Latest Entries

The Economics of uberX (Part II)

N.B. I’ve updated this post to provide additional support for my assumption that (i) uberX passenger behavior mirrors yellow taxi passenger behavior, and (ii) surge pricing alone does not explain the high average fare figure.

If you’re feeling masochistic, you can read Part I here.

Uber released a blog post last week regarding the impact of recent uberX price cuts on overall driver performance. The post appears meant to refute recent criticisms that the price cuts have had a deleterious effect on driver incomes. Uber and its board members have been quite vocal that the net effect of the cuts has been positive: Uber’s customers, they contend, were relatively price elastic, meaning that any percentage drop in prices was more than offset by a greater percentage increase in quantity of trips demanded, leading to an overall increase in total revenue (i.e., gross fares). The data appear to lend credence to this claim, but I thought it would be worth examining the data in a bit more depth to see what else we can tease out.1

What else must be true?

Whenever a company releases data, there’s a strong tendency to fall into the trap of assuming that what you see is all there is. But we should always remember that corporations have every incentive to present the best version of themselves, and that taking their selectively released data at face value is an especially pernicious form of gullibility. For this reason, it’s always prudent to ask, “If these data are accurate, then what else must be true?” In other words, if we are to believe a given dataset, do we also believe the inferences that follow.

Let’s start by reviewing the data Uber released. Note that this data refers only to uberX drivers in New York City in each of the last three Septembers (2012 – 2014).

Uber_-_October_2014_-_Google_Sheets

Right off the bat, I see a problem with this data. It’s odd to me that in 2012 and 2013 the average uberX fare was (i) so large, and (ii) so much greater than gross fares per hour. Even 2014’s average fare seems strangely high. As I mentioned in my last post about uberX, the average NYC taxi fare is $13.40. Given that Uber explicitly benchmarked their price cut against taxi fares, it’s reasonable to assume that Uber views uberX and yellow cabs as near substitutes, which implies that an average ride on one service probably looks a lot like an average ride on the other. And if you don’t believe that line of reasoning, consider Uber’s own request heat map:

[Edit: Andrew Salzburg (Uber's operations manager in NYC) has pointed out that these data reflect e-hails for uberTaxi, not uberX.  This is an important to distinction to make in light of Josh Mohrer's (Uber's GM in NYC) contention that yellow cabs are not a good proxy for uberX.]

heatmap

Here’s the same data visualized as a cartogram (again, this is from Uber’s own blog):

cartogram

Finally, if you’re still not sold, you can always look at the hour-by-hour video of pickup data that Uber released to YouTube last year. It bears out what’s said explicitly in the post: “If you look at all e-hail requests over a 2 week period, you see what you might expect: taxi drivers [who] use Uber get requests in all 5 boroughs, but demand is concentrated in the financial district and midtown Manhattan.”

Knowing this, it’s hard to believe that the average uberX fare is double the average taxi fare, since that would imply a completely different set of behaviors on the part of uberX customers relative to yellow cab customers.2 But for now, let’s call it a suspicion and move on. Maybe once we tease out some more inferences, this will all make sense.

Taxi drivers make money for two activities: (i) pulling the flag and (ii) driving a passenger (i.e., paid miles). While performing these activities, they endure the various incremental costs associated with driving, e.g., fuel and wear & tear (i.e., depreciation). These dynamics explain why the profit-maximizing driver’s dual objectives are maximizing the number of flag pulls while minimizing the total distance traveled. Knowing this, it’s worth asking if we can infer the two operational metrics that most affect taxi economics: (i) trips per hour and (ii) average fare distance. And as it turns out, we can do just that.

Let’s walk through the numbers for 2014. Trips per hour is a simple calculation. We just take the gross fares per hour ($36.16) and divide it by the average fare ($27.11), which implies an average of 1.33 trips per hour.

Fare distance is a bit harder. We know that, post-price cut, gross fare is calculated according to the formula:

Gross fare = base rate + (duration x per-minute rate) + (distance x per-mile rate)

Pre-price cut, though, the formula was slightly different:

Gross fare = base rate + (miles over 11mph x per-mile rate) + (minutes under 11mph x per-minute rate)

We don’t have any information about the amount of on-trip time an uberX spends above or below 11mph, but we do have a good proxy available. Using data from the 2014 NYC Taxi Factbook, we can deduce that taxis spend about 53% of on-trip time under 6mph. Since we’ve decided that yellow cab activity is a reasonably proxy for uberX activity, it seems fair to assume that uberX drivers spend a similar proportion of their on-trip time below 11mph.

Moving on, we know the rates in these equation because they’re posted on UberNYC’s home page. What we don’t know is duration or distance, and we need one to find the other. This would be quite a pickle, except Uber has kindly given us all the information we need to suss out duration. If we divide the minutes per hour “on trip” (29) by the number of trips per hour (1.33), we’re left with an average trip duration of 21.7 minutes. From there, calculating distance is a matter of subtracting time- and flag-related income from gross fares, and dividing the remainder by the per-mile rate, which yields an average fare distance of 7.2 miles. Finally, we can further calculate that average “on trip” speed is 19.8 MPH (distance / duration in hours) and that average utilization is 48% (minutes per hour “on trip” / 60).

If you’re more of a tables person, here’s everything I just said in pretty alignment:

Does this make sense?

As a reminder, we already know these metrics for yellow cabs, which, again, are ostensibly uberX’s nearest substitute. Per the 2014 Factbook (and logical inference):

In addition, the Factbook also tells us that utilization for the average taxi hovers near 50%.

Now that we know what must be true if Uber’s numbers are to be believed, I ask you: do these numbers make sense? They don’t to me. An average speed of almost 20mph? Even if the majority of uberX rides involve the outer boroughs (and they don’t), that seems crazy in a city where taxis barely average 10mph in the most common pickup/drop-off area.3 And besides, are taxi trips in the outer boroughs really so different from trips in Manhattan? According to data gathered for the 2014 Boro Taxi Market Study, the distribution of trip lengths in the boroughs is a little flatter than that of trip lengths in Manhattan, but not that much flatter:

Boro_Taxi_Market_Study_pdf__page_41_of_70_

And what about that average fare distance of more than 7 miles? Do we really believe that the average uberX ride originates in Lincoln Square and ends in Williamsburg? Or that the average trip is an airport run? Or do we believe that when people use uberX, they go on average almost three times the distance as they do when they take a cab, and that the trip takes them only twice as long? And despite all that, they still manage to maintain the same utilization as taxis? What about surge pricing? As Justin Overdorff notes, that’s one a plausible way to explain the discrepancy without assuming widely divergent behaviors. Luckily, that’s a hypothesis we can check using data collected by whatsthefare.com, and it turns out that surge pricing amounts to no more than around a 12.5% fare premium. Not nothing, but not explanatory either.

[Edit: Salzburg and Mohrer each point out plausible explanations for the disparity between uberX and yellow cab trip data: (i) most uberX trips occur in the outer boroughs and NJ, and (ii) uberX has an $8 minimum fare, which taxis do not.]

