# State of Innovation

## Invention – A Financial Analysis

Jacob Schmookler, an economist and author of Invention and Economic Growth, developed a financial analysis of the invention process.[1] The main point of this mathematical modeling of the invention process was to show that the probability of any invention being created is related to the size of the market for the invention.  I intend to present a model of the cost of inventing compared to creating me-too products.  I am not a fan of mathematical models for explaining most economic effects, because the terms in the equation are either unmeasurable or vary in an unpredictable manner.  As a result, I think these mathematical models give the impression of the accuracy of a physical science, which they clearly do not provide.  This can lead to logical errors.[2]

Despite this, I believe a simple mathematical model of the invention process will illustrate some important points.  In addition, some people understand concepts better when presented in a mathematical model.  Here is my model for the costs associated with introducing a new product based on an invention and me-too product:

Ci(n) = (Inv + Mi)/P + NRE + PC*n + OH*n   (New Product based on invention)

Cmt(n) = NRE + PC*n + OH*n  (Me-too product)

Where Ci is the cost of creating a product for the owner of the invention, Inv is the cost of creating the invention, P is probability of that the invention will succeed in the market, Mi is the incremental marketing and sales cost of introducing a new invention, n is the number of products that have been produced, NRE is the nonrecurring engineering cost of setting up production, PC is the production cost of the making n products and OH is the cost of overhead for producing n products, Cmt is the cost of creating a me-too product.

The reason I add the probability that the invention (P) will succeed is that not all inventions are successful.  An economist who wants to capture all the costs associated with introducing a product based on a new invention has to include this probability to determine the true cost of inventing.  This probability will vary based on the type of invention.  For instance, line extension inventions are much more likely to succeed than inventions that create whole new markets.  An example of an invention that created a new market was Webcrawler, which was the first full text web indexing search engine introduced in 1994.  On the other hand adding image or video searching to Google is a line extension.

The cost for marketing and selling a product based on an invention (Mi) is separate from the cost of marketing and selling a me-too product.  It takes significantly more money, time, and effort to sell a product based on an invention that is creating a whole new market than a me-too product.  Any sales person who has tried to sell a truly unique product knows that it is much easier to sell an existing product or a me-too product because you do not have to explain the value of the product, how the product works, and why the customer would want the product.  A true me-too product can be sold mainly on price.  A line extension product takes less marketing and sales effort than a revolution product.  Large companies tend to focus on line extension inventions because it reduces the risk that the product will not succeed and reduces the cost of marketing and selling.  Many start-ups sell through marketing channels in order to reduce this cost.

I include the cost of selling, advertising, and marketing of me-too product in overhead.  Once a product based on an invention is well known, then it will incur the same cost as a me-too product of selling, advertising, and marketing.  I believe this is an accurate characterization.  Non-recurring engineering (NRE) is the same for both the me-too product company and the inventor company.  The reason for this is that me-too products will incur approximately the same cost of setting up production as the owner of the invention.

The values of these variables will vary based on the type of invention involved, the type of market in which the invention is sold, and the point in time the product is introduced.  This model is not exact.  For instance, overhead (OH), production costs (PC), and marketing cost of the invention (Mi) should all be functions of the number of products sold (n).  Production costs usually decrease with the number of products sold.  Marketing costs of the invention (Mi) should be spread out of the first X number of products sold.  In addition, the total marketing cost of the invention (Mi) should not be included for failed products based on an invention, since the owner is likely to kill the project earlier and not spend as much as on a successful product launch.  There are probably other shortcomings of these equations.  However, certain facts are clear even with any flaws in these equations.  The cost of inventing increases the cost to the inventing firm over the me-too firm.  As a result, inventing is a market disadvantage without intellectual property.

Invention Law:  The cost of inventing increasing the expenses of the inventing firm compared to the expenses of the me-too firm.

There are only two common ways to compensate or incentivize inventors.  One is to provide the inventor with a property right (patent) in their invention.  The other is to have the government pay for the cost of inventing.  The first is consistent with a free market economy and has proven to be extremely successful.  The second is consistent with a command and control economy (statism) and has proven to be inefficient and political.

Intellectual Property Law: Inventing is a market disadvantage without intellectual property.

Now I will look at some specific scenarios to provide some insight to these laws.

Pessimistic Scenario

Scenario 1:

Consumer good sold through retail outlet

Inventing = New Market

Target Retail Price \$10.00

Cost of inventing (Inv) = \$100,000.00

Cost of Marketing Invention (Mi) = \$900,000.00

Probability of Success (Pi) = 0.1

Nonrecurring Engineering (NRE) = \$30,000.00

Production Costs per Unit (PC) = \$2.00

Overhead Costs per Unit (OH) = \$1.20

 Unit Inventor’s cost per unit Copier’s cost per unit 1 10030003 30003.2 10 1003003 3003.2 100 100303.2 303.2 1000 10033.2 33.2 10000 1006.2 6.2 100000 103.5 3.5 1000000 13.23 3.23 10000000 4.203 3.203 100,000,000 3.3003 3.2003

It is assumed the manufacturers are selling their products \$4.00, which is a standard double over manufacturing costs.  The 1/5th of retail price is a minimum necessary for the manufacturer to obtain a return that justifies manufacturing the product and selling it through a standard retail channel.  As you can see the inventor has to sell 100 million units (\$1B in revenue) in order to get within 10 cents of the same manufacturing cost as the me-too manufacturer.  The copier’s break even[3] point is somewhere between 10 thousand units and 100 thousand units while the inventor’s break even is point is over 100 times as many units.

