been a real test of my patience; it's shown the cracks in remote-first,
or remote-only.
There are indeed advantages of working away from the office. But it's not without cost. Even the
Buffers and Basecamps of the world, acolytes as they are of remote-first, gather together from time
to time. Surely not for naught?
Now I'm reading everywhere that cities are over. Nice flashy headlines, but these things always take
longer than you think. I think we might have
reached the 90s dream of "paperless office" now, in 2020, but nobody's even paying attention anymore,
and it didn't happen exactly as we thought. What resulted was a synthesis of old and new.
Living far enough away not to be forced into a telephone booth-sized apartment. Somewhere with a
good school district and quality of life, away from the noise, smash-and-grabs, and casual crack
smoking of the big city. None of those are hypotheticals to me. But close enough in to make the
occasional in-person conference, investor meeting, or dinner when you need to.
Regionalism has gradations. I live in Oakland, far enough not to drive to Sunnyvale every day,
though I'd consider a job down there. Housing costs about half as much here. Farther afield,
one could choose Sacramento, or even Tahoe or Reno.
I saw some chatter about the tech scene in Wyoming on Twitter. That seems a little far-fetched.