As it happens, I believe that the declared average uberX fare — a figure that’s more than double its benchmark — is wrong. And not just a little wrong, but horribly, weirdly wrong. So wrong that I have trouble ascribing motive to it.4 But that doesn’t change the fact that this data looks really wrong; in fact, it looks doubly wrong. So I’m going to perform a transformation on it based on nothing more than my intuition: I’m going to cut the average fare figures in half and see what effect that has on the metrics.

WARNING: It should be obvious that if you do buy the implications of Uber’s numbers, then you’ll give no credence to the analysis that follows. That’s your choice, and you are welcome to it. I’ve been wrong a lot in my life; I’m sure one more time won’t hurt.

 

And now we’re really in a pickle, because despite my transformation being based on little more than a hunch, it’s resulted in numbers that actually do make sense. Eleven-minute rides that go for 2.9 miles? Average fares of $13.56? An average speed a few miles per hour over a taxi? Realistic, likely, and plausible (because, outer boroughs).

uberX vs. yellow cabs

Ultimately, the only reason to release these numbers is to prove to drivers that the uberX program offers meaningful rewards. Uber’s made no bones of the fact that they are aggressively recruiting drivers. They’ve even gone so far as to offer income guarantees: for the last three months of the year, every uberX driver who spends 200 hours active in and around Manhattan and accepts 90% of their trip requests is guaranteed a monthly payout equal to $5,000 in net fares. Just how good of a deal is that? Thanks to the first footnote in the Three Septembers post, we can actually figure out the answer.

$25.10 per hour x 200 hours = $5,020, which means that, for the next three months at least, Uber is indemnifying full-time drivers against slow business.5 Good luck competing with that, companies-that-don’t-have-a-$1.5-billion-war-chest.6

Returning to the question at hand, is driving for uberX the “fantastic opportunity” that UberNYC’s GM says it is? It doesn’t seem like a bad opportunity. Of course, $25.10 isn’t really your take-home. There’s gas, wear & tear, car payments, interest, repairs, and insurance to account for, which could easily knock $5-10 per hour or more off your net. All told, a full-time driver probably averages around $15 per hour in take-home, which is right in the ballpark of what was inferable from Uber’s previous data release. To put it bluntly, Uber’s post provides further evidence that the economics of driving an uberX are exactly what we should expect them to be, that is, the same as the economics of driving a yellow cab. Or, at least, the marginal economics are the same. While taxi drivers pay to lease medallioned vehicles owned by others, uberX drivers enjoy the privilege of backstopping all the costs, liabilities, and risks that come with owning your own vehicle, which doesn’t sound so bad until you remember that cars break down sometimes, especially when you put 40,000 Manhattan miles a year on them. Not only do owner-drivers risk having to shell out for repair bills, they’ll have to pay those bills while their only revenue-producing asset sits idle.7 That situation is what the experts call, “operating without a net.”

A comment on utilization

Now, Uber bulls and board members have long argued that the key to this whole thing is network effects, which would presumably push utilization up in the range of 70-80%. Or, to use real numbers, they think that network effects plus algorithms will allow the average driver to produce 42-48 revenue-generating minutes per hour. This is an oft-repeated idea, that the scale advantages of Uber’s network will lead to a virtuous cycle that magically increases both availability and utilization.8 Respectfully, I call bullshit on that belief.

Early last year, the Earth Institute at Columbia University released a white paper entitled, Transforming Personal Mobility. The purpose of that paper was to assess the characteristics of centrally controlled, driverless car networks. Even under the absurd conditions of the paper (9,000 centrally controlled driverless taxis unimpeded by human-driven vehicles and pedestrians, and potholes), average utilization only reached 70%, and that was during peak periods.9 And that actually makes sense, because the behavior that drives utilization — customer demand — has to increase exponentially for utilization to increase linearly, because it’s not about aggregate demand, it’s about proximate demand. If you’re dropping someone off in Harlem at 12pm, it doesn’t much help your utilization if there’s high demand on the Lower East Side. There’s only so much fare chaining to be done, and when you factor in randomness (e.g., stop lights, traffic jams, other drivers, pedestrians, roadwork, potty breaks, fill-ups, blown tires, etc.) and waiting (Does everyone who calls a cab to their home really make it downstairs “in a minute”?), it’s nigh impossible to imagine a realistic scenario where uberX gets its utilization up over 60%, let alone 70-80%. And even if it did get up that high, consider what that would mean for availability. A car, after all, can’t be both utilized and available at the same time.10 If utilization goes up, the supply of cars will have to increase to keep up availability, which will exert downward pressure on availability. In other words, these aren’t metrics that can be independently pumped, they’re metrics that have to be simultaneously balanced. You are of course free to believe that the immediate future New York streets will be infinitely more orderly, predictable, and driverless than the present and past, but I’ll be taking the other side of that bet.

Conclusion

If someone truly wanted to “disrupt” the taxi market, they wouldn’t continue the decades-old trend of pushing risk onto drivers. Instead, they’d do something completely crazy, like buy a fleet of vehicles, hire a raft of high-quality drivers, pay those drivers a fair wage, invest in oodles of technology, and foster a culture that prioritizes efficiency, teamwork, and service. But I doubt such a company would get much interest from investors. Why invest in assets when you can invest in the platform that relies on them? So it goes.

By now it should be obvious that uberX, at least in New York City, is nothing more or less than a lightly regulated, heavily subsidized substitute for yellow cabs. Uber’s own data releases, when examined closely, bear this out. There is no technological magic here, there’s only the magic of subsidized prices, guaranteed wages, and bountiful promotion. Uber did a great and innovative thing by decentralizing the dispatch function and pushing it to mobile, thereby enabling them to build the first truly global taxi dispatch brand. They also did some very smart things with mobile payments, which helped them collect their rake with only a modicum of effort. Combine these innovations with exceptional execution, and it’s easy to see why they get so many plaudits (and so much investment). But at the end of the day, their business model boils down to taking a 20% rake of a large commodity services business while pushing assets, costs, and liabilities off the corporate financial statements and onto the shoulders of their not-quite-employee-partners. They own no cars, employ no drivers, and their vaunted network is built not on real assets, but monetary incentives. If the rest of their metros are anything like New York, they need global expansion to mask intra-city margin compression as their brand becomes less luxury and more commodity. That’s a tenuous position to be in, billion-dollar bankroll or not.