It is likely that this scenario overstates the difference between the costs of the inventor and the copier.  For instance, the inventor is unlikely to spend the full cost of marketing the invention (Mi) for the other nine failed product.  In addition, the percentage of successfully launched products is based on the stated success rate of venture capitalists.  Most VCs state that they have one highly successful company for every ten investments.  They also usually have 2-3 other companies in the portfolio that produce moderate returns or losses.  Not all of the other companies are in their portfolio are a complete loss.

Optimistic Scenario

Let’s look at a much more optimistic scenario.  Let assume the probability of success (P) includes these moderately successful investments and lets also include the idea that the probability includes some line extensions which have a 70% probability of success or higher.  We will also move up the probability of success to compensate for the fact that the inventor is unlikely to spend the full cost of marketing the invention (Mi) on failed inventions.  I will make the wild guess that setting the probability to 45% will compensate these differences.  I will also assume that instead of taking \$1M to launch a new invention that it takes only \$100 thousand.  Part of the justification for this difference is that the inventor and other founders are likely to not take a salary until the company has significant revenues.  I will also lower the overhead significantly, because this is one of the big advantages of a start-up.  My optimist scenario is:

Scenario 2:

Consumer good sold through retail outlet

Inventing = New Market

Target Retail Price \$10.00

Cost of inventing (Inv) = \$10,000.00

Cost of Marketing Invention (Mi) = \$90,000.00

Probability of Success (Pi) = 0.45

Nonrecurring Engineering (NRE) = \$30,000.00

Production Costs per Unit (PC) = \$2.00

Overhead Costs per Unit (OH) = \$0.50

 Unit Inventor’s cost per unit Copier’s cost per unit 1 252224.7 30002.5 10 25224.72 3002.5 100 2524.722 302.5 1000 254.7222 32.5 10000 27.72222 5.5 100000 5.022222 2.8 1000000 2.752222 2.53 10000000 2.525222 2.50

In this scenario, the break even point for the copier is between 1000 and 10,000 units, while the break even point for the inventor just over 100,000 units.  The inventor is still at a significant disadvantage to the copier.  Some of this disadvantage may be offset by the first mover advantage.  However, if the inventor company is a start-up its first mover advantage is likely to be significantly offset by the established relationships of an established copier company.  In addition, the inventing company may sell more of their initial units directly (not through a retail channel) and their margins will be significantly higher for these units.

It is clear that inventing without intellectual property is a competitive disadvantage.  Large companies that invent can offset some of this disadvantage by using other competitive barriers to entry.  For instance, an established company can use its network of relationships to create a barrier to entry from start-up copier companies and may be able to use its relationships to provide some barrier to entry from other large established companies.  The empirical evidence is that established companies mainly produce incremental inventions.  This is because the invention process is risky and as an established company they often have less risky methods of providing incremental revenue or profit gains.

Start-up companies produce all the net jobs in America according to the Kauffman Foundation.   They are also the biggest producer of emerging technologies – see Do Individual Inventors and Start-ups Invent Anything Important?.  Advances in technology are the only way to increase our real per capita income.  We need to encourage investments in inventions, if we want to leave our children a better world than the one we live in.  Technology start-ups need the incentive of property rights in their inventions (patents) in order to justify the investment in these companies.

[1] Schmookler, Jacob, Invention and Economic Growth, Harvard University Press, Cambridge Massachusetts, 1966, pp 113-115.

[2] For instance, the measurement of GDP is said to be consumer spending plus investment plus government spending plus exports minus imports.  This equation leads to the logical error of assuming that consumer spending and government spending results in an increase in the output of a nation.  The reason this is a logical error is that people confuse the cause with the effect.  Consumption does not create goods and services.  Production creates goods and services, which is related to the consumption of good and services.  An engineering analogy is that temperature is often measured by determining a change in the resistance of a resistor.  If I change the resistance being measured by adding resistor in series with the thermistor this does not change the temperature of the environment being measured.  This is what economists are arguing when they suggest that increased government spending will cause the economy to grow.  Government spending does not create any new goods and services; it just either consumes production or transfers the return for production from one person to another.  Similarly, consumer spending is a way of measuring production.  Artificially increasing consumer spending does not increase production.  For instance, giving people income tax rebates when they never paid any income tax does not increase production, it just steals the productive effort of those who do pay taxes.

