Calculating losses

A man walks into a widget shop with the intention of buying a widget. He picks

out a nice widget for $100, but just as he’s about to take it to the counter

to pay for it, he gets an urgent phone call and rushes out of the shop, which

has a 30% profit margin on its widgets. The question is: How much has the shop

lost because of the phone call?

a) The shop has lost nothing. A man walked in, a man walked out. Where’s the


b) The shop has lost $30: the profit it would have made on the sale of the widget.

c) The shop has lost $100: the amount of money it would have received had the

man not received the phone call.

I’m inclined to go with (a) on this one. But let’s say that the shop starts

seeing a pattern: people walk in and always seem to get urgent phone calls just

before they buy their widgets. The manager of the shop calls me up, and says

"we’re losing thousands of dollars a month because of these urgent phone

calls". I can see that, too.

As long as this kind of mathematics stays on the anecdotal level, none of it

really matters. But when corporations spend millions of dollars to work out

exactly what they’re "losing", and newspapers write long articles

on the subject, then it’s worth revisiting the question of what losses, really,


In case you haven’t worked out where I’m going with this, it’s my favourite

subject: counterfeiting. Today’s question (or yesterday’s, actually: we’ll get

to that in a minute) is this: if a Zippo lighter costs $25, and a counterfeiter

sells a fake Zippo for 37 cents, how much has Zippo lost?

a) Nothing. If it’s not there on the P&L, it’s not a loss.

b) $25.

c) $25 muliplied by p, where p is the probability that the

person buying the counterfeit Zippo would have bought a real Zippo were the

counterfeit Zippo not available.

d) The answer from (c), multiplied by the profit margin on a $25 Zippo. Or,

to put it another way: if Zippo makes $10 in profit on every $25 lighter, then

$10 multiplied by p.

Which brings us to yesterday’s news: two big articles fronting the Marketplace

section of the Wall Street Journal. I don’t have an online subscription, but


a link that Google says leads to a story entitled "China Policy Lets Counterfeiters

Off Lightly", and here’s

a link to the other article, entitled "Estimates of Copyright Piracy Losses

Vary Widely".

Let’s take the second article first. "Counting the losses from piracy

isn’t a science – it’s an art," says Geoffrey Fowler, who even quotes

Keith Jopling, at the International Federation of the Phonographic Industry,

saying that "we will never know 100% the exact loss". Fowler’s article

is reasonably fair: it even makes the point that some people who download music

illegally actually buy more money on legitimate music as a result.

As Sasha Frere-Jones points

out in last week’s New Yorker,

The Arctic Monkeys built their audience by playing live shows—constantly,

all over England—and by giving away its songs as MP3s on

When their remarkable album, “Whatever People Say I Am, That’s

What I’m Not,” was released, in January, it sold more than three

hundred and fifty thousand copies in the first week, making it the fastest-selling

début in British history. (So much for the idea that giving away your

music hurts sales.)

Still, there’s a difference between an artist using the internet to market

their work, and a counterfeiter selling bootleg CDs. In the latter case, a criminal

is making money from selling intellectual property which he doesn’t own. But

it’s the next logical step which gives me pause: the idea that if a criminal

is making money, then somebody else must be losing money.

Theft of IP is theft, by this logic, and if you steal something from someone,

then they lose something of value.

In the case of IP, however, the owner of the IP remains the owner of the IP

even after it’s been stolen. It’s not like anybody ever steals a copyright.

So I’m not convinced that the sale of a counterfeit Zippo, or the sale of a

bootleg CD, actually causes any measurable loss at all to Zippo or to EMI.

And that’s where Fowler’s article is weak: he implies that the uncertainty

about losses due to piracy is epistemological rather than ontological. The only

question, in his mind, is over how big they are, not whether or not they exist.

But to give you an idea of how meaningless most of these numbers are, let’s

say that the MPAA is right when it reports that Hollywood studios collectively

lose $6.1 billion per year to piracy. What does that mean?

a) If there weren’t any piracy, then the profits of Hollywood studios would

be $6.1 billion greater than they are now.

b) If there weren’t any piracy, then the gross revenues of Hollywood studios

would be $6.1 billion greater than they are now.

Judging by Fowler’s article, it’s the latter: he notes at one point that "not

every pirated disc equates to lost revenue". So what would $6.1

billion in lost revenue equate to in terms of lost profit? Now there’s

an unanswerable question. But let’s use Disney’s profit margin of 8.22%: it

works out at almost exactly $500 million, which, divided among all the Hollywood

studios, seems a big but hardly enormous number. Is this really the statistic

that some studios were afraid of releasing for fear that its magnitude was so

great it might spook the markets?

And then there’s this very silly bit near the end of the article:

The Business Software Alliance’s Asia regional director, Jeff Hardee, says

the group’s studies show a nearly one-to-one ratio between decreases in piracy

and increases in legitimate sales, suggesting that, on average, people are

willing to "replace" pirate software with the real thing.

Which sounds very impressive until you stop to wonder what these "decreases

in piracy" are, and how they might be measured. It turns out that the BSA

measures piracy by a very simple method: take the retail value of all the software

installed on all the computers in the world, subtract total global software

sales, and the rest is piracy. Now I very much doubt that there’s any market

in the world which has seen the total value of software installed ever go down.