  1. All calculations in this post can be found in this spreadsheet. Feel free to copy and critique at your leisure. []
  2. Maybe it actually is the case that uberX passengers and taxi passengers use each service differently, but that would call into question the rationale behind Uber’s price cut. Mutually unique behaviors would suggest the two services aren’t substitutes at all, which would mean they’re not really competitors, which would obviate the need to “beat” the other’s price. []
  3. See pgs. 14-17 of the 2012 NYC-DOT Sustainable Streets Index. []
  4. When a company releases data that’s not just wrong, but beneficially wrong, it’s easy to ascribe motive. But sometimes, data just gets screwed up. []
  5. This is a good time to point out that, at $25.10 per hour, a driver would have to work 300 hours per month to net $90,766, which is what Uber has claimed to be the “median uberX small business income” in New York. I’m on record as doubting that claim, as are others. []
  6. I’ll now point out that this behavior is very much a form of predatory pricing, in that Uber is using its monumental cash reserves to sell its dispatch service to drivers at a loss with the express goal forcing less-well-funded competitors from the arena. The point of antitrust is to make companies compete on service, not balance sheets. If you’re sympathetic to the Chicago School, you’re unlikely to see that as a problem, because consumer welfare or something. Personally, I do think it’s a problem, but I’m more of a Herbert Simon kind of guy. So it is. []
  7. If you think insurance ever really makes a driver whole in that situation, well, that’s adorable. []
  8. This idea is especially magical when you remember that availability and utilization are countervailing concepts: at some point, increasing one can only come at the expense of the other. []
  9. Transforming Personal Mobility, pg. 25. []
  10. Yes, I’m aware of ride-sharing. No, I don’t think it changes the calculus, because I don’t believe that people in the US will use it in the volumes required to make the math work. I could be wrong. []

Beautiful Illusions: The Economics of uberX

Felix Salmon wrote a post recently on the economics of driving for Uber. Following up on Uber’s recent claims that the “median small business income with uberX” in NYC is more than $90,766, Felix rightly asked Uber to clarify how they arrived at those numbers. In response, Uber sent over the following chart, laying out the annual costs of driving an uberX for 40 hours per week.

Ostensibly, we should be able to subtract $15,080 from $90,766, and arrive at an estimated net income of $75,686. This is exactly what Felix does, noting along the way that these numbers represent a “good-faith” estimate on Uber’s part and lead to the entirely intuitive conclusion that being a self-employed uberX driver is a pretty good gig, especially when compared with driving a Yellow Cab.1 Following some pushback from the Internets, Felix wrote a follow-up post on Uber’s claims, in which he dug deeper and emerged with a greater degree of skepticism. I’d like to continue that line of inquiry in even greater depth.

In the following analysis, I’m going to focus on the unit economics of being a full-time uberX driver in New York City. My goal is not to value the company, but rather to examine whether driving for Uber is as lucrative as Uber’s PR would indicate.2 That said, drivers are Uber’s only productive assets, and they will continue to produce revenue for Uber only if it makes economic sense for them to do so. If the unit economics of driving for Uber don’t work, then any claim to future dominance should be met with skepticism.

A full-time job

Felix asked Uber to come up with an estimate for miles traveled over a year’s worth of 40 hour work weeks. In response, Uber sent data for a driver who covers 40,000 miles per year. Assuming two weeks vacation — what’s the point of setting your own schedule if you can’t take vacation? — that means our hypothetical driver is covering 40,000 miles in 2,000 hours, for an average speed of 20 mph. Meanwhile, the average speed of Manhattan traffic is barely 10 mph, though taxis do a bit better than that, averaging about 11.5 mph.3  There are a few ways to reconcile these numbers:

  • Uber defines an NYC work-week as 70-80 hours
  • Uber defines a “small business” as a car owner, who may drive their own car, but more likely rents it to drivers

Personally, I think the second bullet makes the most sense. A quick check on Craigslist will reveal a number of postings for car owners looking for drivers. Given that Uber’s raison d’être is to increase asset utilization, I assume they fully support an asset-efficient business model. But of course, that’s not the same thing as being an owner-driver making $91k on 40 hours of work a week, which is the claim being made here. Therefore, for the purposes of this analysis, I’m going to assume that the numbers Uber promotes as “potential earnings” are intended to represent the income for a single driver-owner.

Gross fare revenue

Let’s assume that uberX drivers move at similar speeds to every other cab in Manhattan. Under that assumption, a driver would have to work 69.6 hours per week to hit his 40,000 mile target, which yields a gross fares estimate of $26.10 per hour.4

2014_Taxicab_Factbook_pdf__page_10_of_17_

According to TLC data, average hourly revenues for yellow cabs in NYC range from a low of $26 on Wednesday mornings to a high of $44 on Thursday nights. That means that, if we take Uber’s 40,000 mile number as a given, then an NYC Uber driver grosses about as much in his average hour as a yellow cab grosses in his worst hour.

We can also use this information to make some inferences about utilization, i.e., the amount of time a cab spends with a paying customer inside. Intuitively, Uber’s app should cut down on the amount of time drivers spend searching for fares, thus increasing their utilization and, by extension, their hourly earnings. If we assume that yellow cabs and uberX’s are roughly substitutable, then it’s fair to assume that the average uberX trip is roughly similar to the average yellow cab trip. According to the Factbook, the average yellow cab trip runs 2.6 miles, takes 13.6 minutes, and spends 53% of that time stopped in traffic.5 Plugging these parameters into uberX’s published rates yields an average gross fare of $19.19.

Economics_of_Uber_-_Google_Sheets

Utilization

Looking at these calculations, we can infer that an average uberX fare grosses $1.41 per minute. Multiplying that number by 60, we find that if an uberX were 100% utilized by average fares for an hour, it could gross as much as $84.60. By dividing the average gross fares of $26.10 (calculated above) into $84.60, we reach an estimated average utilization of 31%. Compared with yellow cab utilization of ~50%6, that doesn’t say a ton for Uber’s demand management system. However, I don’t actually believe the number is this bad. More likely, the utilization figure is depressed by the amount of time drivers spend “multihoming,” that is, leaving the Uber app open, but picking up fares either on the street or through a competitor’s app. For drivers, the name of the game is utilization; you want to make every minute as profitable as possible. By using multiple hail networks, the driver increases his chances of finding a fare close to his current location. The less time spent driving to a fare, the more time spent making money.

Vehicle financing

Remember that gross fares are not net fares. After Uber’s 20-30% cut, uberX drivers in NYC can expect to net around $19.58 per hour. And then, of course, you have to factor in costs. Since we’re talking about a full-time, 70-hour a week job here, I think it’s fair to classify all vehicle operating costs as business expenses. However, I don’t think that Uber’s representation of operating costs is as fair as Felix thinks. While they’re reasonable for a 15k/year personal car, they don’t add up for a vehicle that’s getting 40k commercial miles per year, nor do they include the costs of actually buying the car. But if we dig into Uber’s much heralded financing plans launched in late 2013, then perhaps we can get a better sense of the actual costs at play.

First, it’s worth noting that at least one of Uber’s financing partners (Santander) doesn’t offer loans to uberX drivers without good credit; instead, it offers what is effectively a “full purchase” lease, where the lessor can purchase the vehicle at the end of the 4-year lease term for $1. Let’s go out on a limb and assume that the median full-time uberX driver doesn’t have fantastic credit and would rather forego the credit check. In that case, our Camry driver will be paying at least $159 per week (before taxes) on his lease, which works out to $689/month, or $8,268 per year over the four-year lease term, plus a $2,000 upfront payment that’s half capital cost reduction and half security deposit.7

When you add it all up, if you take Uber’s financing option, you’re essentially paying $33,072 ($8,268 x 4) for a car you could buy for $24,750, which works out to an equivalent APR of 15%.8 Meanwhile, a trip to bankrate.com will reveal that today’s market rate for a 48-month loan is under 3%. At least you never have to worry about a late payment; under Uber’s plan Santander will deduct your lease payments from your weekly payouts.

