Basic Stats 4 Entrepreneurship

[Velo]

If you learned about (or intuitively understand) binomial distributions already, congrats.

If not, do the following:

1. Go here: Binomial Calculator
2. Put 0.01 in the box "Probability of Success on a single trial"
3. Enter 100 in the box below (Number of Trials)
4. Enter 0 in the box below (X)
5. Click Calculate

Look at this result -

Cumulative probability:
P(X = x)

You'll notice the figure is ~0.366. Translated back into normal human language (although this is basic stats and isn't hard):

If you send out 100 marketing offers, and you think the "true success rate" is 1%, there's a 36.6% chance you get 0 "success" cases (i.e. 0 conversions).

If you're paying attention at all, what this means is 100 is too small of a sample size. The probability you get nothing is quite high.

Try changing 100 to 200, 300, 400, etc. If you put in 400 as number of trials, your chances of getting 0 successes is now about 1.8%. Your chances of getting 1 OR MORE successes (which is what you're after) is now 98.2%.


Let's say you did run 400 trials in real life, and you had 0 successes (i.e. conversions) - the probability of this happening is extraordinarily slim so you can safely assume it's because your conversion rate is not 1%...it's lower.

You can play around with the "probability of success on a single trial" figure...for example, if your true conversion rate is not 1% but 0.1% you'll need approximately 2300 trials before the probability of getting 0 successes drops below 10% (which is what I consider as roughly a "safe margin of error")...

...and then of course your next job is to figure out how to increase that abysmal 0.1% figure.

 

Dislike ads? Become a Fastlane member: Subscribe today and surround yourself with winners and millionaire mentors, not those broke friends who only want to drink beer and play video games. :-)