[3] The break even point is when the cost per unit is equal to the sales price per unit.

## 7 Comments »

1. […] This post was mentioned on Twitter by Peter Meza, Dale Halling. Dale Halling said: Invention – A Financial Analysis: http://wp.me/pwxHH-gm […]

Pingback by Tweets that mention Invention – A Financial Analysis « State of Innovation -- Topsy.com | August 13, 2010 | Reply

2. Dale,

I’ve been struggling mightily and for quite a while in trying to articulate a ‘something’ that I felt was wrong with your above, simplistic financial model (while of course paying due homage to your disavowal of mathematical models –which I agree with– and all the work you did in running those spreadsheets).

What I’ve come up with thus far has the stench of deep irony because you are pioneering this notion of modeling the financial aspects of first-to-invent and I am just another “me-too” frog who is trying leap ahead of you with the hindsight benefit of your backdraft.

Well, that’s kind of it in a nutshell. My second frog’s criticism of your model is that generally are no “me-too” pure copyists. Instead there are obviousness type or even non-obviousness type leap-froggers who jump ahead of the first to invent.

The obviousness type leap-frogger makes merely an obvious improvement over the pioneer’s product, but nonetheless, the leap frogger then has a big advantage in the market place because the pioneer has to re-tool just to catch up or try to do the next frog leap forward.

The non-obviousness type leap-frogger, as the latter name implies, makes a nonobvious –and perhaps patented– improvement over the pioneer’s product, and that gives the second type of frog an even greater advantage in the market place.

Perhaps this is why we see a consistent pattern where first to market loses.

The better financial model might ask, what damages are due to the pioneer to make him whole by placing him on equal footing with those who frog leap ahead of him because, as Sir Isaac Newton once said, their starting point was on the shoulder of the giants shrugging beneath them?

Just a thought from one frog to the next.

Comment by step back | August 16, 2010 | Reply

• Stepback,

While in general most companies are not pure me-too competitors (counterfeiters), they certainly exist. There are plenty of companies that are me-too with respect to part of the technology they incorporate in their competitive products. This is fine if they pay a royalty or some other compensation to the inventing company.

Technology leapfrogging is encouraged by the patent system. This is the point of the disclosure requirement. The patent system ensures that this is not a winner take all situation. For instance, Edison certainly benefited by Swan’s incandescent light bulb. Swan’s light bulb was economically unfeasible. Swan received a patent in England and when Edison started selling light bulbs in England Swan sued him for infringement and won. This system ensures that we do not have a winner takes all situation. Swan is compensated for his contribution, even though he did not create a commercially feasible light bulb. Note that Edison probably argued he was totally unaware of Swan’s earlier light bulb.

Comment by dbhalling | August 16, 2010 | Reply

• Dale,

You and I are generally on the same page with regards to patents and what they are supposed to do –i.e. level the playing field so that the second-comer to the market place does not have an unfair advantage. I thought that was what you were trying to show with your financial model, namely, that if the higher costs of the first to invent are NOT offset by patent rights, then the second comer (the me-too guy) has unfair advantages.

That said, for many inventions, costs scale nonlinearly with increases of volume (n) due to economies of scale. So you can probably tweak your first model by introducing n-squared factors. For example, what if the me-too or leap frogging 2nd company can immediately go into mass production once the first guy has demonstrated the larger market exists? The lines of your tables should not be read straight across as if the first inventor and the me-too competitor are going tit for tat on value of n. Usually, the copyist will be doing 10*n versus the originator’s 1*n production line.

________________
I’ve come here not to drown out your fledgling crop of an idea but to irrigate it.😉

Comment by step back | August 16, 2010

3. Stepback, I agree with you that making certain variables of the equation are actually functions of n or n squared. I think we both agree that these function would be highly dependent on the particular market, so I think they would add a false sense of certainty. There is a difference between resolution and accuracy. This is widely known in metrology, but I fear not understood in economics or finance.

I completely agree with you that reading the charts straight across is misleading. That is why I mentioned that the inventor might have a first mover advantage. In general, I think companies generally steal markets not inventions. As a result, they will wait to see if the market for the invention (product embodying the invention) is large enough before copying the invention.

Comment by dbhalling | August 16, 2010 | Reply

4. […] a new product embodying an invention.   I discussed this cost in Invention- A Financial Analysis https://hallingblog.com/2010/08/13/invention-–-a-financial-analysis/.  This cost is the variable Mi in the equation I developed as part of the financial analysis.  […]

Pingback by The High Cost of Marketing and Selling an Invention « State of Innovation | September 6, 2010 | Reply

5. […] first mover advantage, which intrigued me because I have argued essentially this point in my post, Invention – A Financial Analysis .   One of the responses suggested the book, Copycats: How Smart Companies Use Imitation to Gain […]

Pingback by More Evidence that Stealing Inventions is a Business Strategy « State of Innovation | October 7, 2010 | Reply