Operating costs

Now that we know the monthly payments ($689), the expected mileage (40k, per both Uber’s communication to Salmon and the Santander microsite) and the lease term (4 years), we can start to see the true cost of being a full-time uberX driver. Below is a Kelly Blue Book estimate for the total cost of owning a 2014 Camry Hybrid LE driven 40,000 miles per year over a 4-year lease:

2014_Toyota_Camry_Hybrid_LE_5-Year_Cost_To_Own_-_Kelley_Blue_Book

Net income

[Edit: It's been pointed out to me that I erred in calculating depreciation expense. What KBB lists for depreciation in Year 1 is tax (rather than economic) depreciation, and it's inappropriate to expense both the payment and an allowance for economic depreciation. In other words, I double-dipped. Instead of $16,909, first-year operating costs should be closer to $6,754. As a result, I understated net income in Year 1 by about $2.63 per hour, and $2.92 thereafter.]

In Year 1, this vehicle could be expected to cost its driver-lessor $16,909 $6,754 in operating costs, plus $8,268 in lease payments, plus a $2,000 upfront payment, for a total hit of $27,177 $18,022. Over a year of 70-hour weeks, that works out to around  or $7.81 $5.18 per hour. This means that, after expenses, a full-time uberX driver in NYC, driving 70-hour weeks in a 2014 Toyota Camry Hybrid leased under Uber’s promoted financing plan, can expect to bring in a pre-tax net income of $19.58 – $7.81 $5.18 = $11.77 $14.40 per hour ($40,959 $50,112 annually) in their first year driving for Uber, and around $14.55 $16.36 per hour ($50,634 $56,934 annually) thereafter.

I should point out that the insurance costs here are surely low — although Uber encourages its uberX drivers to find “standard vehicle insurance” as opposed to livery insurance, the coverages required are nearly impossible to find, especially for someone with poor or no credit working 70 hours a week as a cab driver. Meanwhile, taxi insurance in NYC can run $7,00010,000 per year, which would knock our driver’s net income to an abysmal $9.47 $12.39 per hour in Year 1 ($32,956 $43,117) and $12.25 $15.17 per hour thereafter ($42.630 $52,792).

Once we’ve actually taken all of our operating costs into account, it turns out that our uberX driver is actually middle of the pack, with earnings between the 25th and 75th percentile of his peers. Which, in normal times, is exactly where we’d expect a single-dispatch taxi driver to end up.9

The_economics_of_“everyone’s_private_driver”_—_Medium

Deregulation

The story of the for-hire vehicle industry (FHV) has been one long march toward commoditization, with drivers always getting the short end of an increasingly smaller stick. Since the early 1900s, taxi drivers have morphed from employees (prior to deregulation) to independent contractor-lessors (following deregulation) to sole proprietors (following Uber).10 With each transformation, the industry has shifted profits away from the drivers while pushing onto them a greater share of costs and liabilities. This is why drivers tend to push for medallion systems: because only by capping the supply of vehicles can full-time drivers be assured a living wage. Market equilibrium in a wholly deregulated taxi industry comes only when the desperate have driven out the good. The result is something that few cities would prefer to the imperfect gnarl that is a regulated taxi market.

I recognize that it’s not exactly in fashion to side with regulation over laissez faire, and in many instances, I wouldn’t. There are limits to markets, however, and the taxi industry presents an especially vivid example of that dynamic. We don’t need to have this discussion in the abstract — we actually have hard data. The US went through an era of taxi deregulation in the 1970s, only to follow it with an era of re-regulation. From transport scholar Paul Dempsey’s, Taxi Industry Regulation, Deregulation, and Reregulation: the Paradox of Market Failure (pg. 102):

[W]e need not rely on the theoretical assumptions of what unlimited entry will produce. We have empirical results which we can assess to determine what deregulation of the taxicab industry has produced.

Before 1983, some twenty-one cities deregulated taxicabs in whole or part. The experiences of these cities reveal that taxicab deregulation resulted in:

1. A significant increase in new entry;
2. A decline in operational efficiency and productivity;
3. An increase in highway congestion, energy consumption and environmental pollution;
4. An increase in rates;
5. A decline in driver income;
6. A deterioration in service; and
7. Little or no improvement in administrative costs.

I recommend reading the whole paper, then reading this one from the Journal of Transport Economics and Policy (“In every city where the taxicab industry has been deregulated, there has been a significant decline in taxicab productivity as measured by the number of daily trips per cab and trips per shift.”) and finally this one from Price Waterhouse (“the effects of taxi industry deregulation have ranged from benign to adverse”). Theoretical models are important, but they should be shaped by data whenever possible.

Beautiful illusions

I can’t think of any market with more distorted supply and demand curves than the taxi industry today. Billions of dollars are pouring into this industry, and they’re not going to capital investments like R&D or PP&E; instead, they’re covering massive incentive expenses for both drivers and passengers. You could look at what’s happening and plausibly call it a wealth transfer from investors to consumers, and you’d probably be more correct than if you had called it a well-functioning market. Knowing that, I’m not sure how you can take any number reported out of this industry at face value, let alone extrapolate future trends off of them.

What we’re seeing is the very definition of an artificial market environment subsidized by huge infusions of outside capital. But despite the obvious distortionary effects of that capital, we’re treating these trends like they’re secular. Massively profitable cash flow businesses don’t need billion dollar capital infusions every twelve months. Strong brands with differentiated products don’t engage in race-to-the-bottom price wars. And businesses with great unit economics don’t need to increase their unit sales by 5x to grow revenues by 2x.11

“Uber’s brand will support high margins” was a plausible story when the company was focused on the luxury segment, but clearly that focus has shifted to the lower end of the market. If Uber raised the price of uberX by 25%, do you honestly believe that its customers wouldn’t leave in droves? If drivers thought they could make more money on another dispatch network, do you honestly believe that they wouldn’t switch? Until a mass market taxi company demonstrates that it can raise fares over a sustained period without suffering an offsetting reduction in demand, I don’t see how it’s reasonable to believe that auto transport is any less a commodity than air travel. And yet, that’s exactly the assumption we’re making.

Conclusion

Ultimately, the most anyone can do is make reasonable assumptions and think rigorously, and I’ve tried to do that here. I’m sure I’m wrong about many things in this post. But from where I’m sitting, the math just doesn’t work.12

It would be easy to look at this post and say I don’t get it. That I’m anti-innovation and a known Uber bear and I can’t hear Jimi. I don’t really have a good response to that. All I can say is, when I do the math and make reasonable assumptions based on available research, I don’t see where the excess profits are supposed to come from. I don’t mean to reason defensively — quite the opposite, in fact — but the numbers and industry dynamics lead me to the same conclusions as before. You are, as always, free to disagree.

