From time to time people appear on the Business of Software discussion forum and ask “How much money will I make if I create a software product and sell it from a website?”. There are so many factors at play it is very difficult predict, but it isn’t entirely hopeless. We can look for inspiration to the astronomer Frank Drake.

Frank Drake wanted to know many alien civilisations there are currently in our galaxy that we could potentially talk to. That’s a pretty BIG question. But Drake broke it down into a number of smaller questions in his famous Drake equation :

N = R x Fp x Ne x Fl x Fi x Fc L

Where:

- N is the number of civilizations in our galaxy with which we might expect to be able to communicate at any given time
- R is the rate of star formation in our galaxy
- Fp is the fraction of those stars that have planets
- Ne is the average number of planets that can potentially support life per star that has planets
- Fl is the fraction of the above that actually go on to develop life
- Fi is the fraction of the above that actually go on to develop intelligent life
- Fc is the fraction of the above that are willing and able to communicate
- L is the expected lifetime of such a civilization for the period that it can communicate across interstellar space

These are still tough questions to answer. We know that the final value is >= 1 (assuming the success of reality TV doesn’t exclude us from being classified as intelligent) but many of the factors are highly uncertain. In particular there is huge uncertainty over the civilisation lifetime factor L. The equation was first derived at the height of the cold war, when many scientists understandably took a very pessimistic view over whether technological civilisations could avoid blowing themselves up. The more optimistic thought that advanced civilisations would disappear inside Dyson spheres after a few millenia. But at least astrophysicists and xenobiologists can have much more meaningful debates over these individual factors than they can by just arguing about the final estimates. Drake himself came up with a value of N=10.

In the same vein, here is my equation for how much software you will sell in any given month:

S = P x N x Ft x Fb

Where:

- S is the sales per month in units of currency
- P is the price of a licence (in the same units as S)
- N is the number of unique visitors to your website per month
- Ft is the fraction of visitors who try your software
- Fb is the fraction of visitors who buy the software after trying it

This obviously assumes that the Internet is your main sales medium, you don’t have to give many refunds, people will always download and try the software before buying and a raft of other assumptions. But it’s a start. We could break it down into a lot more factors, e.g. to include the fraction of potential customers who didn’t use a crack, but the advantage of the above factors is that they are simple and relatively easy to measure.

Drake only had one data point to work from for most of his factors – our own earth. Similarly a we don’t have a lot of data to work from, as software vendors tend to be rather coy about their sales data. As a data point the stats from my own table planning software site show that approximately 10% of visitors download the software and approximately 10% of these go on to buy a licence. This means that only around 1% of visitors to the site buy a licence. This percentage (Ft x Fb) is known in the jargon as the ‘conversion ratio’. 1% seems really low, but I understand from various sources that 1% is actually quite respectable. Afterall it includes all the people who ended up on my site while searching for “diy picnic table plans” or “multiplication table printing” (if you only include visitors that visited the home page the conversion rate goes up to a more respectable 4%).

So now you can plug in your own numbers and get a rough idea of what your sales might be. Or, perhaps more helpfully, you can re-arrange the equation to find out how many visitors you need per month:

N = S / ( P x Ft x Fb )

If we plug in: S = $10,000/mo, P = $100, Ft = 10% and Fb=10%, we get N= 10,000 unique visitors per month. You can then ask yourself if 10,000 unique visitors per month is achievable for your product.

Some points to note:

- The price will have an effect on the conversion percentage. But decreasing the price will not always increase the conversion percentage. Because price sends a signal about the quality of your product, it may even have the opposite effect. This is a complex topic and beyond the scope of this article.
- Not all visitors are equal. Highly targeted visitors going to a well designed web site with an excellent product might have conversion ratios as high as 5-10%. Poorly targeted visitors (e.g. coming from badly worded online ads or purchased from dubious ‘traffic programs’) will have conversion ratios barely discernible from 0%.
- Just like the Drake equation, if any of the factors are 0, the result will be zero, no matter how large the other factors are, i.e. it doesn’t matter how many visitors you get if the link to the purchase page is broken.

Feel free to leave a comment with your values for for P, N, Ft and/or Fb.

So what is my value for N? I’m not telling. But, I don’t think any of them are aliens.

MarkOne of the most confusing but relative posts I have ever come across, you made my head hurt! – But it makes a lot of sense.

Gavin BowmanYou finally took the plunge and started a blog :), nice one Andy.

Steve LordMaybe you shold run your formula aganst some listed web companies with published accounts?

Google?

Just a thought.

;)

Matt BreckonNice to see you’ve started a blog Andy – you’ve certainly got a lot of relevant and real experience for the MicroISV crowd. Good luck.

Andrey ButovAnd Andy comes out of the gate running strong…

What? No “niche market” quotient?

OK. I’ll play: P = 49.95, N = 5000, Ft = 0.89, Fb = not telling — fairly accurate (but more interesting than anything else) — had to adjust P to factor in fluctuating unit price due to resellers and publishers.

– Cheers!

successfulsoftwarePost authorAndrey,

0.89 (=89%) or 0.89% of your unique site vistors download the trial? If its 89% thats very impressive.

Andrey Butov89% of the unique visitors download the trial as averaged over the last 3 months. It’s a unique product for its market.

Pingback: weekly linkdump « блог разработчиков « developers.org.ua

Bob WalshExcellent post Andy!

davidpaulrobinsonShudder…. math…..

TheBoyKenDoesn’t Drake’s equation assume that there can only be one alien civilisation (worth communicating with) per planet?

For instance, anyone who has watched the newer Star Wars films will know that Naboo hosted 2 such civilisations (ignoring the fish). Similarly Earth might be said to host humans (if we’re intelligent), dolphins, and of course dogs.

Just a thought…

Andy BricePost author>Naboo hosted 2 such civilisations

A factor of 2 error is likely to be lost in the noise. ;0)

Ps/ You forgot mice (ref: The Hitchhiker Guide to the Galaxy)!

Ian TurnerI’ve been playing around with the logisitic function (http://en.wikipedia.org/wiki/Logistic_function) to try and predict hits to my blog recently and so far it’s actually prooved to be surprisingly effect. It is a rather large simplification but the equation is easy to work with and seems to give a good approximation of future number of hits provided the environment you are working in stays the same.

p.s. I’ve never thought of the Gungans as an intelligent race.