3 ms·
> 80/20 rule: 20% of customers were responsible for 80% of revenues > Sometimes you will see a 90/10 relationship or a 75/25 split. It's funny that his exampl
by computator 3y ago
> 80/20 rule: 20% of customers were responsible for 80% of revenues
> Sometimes you will see a 90/10 relationship or a 75/25 split.
It's funny that his examples add up to a nice round 100, but it doesn't have to. The number of customers and percentage of revenue are two independent measures, and they don't have to sum to 100. You could have 20% of customers giving 99% of revenue (a 99/20 relationship) or 10% of customers giving 60% of revenue (a 60/10 split). I assume the author is well aware of this, but likes his examples to sound like the well-known 80/20.
- hantusk 3y agoThe Pareto principle is about the converse as well
- tomsmeding 3y agoThough for any such distribution, there is a point on the curve where the numbers _do_ sum to 100. Surely 0.0001% of customers provide less than 99.9999% of revenue, but surely 99.9999% of customers provide more than 0.0001% of revenue. Somewhere in between, assuming continuity(*), there's a ratio that sums to 100. Not to detract from your point at all — just a curious observation. * In a way assuming continuity is weird because the underlying numbers are discrete. But in approximation it's fine here if the number of customers is large enough, I'd say.
- FabHK 3y agoNicely put. Another way of seeing this: any Lorenz curve will intersect the “other” diagonal (from top left to bottom right). https://en.m.wikipedia.org/wiki/Lorenz_curve https://en.m.wikipedia.org/wiki/Lorenz_curve
- bmacho 3y agoSo the 80/20 rule is about where it happens, and not about that it happens! I've never realized it before! Okay, you are probably not really interested in the point where they sum up to 100, and you just want the 80% of the profit, whether the distribution gives 80/10 or 80/78 (where all customers give you almost the same profit, revenue and cost). Still a useful rule.