  1. I should point out that Felix is wrong to assume that most cab drivers are employees. In fact, most are independent contractors. Indeed, the taxi industry was decades ahead of Uber in realizing the benefits of shifting operating costs and liabilities as far away from the profit-taking organization as possible. []
  2. If you are interested in a rigorous valuation, I’d suggest you read through Aswath Damodaran’s excellent analysis. []
  3. 2014 Taxicab Factbook, pg. 7, fn. 1. []
  4. 40k miles per year / 11.5 miles per hour / 50 weeks = 69.6 hours per week. $90,766 / (69.6 hours per week x 50 weeks) = $26.10 per hour. []
  5. See this spreadsheet for the reasoning behind my stopped-time assumption. []
  6. Factbook, pg. 8. []
  7. See Santander’s FAQ. Funny enough, these awful terms  are kind of a deal when compared with the terms laid out in the financing information sheet given out to drivers in Atlanta. []
  8. You can work this out easily using Excel’s RATE function. If nper = 48, pmt = $689, and pv = -24,750, then RATE(nper, pmt, pv) returns a monthly interest rate of 1.25%. Multiply that by 12 and you’ve got your APR. []
  9. This all assumes, of course, that Uber doesn’t further reduce the price of fares or raise its rake, or boot the driver from its system without process. []
  10. If you’re curious about the history and evolution of the taxi industry, I suggest you pick up a copy of Taxi! Urban Economies and Social and Transport Impacts of Taxicabs. It’s a fantastic read, rich with data and historical detail. It will also disabuse you of a number of facile narratives that have dominated the conversation recently. []
  11. This statement alone would set off alarm bells in a normal environment. []
  12. If my math is wrong, please tell me! I strongly believe that being proven wrong is often the shortest path to being right. []

Play, Money

Founder CEOs are playing with their own money when they make an offer.

Jason M. Lemkin’s answer to What do you say when a potential acquirer asks you what amount of money your startup is worth? – Quora

This is sort of a throwaway line in an otherwise reasonable answer, but I’m going to call it out regardless.

Facebook is unusual for its utter lack of meaningful corporate governance. Mark Zuckerberg could bid $100B for TimeWarner tomorrow and there’s not a thing shareholders could do to stop him. So, from a pure control standpoint, it is kind of like he is playing with his own money.1 But he’s the exception, not the rule.

Generally speaking, the balance sheet of a public company is not a toy for CEOs — even founder CEOs — to play with. Unless you’ve managed to build a benevolent dictatorship into your corporate bylaws, that cash belongs to your shareholders and you are its custodian. Now, one could reasonably argue that founder CEOs, by acting as if they’re playing with their own money, generally advance their shareholders’ best interests. That’s a perfectly valid (if wildly overbroad) argument to make. But it’s also a very different assertion than “Founder CEOs are playing with their own money,” and the distinction is incredibly important.

  1. Incidentally, while it’s SV gospel that control gives founders the freedom to ignore short-term performance and invest for the long-term, available evidence indicates that it accomplishes exactly the opposite. []

Sincerely, Debbie Downer

The problem with being so dependent on processing revenue is twofold: First, the profit margins aren’t great. Competition and the commoditization of payment pipes have driven down the fees payments companies charge merchants, while the cut paid out to credit card companies has remained largely the same. Secondly, the public markets today would likely be forced to value Square at a lower revenue multiple than if it made revenue from software sales, which typically come with better profit margins.

Google, Apple or IPO: What’s Next for Square? | Re/code

It’s not that the margins aren’t great, it’s that a merchant acquirer can only take so much off the top before either the merchants walk or Visa/MC/Amex lock you out. One of the great misconceptions of this investment cycle has been to treat processing volume as revenues (see: Uber, AirBnb, Square). But revenue isn’t the money a processor handles, it’s the money it handles on which it has a claim. In Square’s case, as in the case of most all merchant acquirers, that number is around 0.5% of transaction volume, which for $30b in transactions yields ~$150m in net revenues. That’s nothing to sneeze at, surely, but it’s also not Earth-shattering.

The public markets value merchant acquirers at anywhere from 1/50th to 1/100th of transaction volume. If you believe Square is worth $8b, then you should also believe that it will reach $400 – 800b in transaction volume. And that’s just to break even; if you expect a positive ROI, those numbers have to go even higher. Now, that’s certainly possible, but I’ve yet to see an evidence-based argument explaining how Square can capture 10-20+% of the total US transaction volume, especially considering the strength of its competition.

Payment volumes make for misleading, but exciting, headlines, which is exactly why they’re used so often. I like Square. I think offering payment processing services to individuals and small businesses who otherwise face discouragingly high barriers to those services is a terrific business idea. I think their execution on that idea has been excellent. But I also think the valuations being assigned to the business–and the qualitative justifications undergirding those valuations–are just plain silly, and very much a sign of the times.

Sincerely,

Debbie Downer

Unintended Consequences

In passing the Communications Decency Act of 1996, which, among other things, established immunity for internet service providers for publishing “information provided by another information content provider,” 47 U.S.C. § 230(c)(1), the House explicitly stated its intent to overturn the result reached in the Prodigy case. See H.R. Conf. Rep. 104-58, at 194.

Stratton Oakmont v. Prodigy | Digital Media Law Project

Interesting footnote for those who’ve seen Wolf of Wall Street. Stratton Oakmont also played a major role in the early history of online publishing. In 1994, an anonymous user posting to Prodigy’s Money Talk message board made a number of fraud allegations against Stratton Oakmont. S-O defended themselves vigorously (oopsies) and brought suit against Prodigy as “publisher” the allegedly defamatory comments. The court found in S-O’s favor, ruling that Prodigy was a publisher in this instance and strictly liable for its site content, regardless of who wrote it. However, the ruling’s line of reasoning conflicted with the earlier case of Cubby v. CompuServe, which held that in a similar situation CompuServe was not a publisher because it exercised absolutely no editorial control over third-party content.

For a while, it looked like liability was going to attach to websites’ editorial processes, which acted in some small way to bifurcate the content sites into two flavors: (1) sites that acted like newspapers, and (2) sites that acted like truck stop bulletin boards. The conflict was soon resolved in the Telecommunications Act of ‘96, which granted safe harbor to ISPs for defamatory statements published by users. Ideally, the protection would’ve increased overall editorial quality by letting sites rely on intermediate levels of content control. Whether that happened is an open question.

Also interesting, the section of the Act that offered safe harbor is commonly known as the Communications Decency Act, which has the distinction of being the first concerted Congressional effort to regulate speech and obscenity online. Unfortunately for those seeking regulation, the anti-indecency provisions were soon ruled unconstitutional by SCOTUS in Reno v. ACLU, which left safe harbor for content providers as the sole surviving piece of the CDA.

Talk about unintended consequences.

It’s interesting to note just how quickly the scales tipped last time around. We tend to think of manias in linear terms, assuming that so long as things grow slow and steady that the market’s acting in a healthy manner. But manias, like other systemic contagions, are nonlinear. They only happen once all the pieces have been moved into place. In January 1999, IPO growth looked deliberate and fundamentally sound. Twelve months later, it was hard to hold that line of reasoning (though of course, many tried).

