Yes, we can predict the future (with a certain probability)
Ever since I got into marketing or more concretely—driving users to my own online projects—I have started to notice a fascinating thing: that it is possible to predict how many visitors or leads you will have in the future.
To me it is fascinating because I am used to rigid technical systems and that they are deterministic. You run it — you get the same result every time (or an error).
With people it always seemed to be unattainable. How can I possibly predict what other people would do?
Turns out you can, if the sample size is big enough. And it shouldn’t be even that big, really.
A concrete example: say I get ten thousand of views on Google every month. Out of those I get about 100 clicks — 100 potential customers.
Out of those 100 customers, 30% fill out the form since they are interested (they clicked the link already). That’s 30 leads.
Now let’s say that on average (make note of this!) out of those 30 leads 2 of them get onboarded. That is 2 / 30 = 6,67% chance.
And thats it, that’s your conversion rate. Now you can take this percentage and propagate it all the way back to know how much views on Google you need on average in order to increase those 2 monthly clients to 10, 20 or 100.
At this point you have essentially pivoted the problem. You can focus on improving views on Google which would drive your business.
Of course, in reality there are infinite amount of variables, however in general this principle will work. If you don’t change more than one variable at once.
This way you can effectively predict the future, calculate your required time and money investment and understand if it is at all feasible.
And once you building such inbound system, then you can do the same thing for business metrics: MRR, ARPU, etc. To be honest, I’m getting addicted to building all of these prediction models and graphs. And it is a special type of satisfaction when your simple “30% per month” growth model actually manifests itself in real life.
Just remember: the quality of data matters, so it is best to do fewer but more surgical adjustments over a longer period of time. Not everyone is cut out to the long game.