Startups Stack Exchange Archive

Mathematical Model for user growth

After laying out the initial ideas and concepts we’re now in the phase of evaluating if our business model is valid.

A part of our monetarisation plan is pay-per-click/sale and so the number of users is a crucial number.

Q1: What mathematical models/functions give a realistic estimate of the development of the number of user over time?

So far we considered exponential growth but it does not account for saturaion effects.

Q2: Is the logistic function better suited? The problem seems to be that it does not account for decay.

Answer 769

A1: None.

A2: No.

You’re assuming that all businesses are somehow the same. There are companies offering services that are multiple engagements per day (email, for example), and others that are once-in-a-lifetime (or very infrequent) such as births, marriages and deaths. The way in which people engage with a business, before, during and after purchase is, if not endlessly varied, then at least very broad.

Add to that stuff like post-purchase evaluation, word of mouth recommendation, social media, etc…

I work on one eCommerce site that has a large repeat buyer segment. I work on another where there is no intrinsic repeat business segment at less than a five year interval. They’re both owned by the same company.

There is no simple formula describing all these behaviours. If there was a formula, it would need parameters to describe the nature of the business.

Answer 776

Honestly, more often than not, having to worry about the carrying capacity of a given campaigns run is a problem people wish they had, not that they have, and your time would be better spent on achieving growth, not forecasting it.

To be clear, I’m not saying to ignore the need to understand the limits of a given market in theory, but that forecasting saturation effects is not worth the time; say this in part because true exponential growth often finds the means to grow far beyond what any reasonable forecast would ever account for in my opinion.


All content is licensed under CC BY-SA 3.0.