I don’t know when we’ll enter the non-linear phase of this run-up, but it will happen. I don’t say that because I doubt the fundamental value of the companies coming to market (some I do, others I don’t), but simply because manias are inherent to our nature. Optimism, especially techno-optimism, is especially viral, and I’m not so sure that’s a bad thing. Sure, every 5-7 years it brings pain, and it never does live up to its promise, but the hope, innovation, and camaraderie it breeds in the interim may well be worth that penalty. We need something to believe in, after all, and flying cars (or delivery drones) beat the slow march of steady progress any day.

So there it is: we’re trading hope for volatility, and that’s a trade I’m ok with, even if it’s one I’m temperamentally indisposed to. I just wish more people saw it for what it is.

(Image via Absurdly High Valuations – 42Floors)

How Big of a Deal is Uber?

N.B. This is a repost of an answer first posted to Quora.

I. Taxi & Limousine Industry Overview

Let’s start by looking at the highlights of Uber’s current industry. According to IBISWorld, the US Taxi & Limousine Services industry will pull down $9.7 billion in revenue in 2013 across five major segments: taxi fares (55%), leases to operators, black cars, limousines, and other, which includes in-cab advertising, party buses, and maintenance.

That’s the good news. The bad news is that revenues stagnated a long time ago. Five-year growth is an anemic 0.8%, and the next five years don’t look much better. Even worse, revenues are widely distributed amongst nearly 200,000 players, with no one firm capturing >1% of the market. In business-y terms, that’s considered “extremely fragmented.”

II. Business Model

So, what is Uber? Is it a town car company? A taxi company? A personal logistics service? A car replacement? Indeed, it could be a strip of options that subsumes all of these, or it could be something else entirely. Generally speaking though, its dominant dynamic is most likely one of the following:

  • A taxi authority and call-center lessor
  • A multi-sided platform enabling a two-sided market
  • A brokerage

Now, fair warning: business models are inherently indistinct. At best, they’re carefully conceptualized mappings of the relationships within a firm’s market environment. At worst, they’re seductive narratives that beguile us with their coherence and optimism while bearing only the shallowest relationship to reality. An accurate conception of a firm’s business model provides the foundation for sound analysis. An inaccurate conception provides the opportunity for mental masturbation. With that in mind, let’s examine these options in turn.

Authority and call-center lessor
Under the authority and call-center lessor model (a standard model within the T&LS industry), firms lease out to independent-contractor drivers the right to operate under a proprietary dispatch network. Drivers are responsible for owning and maintaining their vehicles and usually their insurance. Leases are generally offered at a fixed amount per shift, with the driver paying the lessor upfront and retaining all earnings made during the lease period. The main attractions of this model are: (i) low capital intensity, (ii) minimal liability, and (iii) minimal volatility in daily revenues.

At first blush, this model bears a striking similarity to Uber, but there is at least one important difference. Uber does not lease out the right to operate in its network; instead, drivers pay Uber a portion of each fare they collect while using Uber’s dispatch network. Importantly, this arrangement exposes Uber to a great deal of demand-side volatility that would otherwise be spread amongst their drivers. The variance also creates a meaningful difference in incentives. Whereas the typical authority has little incentive to invest in dispatch technology, in-car payment systems, or advertising, Uber has a perfect incentive to undertake such investments, as every incremental fare increases Uber’s take proportionately. For this reason alone, I’d say that Uber is not an authority and call-center lessor, and so will not analyze its business under that model.

Multi-sided platform (MSP)
Although there’s no settled definition of what constitutes an MSP or two-sided market, the literature tends to focus on a common set of dynamics:

  1. direct interaction between heterodox groups of users
  2. positive network effects
  3. value lock-in (high barriers to competitive entry and/or high multihoming costs for users).1

Much of the excitement around Uber derives from the idea that it’s a two-sided marketplace for poorly utilized assets. If Uber is such a marketplace, and if it has the “winner-take-all” characteristics that such marketplaces often exhibit, than its comps include the sort of sharing economy/collaborative consumption companies that VC dreams are made of (e.g., Airbnb). So obviously this is an important issue to settle. With that in mind, let’s examine each of these dynamics in turn.

  • Direct interaction between heterodox groups of users. Uber clearly exhibits this dynamic. Passengers and drivers are distinct groups with distinct needs, communicating directly via Uber’s app platform.
  • Positive network effects. While some positive network effects do exist, I would argue that Uber’s business is actually dominated by negative network effects. As anyone who’s tried to order up an Uber during one of their land-grabs can attest, adding users to the platform quickly leads to a worse experience for everyone. Unlike information, a cab ride is an exclusive product — if you’re in a cab, then I can’t call it, which means your usage degrades my experience. What’s more, a cab ride is a commodity product, which means that every passenger (and driver) in an area is competing with their peers. Again, this implies that adding users subtracts value from existing users of the network.
  • Value lock-in. Let’s be clear: Uber does not have any. There are no real barriers to entry in the cab market, and that’s doubly true for the asset-light dispatch services sub-market. The problem is the product. Not only are cab rides commodity products differentiated almost solely on the basis of availability, they’re further perceived (at least to Millenials) as a universally inferior mode of transport.23 You can talk all you want about what a great experience Uber is, but the fact remains that, on balance, when people need a cab, they care about getting from A to B far more than the “experience.” If an Uber’s not available, people have a plethora of perfect substitutes at their disposal. And thanks to a bunch of evil, innovation-crushing regulations, they can make their decision without regard to unsafe conditions or price gouging. Add it all up, and you can understand why the T&LS market remains so incredibly fragmented, and why Uber’s chances of changing that dynamic are most likely nonexistent.

In conclusion, Uber is an MSP, but because users are free to use whichever dispatch service they’d like, and because there’s no penalty for multihoming, Uber is not meaningfully shielded from competition by either network effects or multihoming costs. Therefore, it should not be analyzed in the context of a “winner-take-all” market dynamic. While the MSP model should certainly inform Uber’s business strategy, it’s less helpful in our quest to determine Uber’s value as a business.

Brokerage
Brokers arrange transactions between buyers and sellers in exchange for a commission paid upon execution. They often facilitate these transactions by supplying information on market conditions, prices, and available products. They also offer their clients access to prospective customers with whom they’ve already formed relationships. Brokers tend to be judged by the quantity and quality of their customer relationships, as they speak to the broker’s ability to consummate transactions.

In my opinion, Uber is best analyzed as a non-exclusive broker operating on behalf of drivers. Drivers use Uber’s dispatch platform as a means of finding passengers, and by handling the payment infrastructure, Uber is ensured a commission from the driver on the transactions it effects. However, neither the driver nor passenger commits to Uber exclusively; indeed, should an Uber-driver and an Uber-passenger meet without the help of Uber’s platform, they’re under no obligation to pay Uber a dime.

III. Sizing the Market

Total Available Market (TAM)
TAM measures the theoretical revenue opportunity for a product or service. For Uber, I define TAM as fare revenue for taxis and black cars. According to the revenue segmentation below, that comes to about $6 billion.