Which means that if the BSA measures a decrease in piracy, then the only way

that’s going to happen is if legitimate sales rise. No wonder there’s "a

nearly one-to-one ratio between decreases in piracy and increases in legitimate


Now, let’s return to those Zippos, as reported by Nicholas Zamiska. The story

is this: Chinese authorities raided a factory which turned out to have 32,980

fake Zippo lighters. Here’s Zamiska:

The problem stems from the way China values seized knockoffs. Unlike many

wealthier countries, China values them at the price the counterfeiter meant

to sell them at, not at the retail price that legitimate versions go for.

Yes, that’s "the problem".

To be fair, Zamiska does report both sides of the story. He gives the Chinese

view, which seems sensible to me:

Chinese courts have defended the practice by arguing that valuing seized

goods at the prices they would fetch in developed countries would grossly

distort penalties for counterfeiters, who stand to make a pittance in comparison.

At a news conference in April, Xiong Xuan Gao, the vice president of China’s

Supreme People’s Court, brought up an example of counterfeit watches, pointing

out that even though a genuine luxury watch might be worth more than $12,500,

the vendor sells it for only about $1.25.

"The sentence of conviction and criminal punishment should be measured

by the real profits he has made for himself, from the viewpoint of social

harmfulness," Mr Xiong said, according to an official media report. "On

one hand, we should make more efforts on cracking down on the infringing upon

intellectual property rights; on the other hand, we also need to be fair on

the suspects and protect their own rights."

The main thrust of the article, however, is that this policy is extremely misguided

and harmful.

My view is that China, with its weak IP laws, is actually much better placed

to be an economic powerhouse in the 21st Century than economies which stifle

innovation with draconian copyright laws. The Zippo lighter is a design icon

because of its beauty and simplicity – in that respect it’s very similar

to the Noguchi table which

I’ve blogged about in the past. We live in a world where great simple design

can and will be brought to the masses cheaply. The successful companies of the

future will be the ones who embrace that fact; the failures will be those who

litigate against it. Law might be on their side, but history will not be.

This entry was posted in Uncategorized. Bookmark the permalink.

4 Responses to Calculating losses

  1. john t unger says:

    As an artist I get inquiries pretty much every day about commissions or existing work for sale. Most of these conversations end with the potential customer telling me that they definitely intend to buy or commission work. About 1 in 10 of the people I talk to actually end up writing a check. That seems to be pretty much the going conversion rate for most of the other people I know who do custom work as well.

    So am I losing 90% of my income? Um, no. And do I plan my spending before I get the check in hand? Again, no. I tried that out the first month I was in business and it was almost the end of me… No matter how much interest someone professes, you can’t spend the money til you have it. Pretty basic.

    It’s always one of the most painful early lessons for a new business owner that most people are just talking. I’m sure it probably happens in real estate, law, and lots of other fields as well.

    I’m not saying this is the same exactly as failing to make a sale due to counterfeiting, but it did occur to me as another anecdotal way to look at the situation.

    Yeah, I could *say* that if everyone I talked to came through, I’d be making about 500K a year, but lets face it, I’d be blowing smoke (or smoking crack).

    To me, when I hear the music or software industries whine about “lost sales” that’s all I hear: whining and greed. With a bit of fear of change thrown in. The music industry is the worst, of course, and every time I hear about a new RIAA lawsuit I become less inclined to buy their product (to a degree where by now, I think they’d have to pay me to take it). If they want my business, they’re eventually going to have to find a way to compete with other providers in the marketplace and not the courtroom. If they can’t do that, well, heck, I guess the problem is in either the marketing or the business model. (hint: I’ll never buy another cd again. period. I don’t need the plastic disc. If they can sell me the file at a price that works for me, without DRM, they’ll get my cash).

    I have a wee bit more sympathy for people who face physical counterfeiting of products but again it’s really a marketing or business model problem. Zippo can either sell $40 lighters because they’re “the best,” “the original,” “the styliest,” or they can get the same factory that got raided to make them for a price equal to what the counterfeits go for and sell them as “the cheapest.” They have enough brand capital to lock in a good share of the market for either luxury or cheapo lighters, but whichever they choose they’re going to lose some sales to the people who want the other. Or, if they were really clever, I guess they could do both under different names and hope no one ever found out.

    Ultimately, “lost sales” is BS. It’s a failure to grasp reality, not a failure to grasp profits.

  2. To a certain extent, the fake products in China are a free form of advertising. Those who buy fakes in China typically cannot afford the real thing anyway.

  3. james lee says:

    i have never seen it explined like this but this is well put. I love your blog and the information.

  4. Ahmad Cheragh says:

    Thanks for the very useful view over the “lost sales” concept. Now consider the below example:
    A customers calls your dealer for 5 units of product A, the dealer calls the factory but only 2 units is in the factory stock. The dealer looks in other dealers and shops for the product to make up the order. In one hours several dealers and shops will look for a few units of product A as they reckon there is a good order in the market. Dealers put orders for the factory to produce product A. The total of 50 units of product A is received to be produced and the lead time is 8 weeks. 8 weeks later when the logistic and production process has all been done and the products are ready for delivery only 5 to 10 units are really sold and the remaining is actually incorrect sales forecasts.
    This problem is not rare I have faced it. The accuracy of lost sales calculation is also of importance not to miss guide the management.
    please provide formula and models for …

Comments are closed.