Serviceable Available Market (SAM)

I’ve yet to find an authoritative definition of SAM, but I tend to think of it as the portion of the TAM a company can reasonably expect to reach given its business model. As an app-based platform, Uber is aimed at pre-arranged taxi services.4 By some estimates, pre-arranged services account for more than 80% of overall demand nationwide.2 So let’s be generous and say that Uber’s SAM (i.e., cabs not hailed on the street or at a cab stand) is 90% of its TAM, for an SAM of $5.4 billion.

Serviceable Obtainable Market (SOM)
There are many ways to derive this number, but let’s go with the simplest: feasible market share. Personally, I would estimate that the best possible outcome for Uber would be to capture 5% of the market. My reasons are as follows:

  • The taxi industry is extremely fragmented. The top four operators combined generated less than 3% of total industry revenues in 2013.
  • Uber has no structural advantages (i.e., moats). Because their drivers are independent contractors, Uber has no real means of forcing exclusivity on the supply side. Passengers, meanwhile, are free to use any cab company they like, using whatever means of hailing they please.
  • Uber’s brand excludes over 50% of riders. Uber is quite clearly positioned as a corporate/up-market option for riders. Given that corporate ridership accounts for 60% of black car revenue and 24% of taxi revenue, that seems like a good decision. But taxicab ridership follows a bimodal distribution: those who use cabs the most are either very rich or very poor. In fact, households with <$50k in income (44% of the population) use taxis at about the same frequency (relative to overall trips) as households with >$125k in income (13% of the population).25 Furthermore, households with zero or one cars — i.e., low-income people and Manhattanites — make 5-10x more trips by taxi (as a percentage of total trips) than households with two cars or more. Toting it all up, I’d estimate that mid- and up-market taxi riders account for less than 50% of all taxi riders.
  • Local operators know how to compete with a national brand. 1-800-TAXICAB is a cooperative of independent operators who rely on a national endpoint brand to drive customers to local members. Coordination in response to the brand localization problem operates to blunt the effectiveness of Uber’s national branding strategy.
  • E-hail technology is fully commoditized. Don’t believe me? Go ahead and search for “taxi” in iTunes or Google Play.
  • Mobile payments don’t make a (meaningful) difference in dispatch decisions. If you really want to argue with me on this point, you better come with data, and your sample better cover more than SF/SV.

So, assuming that Uber can capture 5% of its SAM (a generous assumption for all the reasons above), that leads us to an SAM of $162 million.

Revenue model
It’s easy to look at that $162 million number and get excited, but remember that we’re talking about $162 million in fare revenue. As a broker, Uber will only ever see a small portion of that amount. According to Xconomy, Uber takes 10% of the base fare plus a $1 per ride surcharge from cab drivers, and 20% of the base fare for black cars.6 Again, let’s be generous and assume that Uber can get its average cut up to 18%. This implies a best-case total US revenue opportunity of $29 million per year.

IV. What if…

Obviously, this number can be made bigger in any number of ways. International expansion is an easy assumption to make, but before assuming too much there, I’d recommend reading up on the cultural dynamics surrounding taxicab usage in other countries. The decision of whether to take a cab — and the subsequent decision of which type of cab to take — depends very much on the availability or lack of substitute forms of transportation. You can’t just say things like “London is a reasonably dense metro, so that’s a big opportunity!” and leave the rest to the imagination. Narratives are fun, but they have to match with the underlying infrastructural, cultural, socioeconomic, and psychological realities.

Similarly, we could grow this market by assuming that increasing efficiency will increase demand, but there’s no evidence to suggest that increasing supply-side efficiency will lead to increased travel demand. Nor is their evidence that society is waiting for a shared “replacement” for personal autos. I’ve heard it argued that Uber frees people to travel by car more often because it’s more convenient. While that may be true in the parking desert that is SF, research suggests that owning a car causes people to increase their travel frequency by more than 2.5x. So far as I can tell, there is no evidence that people actually want to replace their cars, regardless of how great Uber’s UX may be.

Finally, we could assume that Uber will become the switchboard for all things transit, a “Cisco for the physical world” as Michael Wolfe puts it. While I applaud the imagination, the evidence is lacking. And by evidence, I don’t mean a smoking gun, I mean signals and indicators. What is it in our psychology that will draw us toward such a future? Do the fundamental economics of the physical world lend themselves to such an enterprise? What about our regulatory or liability regimes? These questions are rhetorical, but my point remains: our world is complex and chaotic, and certainly not subject to facile narratives drawn from shallow analogies.

V. Conclusion

So how big of a deal is Uber? Well, to seed and angel investors who sold shares in the last round, I’m sure it’s a very big deal indeed. It’s also a big (if temporary) deal to passengers in land-grab cities like DC and SF who get to take highly subsidized cab rides for as long as Uber’s war chest lasts. Of course, as soon as those subsidies run out, it will be just another transport option, but so it goes. There’s nothing wrong with a business that tops out at under $100 million in net revenue with fat margins — that’s pretty amazing, in fact,. Probably not $3.5-billion-plus-a-reasonable-return amazing, but still pretty great. Ultimately, the question depends on you. If five years ago you spent your weekends gently sobbing while cab after cab failed to show up, and now you don’t, then Uber’s a very big deal indeed. If you live in Manhattan and make around $100k, then it probably doesn’t register at all.

VI. One Last Note

I’ve made a boatload of assumptions during this exercise and you are free to disagree with any or all of them. But remember that we’re not talking about a seed-stage company anymore. It’s a real company with established business lines and actual revenue, and that requires more than a facile narrative to value.

  1. Note that these elements do not constitute a dispositive test; rather, they form a useful framework for evaluation. Our understanding of network effects, especially those that connect the internet with the physical world, is still very much a work in progress. []
  2. The Taxi – Friend or Foe, page 15 [] [] []
  3. Millenials & Mobility: Understanding the Millenial Mindset []
  4. Taxi service comes in two flavors: on-demand and pre-arranged. On-demand service includes street hails and taxi-stand pickups. Pre-arranged service encompasses rides arranged in advance of pickup. Operators can and do provide both types of service. []
  5. Income, Poverty and Health Insurance in the United States: 2012 []
  6. Uber’s Fare Numbers in Boston Revealed, Barely Dent Cab Market []

Things Not Seen

In Thinking Far and Slow, Daniel Kahneman describes a cognitive phenomenon he terms, “what you see is all there is.” The idea is that we fail to account for the things we don’t know, even when we *know* that we’re only seeing a small piece of the bigger picture. One consequence of this shortcoming is that we tend to place undue weight on the representativeness of our own experiences, leading us to believe that our problems are everyone’s.

Unfortunately, if you’re a student at an Ivy League college in Manhattan, chances are your problems are not representative at all. I know Paul Graham tells people to build companies that scratch a personal itch, but here’s the thing: to a venture investor, the things that make you itch right now just aren’t all that interesting. In all likelihood, they’re not very lucrative either. Sure, maybe you’ll invent the next great flip opportunity, or become the next Zuckerberg, Levie, Houston, or Gates — but odds are you won’t. And we’re not talking 100:1 here, we’re talking odds so long that they are effectively zero. Mathematically speaking, you’re better off playing the lottery. So the million-dollar question becomes: how do you increase your odds? Well…

Work for a company that makes lots of money — companies that make money are generally solving the problems that people care enough to spend money on. If you want to become a social entrepreneur, substitute “impact” for “money” — the song remains the same.

Note with particularity the traits you admire in your co-workers — this information will prove essential when the time comes to build your team. Build a network, within your company and without. No one has built a startup without help.

Gather feedback, positive and negative, from co-workers whose judgment you trust. You have to understand your weaknesses if you hope to overcome them. In the same vein, spend weekends trading notes with your more entrepreneurial-minded friends. See whose notes you listen to and whose you ignore.

Save as much money as you can. Trust me, you’ll need the cushion. There’s something poetic in the image of the young founder living on Ramen noodles and credit cards, testing and iterating his way to success. In reality, starvation and seclusion are counterproductive and stupid. More than effort, startups require insight, and insight is built on a foundation of knowledge.

And therein lies the problem: without real world business experience, all you know is what you’ve heard or read, and the world’s a whole lot bigger than that. So go out there and grow your world. What you see is not all there is, so keep looking for more. Soon enough, you’ll find an itch worth scratching — for those who truly want to be founders, it’s only a matter of time.

This post originally appeared in the Columbia Spectator as part of a series for Columbia’s Startup Week.

All of the Smart People

Have you ever wondered how Wall Street became Wall Street? It’s quite simple really: they hired all of the smart people. And I don’t mean “all of the smart people with a passion for stocks” or “all of the smart people with experience in banking.” They started with the business schools, and when they ran out of smart MBAs they moved right on to smart non-MBAs. And when they finally filled their quotas, they took all those smart non-MBAs and did something truly crazy: they trained them. With good mentors and great paychecks, the banks molded those philosophy, comparative literature, and political science majors into financial wizards, and over the next few years those wizards repaid their employers’ investment with ridiculous interest. And that’s how Wall Street discovered what law schools and consulting firms had known for years: when properly motivated, smart people can learn things.

Today, though, things work differently. Employees aren’t trained so much as refined. Entry-level job specs ask for — and get — previous experience and expertise. And yet, at the same time companies are holding new graduates to mid-career standards, those companies are running to Congress and universities alike to complain about a lack of available talent. But if the talent well is in fact running dry, there’s no need to rely on solutions that run through Congress. Instead, do what growth industries have been doing for generations: find all the smart people, and train them.

Nothing is that simple, of course. Paradigms don’t shift overnight, especially when those paradigms produced the people in charge. It’s hard to imagine how someone else would do our job, let alone someone with different set of skills or experience. On the other hand, it’s quite easy to imagine a mini-me handling our workload. Consequently, when we sit down to write out a job spec, we instead write out an autobiography. Clay Shirky once wrote something along the lines of “organizations tend to preserve the problems to which they are the solution.” Well, that’s really a rollup of a more generalized problem, that people tend to preserve the systems that validate their life-choices. Good coders who happen to have gone to Stanford will seek out Stanford graduates who happen to code. As coincidence becomes condition, hiring managers increasingly mistake proxies for the traits they’re meant to signal, and an inevitable narrowing of opportunity results. Pretty soon, you start to hear great employees making offhand comments about how they “could never get hired today.” But such comments should terrify the HR managers at “innovative” companies. Mandarinization is a wonderful strategy for maintaining the status quo at equilibrium, but it’s hardly the best choice for a company hoping to live at the bleeding edge of disruption.

Fortunately, some companies are starting to push back. Google’s VP of HR recently went on record that many of their most famous hiring practices lacked any value at all. It turns out that tech interview standards like brain teasers actually do more to make interviewers feel smart than to help them assess candidates. GPAs and test scores? Hardly a useful predictor. The problem, he notes, is that the skills someone develops over their first few years in the workplace renders them a fundamentally different person. Turns out that people tend to learn things at work. Certain resurgent hiring practices — specifically behavioral interviewing and criterion-sampled job descriptions — can help companies cast wider nets over the talent pool, which when combined with a commitment to on-the-job training (possibly through a half-salary apprenticeship) offers a feasible solution to the engineering talent shortage.

We all know that engineering isn’t for the faint of heart, but if you think that top-shelf attorneys (or hedgies or consultants for that matter) don’t have the minds to program, then you really need to get out more.1 The people who thrive in professional services are as logical, insightful, and detailed of thinkers as the best engineers, they’ve simply applied their talents toward a different end. But there was a point not long ago when they could have easily chosen to apply their talents to software engineering, if only Google, Facebook, or some other tech giant had been waiting with open arms and the promise of an interesting future. Increased emphasis on STEM education and comprehensive immigration reform are worthy issues, but if Big Tech can’t find the engineers it needs, there’s no need to wait on Congress. Be the change you want to see in your industry: find all of the smart people, and train them.

  1. As an aside, I don’t completely understand the malice developers direct toward attorneys, especially given their similarities. I had a professor who described good legal writing as writing “free from irony.” What he meant was, good legal writing can only be interpreted one way — it permits no ambiguity. Unfortunately, that standard is impossibly hard to achieve. I know that legal documents look redundant, but consider this: how would your code look if you were forced to write in an ambiguous language for an unpredictable parser? []

True Histories: ESOPs

There are fifty or so billionaires and tens of thousands of millionaires in Silicon Valley. Think about that for a second: tens of thousands of millionaires, almost all them created by companies that didn’t exist two decades ago.

The defining difference between Silicon Valley companies and almost every other industry in the U.S. is the virtually universal practice among tech companies of distributing meaningful equity (usually in the form of stock options) to ordinary employees. Before companies like Fairchild and Hewlett-Packard began the practice fifty years ago, distributing stock options to anyone other than top management was virtually unheard of. But the engineering tradition that spawned Silicon Valley was much more egalitarian than traditional corporate culture.

Steven Johnson (via brycedotvc)


This would be nice if it were true, but ESOP plans existed (and were used often) long before Silicon Valley. To be fair, the modern ESOP was invented in SF, but at Peninsula Newspapers, not Fairchild. It’s more accurate to say that SV perfected the art of distributing wealth historically concentrated in one or two multi-digit billionaires across the bank accounts of four or five single-digit billionaires and a few multimillionaires to boot. An improvement for sure, but not quite the egalitarian miracle we’d like to believe.

That’s not to say that SV doesn’t have its structural advantages. For instance, there’s ample evidence that California’s distaste for non-compete clauses has played a very large role in productivity enhancement, which makes sense if you believe (as Steven Johnson does) that innovation derives from the continuous interplay of ideas. But does anybody in SV actually believe that technological innovation is driven solely (or even mostly) by the profit incentive? Because the evidence (at least according to TED) suggests otherwise…



Copyright © 2009–2013. All rights reserved.

RSS Feed. This blog is proudly powered by Wordpress and uses Modern Clix, a theme by Rodrigo Galindez that's since been screwed up by Justin.